Breaking News
Comments
Kirby2008
User Rank
Rookie
There are already some Low Cost PnP Projects
Kirby2008   6/18/2014 6:26:36 PM
NO RATINGS
Good day,

There are already some low cost PnP projects started and so perhaps it would be best to work with these groups, as opposed to starting another one.  Here are some that I have found:

http://hackaday.io/project/963-%24300-Pick-and-Place-%2F-3D-printer

http://briandorey.com/post/DIY-Pick-and-Place-V2-Project-Complete.aspx

https://github.com/openpnp/openpnp

http://www.placecomponents.com/

http://www.vbesmens.de/en/pick-and-place.html

Cheers,

Sam

Aeroengineer
User Rank
Blogger
Re: There are already some Low Cost PnP Projects
Aeroengineer   6/18/2014 8:06:20 PM
NO RATINGS
Thanks for the comment.  I am sure that there are things that can be leveraged off of these efforts, but there are some fundamental aspects that we are seeking to improve upon.

 

One of these areas is to completely eliminate the style of feeders that are used.  There are two reasons for this.  First the cost.  Feeders drive significant cost into the designin a few ways. Their mechancial nature makesthem expensive to produce, and because you need to have a lot of them for building any sort of board, this further increases cost.  The other issue with this style of feeder is that they significantly increase the footprint of the device.  I think that we as a group can do better while only suffering some minor tradeoffs

 

This is not to discourage posting other references.  We are always open to reference material.

 

Adam

salbayeng
User Rank
Rookie
Re: There are already some Low Cost PnP Projects
salbayeng   6/18/2014 10:19:30 PM
NO RATINGS
Hi Kirby, 

That's an impressive list of suppliers, I particularly like the PP4 unit from www.vbesmens.de 

One of the things to consider when scratch building your own is your effective labour rate, if you spend 200hrs of effort to save $400 of stuff you could have bought, that's a wage of $2/hr you are paying yourself.

I work closely with a colleague with a properly setup fab shop, so I always try to setup my designs to improve productivity (e.g. use only 10,22 and 47 Decades for R and 10's for C's). What we used to call "Muntz'ing" is now "KanBan". 

As others have noted the biggest Archilles heel are(is?) the feeders. Even a small pcb with 60components will have 50 unique parts, half will be 8mm tape, so you need to have perhaps 80 feeders available. This is expensive and bulky whichever way you do it. To keep unit size down, you might consider using cassettes that hold say 20 strips, and then do 3 cassette swaps during a job. You could then leave the strips permanently in a cassette, bag them and store in a library. The biggest issue with home made feeders is handling the cover tape, solve this problem first before embarking on the rest of the feeder design.

I would definitely incorporate the paste dispenser, I would use 1ml syringes (What I use for hand pasting) to get much better control of dispensed volume, and much less wastage ( A 5ml syringe will go hard before you use it all) 

Incorporating the oven into the project is a silly idea for several reasons:
  • the FUMES will give you bronchitis and cover everything in your lab with an icky film
  • the warm temperatures near the PCB's being placed will dry out the paste (make it less tacky) (parts will fall off) 
  • you need somewhere to put all the cooked boards while cooling down
  • you need good temperature control, really only done with a dedicated oven

So unless you have an exhaust system set up, you really need to cook the boards outside. I just use a $40  Kmart toaster oven plus a thermocouple attached to PCB, cook at 200C till 160C on thermocouple, then maximum until temp = 205C (210C if large inductors on PCB) , then oven off, open door , slide out tray, wait till 160C remove pcb.

 

Aeroengineer
User Rank
Blogger
Re: There are already some Low Cost PnP Projects
Aeroengineer   6/18/2014 11:06:40 PM
NO RATINGS
I think that you have the right idea there, and somewhat more along the line of some of the concepts that we have been throwing around as we were preparing this post. 

 

My thought is that because we are talking low volume here for boards to just use pieces of cut tape.  Most of my projects I am purchasing cut tape already, not full reals for these components.

 

I also like the recommended size for the syringes, this will promote more accuracy in dispencing the solder paste as it will take more stroke to dispense the same amount of paste.

tb100
User Rank
CEO
Desktop pickup and place
tb100   6/18/2014 6:44:09 PM
NO RATINGS
The way my brain works: when I saw "desktop pick and place machine" I immediately thought of a robot that you put on your desk which cleans up your desktop. The question is, after it has finished placing everything, will I be able to find anything?

Aeroengineer
User Rank
Blogger
Re: Desktop pickup and place
Aeroengineer   6/18/2014 8:07:16 PM
NO RATINGS
Glad to see that I am not the only one that functions well with a little clutter on the desk ;)

Aries1470
User Rank
Rookie
Interesting
Aries1470   6/18/2014 7:20:44 PM
NO RATINGS
Hi,

Well, here are my 2¢ worth.

For feeding, instead of having the huge things - aka feeders for large runs, why not implement an 'old style' dot matrix printer feeder, but with adjustable width spokes.

Then on the side, depending on the maximum pcb workable area, have an attachment, that will be a small oven, with side walls that can open and close, as to save space and energy.

I can input more, but... some others have referred to other projects.

Aeroengineer
User Rank
Blogger
Re: Interesting
Aeroengineer   6/18/2014 8:09:58 PM
NO RATINGS
No, please keep the input coming.  I think that you are seeing why we are branching out on our own.  There are some reallyunique things that can be done within what we are trying to do. 

 

Your feeder system for example is somewhat similar to one of the concepts that we have batted back and forth.

 

Being able to add the oven feature will also be able to really add functionality to this to be a one stop machine for turning small proto boards in say half an hour to an hour.

 

 

Thinking_J
User Rank
Rookie
Re: Interesting
Thinking_J   6/18/2014 9:16:01 PM
NO RATINGS
fun project..

Wish you luck ..

Having been in the industry for some time.. I would like to point out some of the most commonly over looked issues that a project of this type would have:

- component storage .. soldering sufaces on components do oxidize and plastic components absorb moisture from the air. Most hobbiest are not prepared to keep components in a low moisture atmosphere during times the system isn't running.

It only takes a few days of exposure to high RH.. then you have to "bake out" the parts. If you don't, the parts will "pop corn" during solder reflow (drive the moisure out too quickly).. a real issue for many thinner SMT packages

- solder paste knowledge. If you plan on building with smaller parts the solder stencil opening become very small (fine pitch BGA, etc..) the quality of the solder paste become critical - and it solder paste is expensive! Generally best practice is to keep refrigerated - too keep the flux from going bad. This is an issue regardless of the method of applying the solder to the pcb (stencil, direct deposit, etc..)

- For the desired desktop space mentioned.. your component feeder/submitting system will be the hardest part of the system's mechanical design. Tape and reel feeders can be small , but I don't think you could get very many components "on-line" in the space mentioned. Bulk feeder system can be small , but only work for passive components..

- Reflowing single, simple boards in the same space will have another problem. The impact of heat on the accuracy of the placement. I have a quality SMT line, and we cycle the machines for awhile - then calibrate their placement accuracy with a zero temp coeffient optical target based on the thermal expansion of the machine. .. and I only consider it good enough for 0201s.

Best to keep the furnace/oven separate.... regardless of if you can put it in.

For prototypes: I would recommend use of conductive epoxies instead of solder for your design goals... it would reduce or eliminate many of these issues.
 Easier to dispense, no high temperature excursions to subject components to, less sensitivity to oxidation on connection points.

Aeroengineer
User Rank
Blogger
Re: Interesting
Aeroengineer   6/18/2014 10:14:45 PM
NO RATINGS
Thanks for your detailed input.  As you mentioned the feeders and how ling parts are stored in that feeder can be problematic. This along with the size of the components that you mentioned will put some bounds on what one may want to do with this machine.  I am thinking that 0402s are going to be the target size for these components.  There are going to need to be certain limitations that are not typically encountered  for volume production, but then again, this is not geared towards that world.  One of the things that could be done if really necessary for moisteure is to have a drying routine.  I have not looked into it too much for electronics, but I know that for many dry composite materials, we do a 150 degree F for an hour to remove any moisture.  I know that I have seen some drying procedures for electonic components, I just have not had to deal with them yet.

 

As to the thermal issues that you mentioned, one of the things that I am looking to mitigate some of these things are the use of materials in the mechanicals that are matched to the Cte of FR4.  Standard Steel or Stainless Steel would be the materials of choice to match the approx 6-8 microstrain perdegree F of the FR4.

 

Please, though, keep the ideas coming.

salbayeng
User Rank
Rookie
Re: Interesting
salbayeng   6/19/2014 12:19:22 AM
NO RATINGS
@Thinking_J.

Wholeheartedly agree with all your points except the last (conductive epoxy).

I have actually used conductive epoxy with large area PV panel bonding. But maybe you've had better experience than
me.

I accept that the epoxy will have better dispensing properties (smaller beads, better wetting) than solder paste.
However for the hobbyist / small jobber there are some issues.

Cost: a 50g pot of conductive epoxy is ~ $400 vs a 250g pot of solder paste at $50 (both pots would produce roughly the same volume of joints)

Shelf life: Solder past is good for a year or more in the fridge, premix epoxy has to be delivered refrigerated and lasts maybe 3months, two pack epoxy is good for ~ 6months on the shelf, maybe 24hrs after mixing (depends on type),
you have to discard what what you don't use.

Curing: You still need to heat cure the epoxy, maybe 100C for an hour, not as critical as solder profiles, but slows down production.

Smearing: If you smear solder paste (e.g. due to bumping a component, it will just ball up, and if you smear across pads it will usually fix itself during reflow, epoxy would just make a hard to fix short circuit). Epoxy pays no heed to the solder mask.

Wicking: You can lay down a long thin paste line, and the solder from all over the pad will suck to the actual area where the pin makes contact. Secondly a slightly misplaced TQFP or fine pitch will suck itself into position due to surface tension.

salbayeng
User Rank
Rookie
Positioning / verification
salbayeng   6/19/2014 12:45:52 AM
NO RATINGS
Not sure if it's been covered earlier, but:

Another important issue is the alignment of parts as they are picked up, typically the parts on the tape will be anywhere with +/-0.5mm of centre and up to 5deg off alignment, so for parts bigger than 0805 one can rely on surface tension to straighten them when reflowed , but 0.65mm pitch could be be misplaced by one pad.

Early PnP machines use two plates that "squeeze" on to the part , simultaneously centreing and aligning in one axis, if these metal plates are gold plated and insulated, you can also check resistance/polarity of the part. Later machines used lasers to check the part was the correct size and right way up. Early machines could only rotate 0deg or 90deg (using a on/off pneumatic cylinder), to get 180deg you needed to "shuffle". Later machines can rotate to any angle. For the hobbyist a 0/90deg may be all that is needed (except you can't do a radial array or arc of LED's.) 

Verification: One strip of MLCC caps looks the same as all the rest, how do you mark them? How do you ensure you haven't mixed up two strips when placing on the robot? How can you tell the part hasn't somersaulted during pickup (SOD-323 and SOD-523 are bad for this) ? Should the pickup head read a barcode on the strip? Can you even measure the suction to prove there is a part there? 

salbayeng
User Rank
Rookie
Nozzles
salbayeng   6/19/2014 1:04:40 AM
NO RATINGS
Don't think anyone has mentioned nozzles yet.

I'm guessing you would plan on using luer lock type nozzles to keep costs down. So some issues here:

The taper fit of luerlock means the actual height might change depending on how hard you push it on, and different diameter nozzles would be different heights too, so some method of calibrating installed height is needed (maybe just drive down to a stage and see where the pressure changed on nozzle suction).

You could use a 16g nozzle to pick up everything from 0603 to a 44pin TQFP, otherwise you need to change nozzles, so need a nozzle rack somewhere, and a means of changing nozzles.

And related to picking up off the feeder, how do you avoid going too high/too low when picking up with the nozzle (too low can flip the part or its neighbour)? What if the nozzle snags on the tape (or more usually snags on the cover tape) ? Is there some sort of spring to avoid damage?

Aeroengineer
User Rank
Blogger
Re: Nozzles
Aeroengineer   6/19/2014 9:44:52 AM
NO RATINGS
I have given some thought to this, and I am not too sure that I have a solution that I completely like. 

 

I would love to hear some more thoughts from the readers on this area.  I can say that this is one area that may not be as easy as it might first appear.  The other professional pieces used in pick and place machines cost much more than the (order of mag or two).

 

I had not only thought that there would be a need to calibrate the depth, but also the runout.  One thought is to have it so that the head would have a spring stop in place of a hard stop.  This would mean that as the component comes to the surface, you could even command it to go past theh surface, but a spring in the system would prevent stalling of the motors.

Aeroengineer
User Rank
Blogger
Re: Positioning / verification
Aeroengineer   6/19/2014 9:52:30 AM
NO RATINGS
These are all interesting things to think about.  I will give some thought to it.  I am leaning on doing a full 360 rotation, but there are some electromechanical challenges.

 

The suggestions on how to do polarity and centering are interesting.

 

On the suction, we might be able to put a pressure sensor in line to watch of a delta after picking up a part, and the delta after placing the part.

 

Thanks for all the great points to ponder.

Aeroengineer
User Rank
Blogger
Re: Positioning / verification
Aeroengineer   6/21/2014 3:09:37 PM
NO RATINGS
It is a whole lot easier to reply to comments when I am not working 12-14 hours a day. 

 

I wanted to elaborate a bit on some of the things that you have brought up in this comment.  The alignment of components will be critical to be able to push into the smaller component sizes and pad spacing. 

 

I see this as a two pronged effort.  The first effort is to ensure that the electromechanical design can actually provide the accuracy and precission that is required for the task.  I have seen a lot of systems that are not designed for this, and they suffer.  Even in the 3D printer world, many of these groups only speak of the min step size, but not about the repetability of their process.  I think that this causes a lot of confusion and is perhaps borderline false advertising. 

 

The other prong of this effort is to look at how to integrate a camera vision system to help in the alignment of the component. 

Aeroengineer
User Rank
Blogger
Re: Interesting
Aeroengineer   6/19/2014 9:31:23 AM
NO RATINGS
It is interesting to see some of the facts on solder epoxy.  I have never used it before.  I think that for familiarity it may be best to go forward with standard solder past, but we can keep the other in the mix.  I think that many of the items may be interchangable.  The main issue that I worry about with conductive epoxy is the fact that solder mask will not give any help in reducing shorts.  This will mean that not only will precission despensing be a requirement, but also placement of the epoxy and components as the tolerance stack is now dealing with three different steps.  The solder past gives a much larger tolerance on position, volume dispensed, and component placement.

 

How are these issues controlled with conductive epoxy?

Thinking_J
User Rank
Rookie
Re: Interesting
Thinking_J   6/19/2014 4:48:30 PM
NO RATINGS
@Salbayeng.

Your comments show much experience. Agreed - conductive polymers are not going to replace solder for every application.

Like so much of life... the devil is in the details.

Cost: I can get both solder and conductive polymers cheaper than what you quoted... and what ever it is today, will be different tomorrow. Often the difference is based on where you are.

Shelf life: similar observations to above (we don't like to keep solder for more than 6 months and never reuse by placing excess back into storage)

Curing: agree.. some curing heat is generally required. But much lower temps than solder. Especially the higher temps of ROHS solders.

Smearing: correct can be easier to be corrected when using solder. But always a bad situation. Even solder balls create problems.

Wicking (surface tension): this depends on the pcb surface finish. OSP (organic surface protectant) type finish,  solder doesn't generally wet beyond where you put it. ENIG or HASL variations generally the solder will wet entire surface.
And, yes , the surface tension can re-align a component with solder. Some designs and facilities depend on this characteristic of solder. But if often this creates as many problems as it solves (tomb stoning, floating off center because on thermal issues, solder balls floating around in the assembly).  Better solution: just put the part where it belongs... and expect it to stay there. And don't bump it until it is cured or reflowed! I have seen articles by "experts" claiming the lack of wetting and automatic re-alignment with many new ROHS solders helps them reduce solder bridging! .. go figure...everyone seems to view this differently.

An item many have ignored: ROHS vs NON_ROHS compliance on components, specifically the surface finish of the connecting points. It is generally not recommended to mix soldering materials and component types. This is not an issue with conductive polymers.

 

zeeglen
User Rank
Blogger
Re: Interesting
zeeglen   6/19/2014 2:13:08 AM
NO RATINGS
@Thinking_J I would recommend use of conductive epoxies instead of solder for your design goals.

I would not recommend conductive epoxy for transient voltage suppressors or current sense resistors.

Aeroengineer
User Rank
Blogger
Re: Interesting
Aeroengineer   6/19/2014 9:37:07 AM
NO RATINGS
Is this due to the higher resistance of the epoxy compared to solder, or due to variability in the materials?

Thinking_J
User Rank
Rookie
Re: Interesting (limiting uses of conductive polymers)
Thinking_J   6/19/2014 3:52:11 PM
NO RATINGS
Zeeglen: I understand your basic concern: thermal / electrical performance of conductive polymers comparted to solder (primary metal: tin .. assuming ROHS compliance).

You are correct .. conductive polymers are not appropriate for everything. But I was directing my thoughts to whom I thought would be using a table top machine with all the processing done at one station... for prototying or very small production runs.

Generally,

- Thermal conductivity:

Yep.. metal is better. But most applications don't approach the limits of either material.

- Resistance:

The differenences over 0.0005-0.001" thickness in the junction typically larger than 20 sq mil, are minor. Assuming use of quality conductive epoxy intended for this application. Unless you are sensitive to variations of less than 0.00001 ohms (you might be).. I don't think it will matter much. Many are not aware of the variations they may be experiencing with solder on current sense resistors. Good pcb layout methods are the primary concern for current sensing.

- Mechanical differences:

Many are not aware of the sensitivity of the some of the components to variations in connection process to the mechanical strength of the joints. Ceramic capacitors with high plate density (capactitance per volume) often demonstrate large deviations in quality due to this.

Example: the measured shear strength of a soldered X7R chip capacitor varies significantly with capacitance for a given size.
 I have measured a range of 2-17lbs for single location/size. While the same location/size with conductive epoxy showed very little standard deviation from 7 lbs of shear strength.

It was determined the capacitor's end platng (based on density) was the primary reason for the large std dev in the soldered joint performance. And the primary reason for this sensitivity to the quality of the capacitor end plating was exposure to higher temperatures and sensitivity to flux activation level (all mild).

Because of the physical space to work with and the minimum capacitance over temperature extremes and voltage required.... we had to work with capacitors that were (at the time) near the limits of how much capacitance we could get in a given volume (size).

I couldn't find a capacitor manufacturer that didn't display some form of this problem. I couldn't tolerate less than 5lbs shear strength on ANY product (avionics -55.+125C rating with high shock and vibration requirements). So, the ONLY solution at the time was to replace solder process with high temp rated conductive epoxy.

Connection junctions are typically a very small fraction of a square. (bulk resistance of a given material, generally specified in "squares" with a given thickness or resistance per cm)

Which is a very, very different animal compared to conductive epoxies vs copper used for traces (many squares) , often an issue when using bendable circuitry (copper vs conductive polymers).

 

Aeroengineer
User Rank
Blogger
Re: Interesting (limiting uses of conductive polymers)
Aeroengineer   6/19/2014 4:12:03 PM
NO RATINGS
Once again, thank you for the detailed information.  This provides interesting food for thought not just for this effort, but for efforts to manufacturer other PCBs that are in production.

betajet
User Rank
CEO
Ariel Electronics Circuit Writer
betajet   6/19/2014 4:50:01 PM
NO RATINGS
Regarding conductive epoxy, here's a repeat of a comment I made last year:

Around 1990 there was a promising attempt to use CNC technology to make circuit boards by Ariel Electronics (California).  They created a gadget called the Circuit Writer which extruded conductive plastic wires onto a substrate, basically a 2-D plotter with an extrusion head.  I actually visited Ariel and saw a Circuit Writer working.  I don't think the technology got anywhere, but maybe it was just ahead of its time and with newer 3D extrusions this could be done practically.  For more info, Google "ariel electronics circuit writer".

Thinking_J
User Rank
Rookie
Re: Ariel Electronics Circuit Writer
Thinking_J   6/19/2014 5:28:52 PM
NO RATINGS
Many manufacturers of membrane switches use conductive polymers for their circuit traces.. and Ariel's solution would be great for their prototypes. But most of the rest of the industry can't tolerate the higher resistivity of this material for traces.

Low current, low speed, low density.. generally will work fine using conductive polymers.

But even my simpliest designs today often involve USB 2.0 lines (with controlled impedances) or modest levels of current (0.5 amp). All point back to copper.

 

 

Aries1470
User Rank
Rookie
Re: Interesting
Aries1470   6/18/2014 10:38:08 PM
NO RATINGS
I hope the image works.

It is mostly for short runs, so the barrel is not needed, you can feed it with short smd tapes. Spare sprockets can be located on the left for the different sizes that exist. The dot matrix printer had a single bar, with one fixed, or none, dependant on the brand and model, and the other end was movable. It is a low production run anyway, not for large batches. There is also a parts tray on the left, for manual usage to or for odd shape etc.

The drawing is just a raugh guide, to convey the idea, and was made using 'paint' ;-)

simple idea description drawing


Now on the right sise, can be the exit, that can feed a small 'oven'. So it can have a conveyer system to transport it along or it can be on the left, or both in an ideal world, so it can suit more people depending on their desk layout. So the oven have have 2 side doors, and a front facing window door, for manual loading.

Trays / bins can also be located on the other side, and have an option for a "small" monitor or just the output from the camera, and have a 'joystik' to be able to use for manual pick and placing.

I hope that this is the type of input you are looking for.

 Edit:

image may need to be opened in a new tab, due to the width constraints of this blog ;-)

Aeroengineer
User Rank
Blogger
Re: Interesting
Aeroengineer   6/18/2014 11:08:58 PM
NO RATINGS
The pic made it though, but I am looking at it on a small tablet, and it is clipping the picture.  I will take a closer look tomorrow when I am at a better computer (on travel right now).

Sanjib.A
User Rank
CEO
Desktop PCB assembly
Sanjib.A   6/18/2014 11:05:59 PM
NO RATINGS
It is a great idea...wish you All the Best! The footprint of the machine 11" x 17" or smaller would be great, and as you have mentioned there should be an option to add the reflow oven as a module later, hence the provision shall be kept for that...if you could convert the pick-and-place chamber to an oven, that would be great...but I feel that might be too ambitious to start with.

Aeroengineer
User Rank
Blogger
Re: Desktop PCB assembly
Aeroengineer   6/18/2014 11:10:33 PM
NO RATINGS
I too like you would like to have just what you mentioned.  I am scheming on how it might be done in a clean way, and in such a way that it will not require too many trade offs to accomplish.

seaEE
User Rank
CEO
Desktop pick and place
seaEE   6/18/2014 11:33:37 PM
NO RATINGS
This sounds interesting.  I wonder what the smallest component it would be able to place might be?

Aeroengineer
User Rank
Blogger
Re: Desktop pick and place
Aeroengineer   6/18/2014 11:42:07 PM
NO RATINGS
I would like to target 0402 parts as that is what I am typically using in my projects and doing those by hand

elizabethsimon
User Rank
CEO
Re: Desktop pick and place
elizabethsimon   6/19/2014 11:59:49 AM
NO RATINGS
would be really nice if you could do DFN and QFN packages since theres a lot of newer parts that only come in those pacakges. BGA would be nice too but probably not practical.

Another consideration would be how to hold the PCB. Especially a small one.

I'm sure that you've given that some thought.

 

Aeroengineer
User Rank
Blogger
Re: Desktop pick and place
Aeroengineer   6/19/2014 12:08:02 PM
NO RATINGS
I agree that DFN/QFN/QFP parts are a must.  I think that BGA's could be accomodated, though I think that the biggest issue will be the overall size and weight of the package.  I have given some initial thought into how to handle large (>100 pin QFP sized parts) packaged, though it is far from fully baked.

 

In the area of holding small PCB's, this is still very notional.  I need to do some research on this.  If you happen to have any links that might give some great solutions to small PCB holding, let me know and I will use them as research.

Garcia-Lasheras
User Rank
Blogger
Re: Desktop pick and place
Garcia-Lasheras   6/19/2014 12:31:17 PM
NO RATINGS
Adam, I'm glad to see your desktop pick'n place machine is on a roll !!

About holding little boards, I've always seen these produced by building a grid panel with the appropriated total size.  I'm eager to see how do you plan to solve issue ;-)

Aeroengineer
User Rank
Blogger
Re: Desktop pick and place
Aeroengineer   6/19/2014 1:27:46 PM
NO RATINGS
Yes, Javier, it is started.  The idea is taking shape, and it should be a fun ride.  I am sure that we will tap you for some help on it ;)

Mark.Kirkwold
User Rank
Rookie
Desktop PnP--some inspiration
Mark.Kirkwold   6/19/2014 11:31:43 AM
NO RATINGS
In my opinion, making and using a stencil is too easy to try to automate the solder-paste dispensing in such a small machine. See, for example, http://imajeenyus.com/electronics/20100109_solder_stencil/index.shtml

While you are at it, you should also have a look at

http://imajeenyus.com/electronics/20130504_pick_and_place/index.shtml

Aeroengineer
User Rank
Blogger
Re: Desktop PnP--some inspiration
Aeroengineer   6/19/2014 11:58:25 AM
NO RATINGS
Thanks, this is some great info. 

 

In some cases, stencils make a lot of sense, in other places, they do not.  I for example will turn a handful of designs in a short time period, and the board itself only costs me $1-5 for three copies.  The stencil will cost more than the board.  In this case, I can save a significant amount by having this integrated into a machine that can print the paste.

 

On the other hand, if I were trying to do a few hundred boards, then the case you presented might argue for going the stencil route as you can do this faster, though it still will require more workspace.

 

Perhaps we can do a trade and then we will post the results.

Etmax
User Rank
Rookie
Re: Desktop PnP--some inspiration
Etmax   6/20/2014 1:46:46 PM
NO RATINGS
I was talking to a company that made a small desktop machine and they said that doing 0603's was about the limit for paste dispensed from a pneumatic driven syringe, and that finer pitch IC's than 0.8mm were also a problem. I for one would be happy to place all the fine pitch IC's and QFN's by hand and then us the machine to place all of the jellybeans. Currently I do the lot with tweezers for prototypes. DFN's are a breeze to do by hand as are TQFP's, the only things I have limited success with is BGA's. If I used a paste mask they would probably work out, the solder on the balls seems to be unleaded which has such lousy flow characteristics.

Aeroengineer
User Rank
Blogger
Re: Desktop PnP--some inspiration
Aeroengineer   6/20/2014 8:14:09 PM
NO RATINGS
A lot of what you are talking about gets to the heart of the mechanical design which governs accuracy and precission.  Calibration also plays a part.  If the machine has a constant offset, this will continue to propagate into the component placement.  I feel that many of these things can be overcome with good electromechanical design.

Aeroengineer
User Rank
Blogger
Re: Desktop PnP--some inspiration
Aeroengineer   6/21/2014 1:54:37 PM
NO RATINGS
I have a bit more time now that I can elaborate on my previous comment.  One of the reasons that I think that they were having some issues for paste dispensing is that they were using a pneumatic method.  While this is common, there are issues of compressability and the fact that you need to control the built up pressure in the system in a way similar to how you would control an inertial device.  Added to this, there is a lot of flex in the tubing and such that connects the pump to the syringe.  This is greatly influenced by temperature.  This means that efforts to calibrate it are going to be difficult.  I can also imagine, but I have not actually looked at in depth, that there are differences in viscosity in the solder pastes and that these too can change with respect to temperature. 

 

To combate these, I am thinking of one of two things.  The first, if going with the airpump method, there is a need to reduce the amount of line between it and the syringe.  I would look at doing that by mounting the pump right to the head of the divice.  There are a lot of small pumps that would be suitable for this.  The other option is to go to a mechanical system.  The mechanical system is going to be the most precise as it does not suffer from the "inertial" effects of the pressure system, nor the variability.

salbayeng
User Rank
Rookie
Re: Desktop PnP--some inspiration
salbayeng   6/21/2014 10:57:38 PM
NO RATINGS
Re Pneumatic pasting;

All of the concerns about length and size of plumbing are irrelevant (provided they are not excessively long or large) the pressure builds up in milliseconds, and more crucially the plumbing is constant during the production run, so it merely adds a calibration constant if anything. 

It's all done with TIME so 200ms = 0805 pad, 400ms=1810 pad. you manually dial up pressure to suit the nozzle / viscosity. Some machines do a suckback after a dispense pulse to avoid after-dribble. 

With bigger pads you might consider going diagonally across the pad (as you would when hand squeezing. 

For hand squeezing paste I use a 20g or 22g tapered plastic nozzle on a 2ml syringe. I have tweaked the 22g nozzle by slicing the end at ~ 30deg, this makes it easier to lay down a sausage.

A colleague has a hand placement machine with pneumatics, that uses a 5ml syringe, and typically a 18g or 20g straight nozzles. It has an adjustable timer, every pedal press gets you a dot of paste hold pedal down & it goes dot..dot..dot.. with a bit of practice you can do ~ 100pads/minute, your machine would about the same.

Note if you are an infrequent hand user, then the tapered nozzles are a better choice (harder to clog, easier to clean, less pressure) 

If you want to do fine pitch or 0402 pads, you will need a small nozzle (22g?) and you have a lot of clogging issues if not used everyday, and you will need ~ 100psi so you need a real compressor (You can get some really nice, quiet shoebox size units designed for airbrush use) Note if using pneumatics you can use any diameter syringe, (as the pressure in nozzle=air pressure)

For a hobbyist type application, I'd lean toward re-using the same 2ml syringes you use for hand squeezing in the machine, this keeps the paste fresher, and then use a leadscrew with a tiny DC motor to push on the plunger, and just use a timer to control dispense volume. The leadscrew/motor will be heavier and larger so will slow down the motion system. It's kind of a tradeoff, (more versatility + slower) vs (more dots per second + hrs wasted unclogging nozzles). 

It's probably preferred to <design> for a motor/leadscrew, as this can be swapped to pneumatic in a matter of seconds, much harder to retrofit the motor assembly. 

 

Aeroengineer
User Rank
Blogger
Re: Desktop PnP--some inspiration
Aeroengineer   6/21/2014 11:08:36 PM
NO RATINGS
As I have been looking into the pneumatic systems, most have a popoff valve to help with the "inertial" effect.  I actually would prefer to use a squiggle motor for this type of system.  I need to see if I can convince them to help us out with a motor to use as a prototype.  This is great as it has both the motor and the leadscrew already integrated.  They are piezo electric systems.  Pretty neat little devices that have very high force in a well concentrated device.  By actually making it a closed loop system based upon displacement, then I think that very fine control can be had in a dispensing effort.  Thankfully this will come as a later addition to the machine, but the basic framework needs to be laid out from the beginning so as to be able to integrate it later. 

 

As a side note, I was working out some basic calcs for speed of the machine if we were to want to have it place 60 components per second.  If the average travel distance were 7.5" per component, then that would mean that the average head speed would be about 50mph.  This seems like it might be managable, but will have to look into the wear characteristics of the components.  By going with a smaller build area, this would allow for leadscrews to be used and compete in cost with other components for a belt system.  Leadscrews will allow for better positioning of components, but one of the maintenance items will have to be the lead nuts.  Anti backlash nuts could be used, but are pricy.  Might be able to use two nuts and make our own anti backlash nuts for cheaper.  Will have to look into it as things get closer. 

salbayeng
User Rank
Rookie
Re: Desktop PnP--some inspiration
salbayeng   6/22/2014 4:41:45 AM
NO RATINGS
Mechanics:

The piezo motor sounds like an "inchworm" drive, are these available at sensible prices nowadays? (they used to be ~ $1000 each)

Ok 60 parts per MINUTE might be manageable for a low budget machine.

That's 7.5" there and back per second = 0.42mph , it's much better to do all the mechanics calculations in metric units (=200mm/sec avg), you will eventually come unstuck persisting with furlong-fortnights , to the order of 32 or worse.

Lets consider  1 cycle/sec, if we limit the ramp to 1g , then we have 250ms to accelerate, 250ms to decelerate,250ms to accelerate, 250ms to decelerate,  so the maximum distance traversed for the ramp of  250ms,1g = 300mm or 600mm = 2ft point to point. So you would be ramping ~ 10% of the total trip.

It takes time to pick up the part and place it. So a  full cycle might be 100ms build up suction. 80ms Zacc , 80ms Zdec,100ms X acc, 100ms X dec,  80ms Zacc , 80ms Zdec , 100ms to release vacuum,  then 80+80+100+100 +80+80 to get to next part.  = 1.24secs / cycle. These numbers give 48 cycles/sec. for 4" point to point move, the actual distance doesn't change the calcs much if you have a servo drive, (8" point to point is 1.44secs, 16" is 1.64secs)  

Acceleration is important too, and given you may not have the large variety of nozzles, and suction variability as commercial machines, you will need to limit the acceleration, particularly with larger parts. Stepper motors can be jerkier than DC servos too.

Stepper motor will need a coarse leadscrew to get any reasonable speed.

Rotation speed will be limited to ~ 1000rpm for metal nuts, to prevent galling.

A double start 5mm pitch leadscrew with acetal nuts would be something I might consider.

Backlash can be fixed by using two nuts with a spring between them (and anti-backlash positioning algorithm)

Torque vs speed is a tradeoff with motor type and leadscrew vs pulley. So to get 8inch per sec , you need 2440rpm with a 5mm leadscrew or 338rpm with a 1/2" diam pulley. 2440rpm with a stepper motor generally requires a very high voltage drive (e.g. a typical "5v 6A" stepper motor might need 100v to get 1200rpm). Unlike a servo motor, the available torque at full speed is severely reduced in a stepper motor, (typically you need to use half the rated speed and half the rated torque in your calcs) 

Be careful in your calculations from linear values into rotational (motor)  values , there is almost always a 2 x pi in there somewhere. 

Aeroengineer
User Rank
Blogger
Re: Desktop PnP--some inspiration
Aeroengineer   6/22/2014 1:55:53 PM
NO RATINGS
You are correct that my calculation was rather simplistic.  The calculation that I did was more than anything to bound the problem on the high end.  Removing all other aspects including acceleration (I recognize that this is a futile assumption), what would be the bound on the high end.  60 components per second with an average of 8 inches of travel would require a .25 inch lead, lead screw and a 9600kV motor at 12V.  All of these things are in the relm of possiblity.  Are they practical, no, and are there other factors that I need to look into, yes, just have not had time to do that yet.

 

I am going to take the information that you presented and put it into a spreadsheet so that we can play with accel rates and other fixed time variables to generate a more realistic bound for the speed of the machine. 

 

I also think that a 1g acceleration is a little low, without having given it any great thought, I do not see right away why up to a 10g acceleration cannot be achieved.  Some of the details that would have to be looked into would be power/torque capacity of the motor, structural concerns (though I do not think that this will be a major concern), vacuum capacity to hold the part (friction), and positional accuracy

salbayeng
User Rank
Rookie
Re: Desktop PnP--some inspiration
salbayeng   6/22/2014 11:34:59 PM
NO RATINGS
Ok,

It's the second time you have posted "60 components per second "  instead of 60 per minute. i.e. one part placed per second. 

You need to be careful of getting scale factors mixed up in calculations, especially in spreadsheets.

60       Converting from RPM to revs per sec , also needed with fpm and ipm linear measurements.

2 x pi   Converting from "natural units" i.e. R L C & angular phase/velocity/acceleration to "human units"  specially note that torque to linear calculations almost always have a 2pi somewhere. The constant may appear as 0.159  (e.g. a motor with a 1mH + 1Ω winding will rolloff at 160Hz)

57    degrees per radian , need this for cos(angle)

32 or  9.8   Converting from gravitational to inertial units, need this to get from oz-ins to real torque. Imperial units have g embedded in odd places. Be careful when expressing accelerations in g , when you really meant to express a ratio. e.g. a block slides down a ramp at 30° implies a friction force of 0.5g , that calculation works on earth but not on the moon.

I do layout my boards on an inch pitch, and use materials in inch thicknesses and foot lengths, BUT when it comes to any sort of mechanical calculations, I always use MKS units. So I would strongly advise you to use MKS units as the main column in your worksheet , and have columns to the side of that for other units. I shade human input cells in yellow and computed in blue ,  so e.g. distance(m) in blue = 25.4/1000* (inches in yellow). If your primary calculation column uses metres, Newtons, Kilograms, and secs and derivatives thereof then you automatically get the right answers for velocity, acceleration, watts and joules and volts and amps and heat flow and temperature rise as well. Note that the correct representation of the inertial mass of 1kg is 9.8N/m/s/s. 

You might want to double check your numbers, a 9600Kv motor on 12v spins at 115,200rpm !!! ,  (the linear speed inside the nut is 7200fpm, about 10 times the max rating of bronze of 750fpm)

 

Aeroengineer
User Rank
Blogger
Re: Desktop PnP--some inspiration
Aeroengineer   6/23/2014 12:26:40 AM
NO RATINGS
Two things, I think that in many ways you and I are sayign the same things, but going about it in different ways, and second I think that you think that I am much further along than I really am.  The calcs that I am doing are just some quick back of the envelope bounding calcs to get me in the order of magnitude that I shoud start my searching, and are in nowise representative of a final work.  As was ilustrated with the motor comment, you can have a motor with 9600kV, but runing on essentially 3s is not really practical, this is why I said it is in the relm of possibility but not practical.  Though if I had come up with I needed to have something with 100,000kV and 100V, then I know that I am not even in close to being in a realistic bounding box. 

With that, I really did mean to say 60 per second, not per minute, though not because I am stating that this is the goal of the project, but I am trying to bound the problem on the high side of things.  I also understand that there are other limiting factors.  One of those things is as you mention the lead nut, but more than that, it is going to be the critical speed of the shaft.  Without having looked at it, I can imagine that this will be the biggest limiting factor of all. I am going to guess (do not take as a statement of fact) that for a 3/8" leadscrew with two rigidly supported ends that the critical speed is going to be between 10k-15k rpm

In reality, I expect that the actual max munber that could be achieved would be on the order of 1-5 components per second once things such as friction, inertia, and actions which must happen in series.

Another thing to take into account is that I am writing most of this from a hotel room on an 8 inch tablet, and doing some quick checking for components here and there, though not really doing anything exhaustive, yet.  Once I get home and have access to a monitor that is better suited towards setting up a nice spreadsheet, I will do so. As for the units that I amworking in, I use IPS generally because that is what my work requires.  I even use slugs for mass.  I know what an awful system, but becaues of my location, I have more access to a larger variety of parts that are inches than metric.

This is not to say that I do not appreciate the effort that you are putting in.

 

 

 

Aeroengineer
User Rank
Blogger
Re: Desktop PnP--some inspiration
Aeroengineer   6/23/2014 12:55:47 AM
NO RATINGS
Just for reference, the all knowing wikipedia seems to indicate that older pnp chip shooting machines were doing around 15 components per second on a single head, and that current machines can do around 40, though it is ambiguous as to if that is with a single head, or with multiple heads. 

 

So 60 as a high bound was a good guess to get in the ballpark, and I think that the low end may be around 1 per second with a max reality no higher than 5 per second, and even that may be dicey once we get looking into it in more depth. 

salbayeng
User Rank
Rookie
Acceleration and nozzles
salbayeng   6/23/2014 12:02:00 AM
NO RATINGS
I tried a bit of googling, but couldn't find any actual PnP accelerations.

I did do some experiments using my hand pickup pencil (with an aquarium pump for suction)

Using a 15g luer lock nozzle (approx 1.4mm ID), I could pick and hold SMC,DPAK,tqfn32 packages vertically so 1g is definately ok , and noted
  • SMC,DPAK would fall off quite easily if gently shaken (maybe 2g)
  • the tqfn32 took a bit more to dislodge(maybe 5g)
  • and 0805 and SOD323, SOD123 nearly impossible to dislodge
  • can't pick up a D2PAK, 
  • can't pick up a 0603 (they get partly swallowed)

 

(one could repeat with an accelerometer for accurate results),

This implies that 10g should be possible with the correct nozzle. Problem with homebrew machine is that you may not have enough "correct" nozzles, so will need slower acceleration and slower speeds for some bigger parts.

A second aspect of G forces is the reaction force,  so if the gantry weighs 1/10 of the rest of the machine , then the rest of the machine will be subject to 1g forces when the gantry moves at 10g, this can shake parts out of the feeders and or move the machine around the table top.

Thirdly if the c.g. of the gantry is a long way above the rails, then there will be a strong pendulum effect trying to bend the rails and de-stabilise the servo loop. 

---------some reading--- 

Here's an interesting read on PnP dynamics (some tables about halfway through) he basically says that acceleration doesn't have much effect on throughput (as most moves are velocity limited)

http://red.cs.nott.ac.uk/~gxk//papers/mxaPhdthesis/Chapter%208.pdf

maybe something here

http://www.cree.com/xlamp_app_notes/pick_place/

more reading:

http://mdc-smt.co.jp/pdf/eng/NXHANDBOOK.pdf

a comparison:

http://www.researchgate.net/publication/222695003_A_survey_of_surface_mount_device_placement_machine_optimisation_Machine_classification/file/e0b495231cd4d09291.pdf

Aeroengineer
User Rank
Blogger
Re: Acceleration and nozzles
Aeroengineer   6/23/2014 12:43:10 AM
NO RATINGS
Ah, and now you get to the heart of an idea that I am contemplating, but not sure the best way to make it work, and that is using counter weights.  It would slow the actuall acceleration of the machine, but also make it so the thing is not jumping like popcorn in a hot frying pan.

 

I will take a look at the links that you posted.

 

If you want to get an idea of the sorts of mechanical designs that I am capable of, you can go to my twitter page and look at the background pic.  That wind tunnel model had a span of almost 40feet and had two installed jet engines.  I designed all the wing structure, which included a blowing slot and flaps, the engine intakes, systems placement/attachment, thrust reverser, and even designed a rotary actuator for this thing as we had a flap that hung out into the exhaust and was on a 12" moment arm.  Nothing like designing an actuator that had to work in the heat with 4000lbs of thrust upon it and had to hold position to within .05 degree.  The final design used a hypocycloidal gearbox to a low speed ac motor.  It was capable of 9000 ft lbs of holding torque, and had an inline set of strain gauges to measure the torque.  The part had to be designedso that it could handle the expansion due to heat, which was on the magnitude of .25" from nominal room temperature.

 

So have confidence that right now, I am doing very, very high level calcs to understand how big the design space is, and I hope that this week I will be able to sit down and really spend some time putting things together in a much more refined and detailed fashion.  These will take into account someof the realities that I have currently been ignoring up till now, though you are good to point these facts out ;)

salbayeng
User Rank
Rookie
Re: Acceleration and nozzles
salbayeng   6/23/2014 3:26:34 AM
NO RATINGS
Ok the counterweights are a clever touch,  and should work (marginal with stepper motors maybe), if the trajectories are mostly constant velocity, then lower accelerations have lesser effect on transit time. Also if the motor is selected for dynamic stiffness, it should have excess torque available, and if the acceleration is limited by stability issues, then adding a counterweight with damping (i.e.attached with sorbothane) will allow more acceleration, and less "ringing".

Have you considered using a stainless wire and drum drive? (instead of belts/leadscrews), it's a bit old fashioned, but you can move the Y motor off the gantry, and attach a counterweight on a dead leg. 

Wow impressive history of achievements, I was once lucky enough to visit Moog in Aurora, you can't really appreciate the power of the shuttle engine controls until you stand next to an actuator- Awesome.  I once built an aircraft simulator (payload) shaker for Lockheed Martin, impressive and scary at the same time!

Have you had a look at radio telescope drive systems?You would appreciate them, Monstrously slow and heavy, but insanely accurate, they use two huge motors and high reduction boxes, with torque control operating in a tug-of-war mode to reduce backlash to zero, some still have selsyn transducers.

Tough working on a laptop in a hotel room,  I try to carry a "lab in a suitcase", but I sure do miss the dual 24" monitors when I'm on the road!

 

 

 

Aeroengineer
User Rank
Blogger
Re: Acceleration and nozzles
Aeroengineer   6/23/2014 10:44:34 AM
NO RATINGS
I see that you found more than my twitter page ;)  Yeah, I am an engineer because if I were anything else, I am not sure what I would do.  I love it. 

 

My back of the envelop crazy calcs are just that, crazy, trying to help me understand the absolute limits.  I did a quick calc with a whole sequence last night, including a trap accel.  The accel definitely will play into the time on this machine as the travel distance is so short.

 

The other thing that I need to look into is the PV capacity of the nuts.  Usually they are not just rated for a linear speed, but a value that is proportional to the load times the speed.  This will bring some more reality into this effort.

Aeroengineer
User Rank
Blogger
Re: Acceleration and nozzles
Aeroengineer   6/24/2014 12:15:14 AM
NO RATINGS
I had never heard of that type of transducer.  It seems to be a variant of a resolver.  Learn something new each day.

 

As to the wire drive vs belt drive.  The nice thing about the wire drive is that the design space is a lot smaller than for a belt.  There is the issue of ensuring sufficient tension is maintained to control for slippage, but I like it because it is a very space efficient design.  There is another aspect in that it could be set up to carry current.  This is something that would have to be looked at very closely, but could offer some interesting possibilities in that it would be easy to shield this so that you would not have to worry about it being open (unlike carrying power through the slides (which I think is an interesting idea, but hard to shield).

 

The advantage to using a belt is that you can get grooved belts, but they are more prone to stretching and hence might lack in precission.  The would also not have the ability to carry power if necessary. 

salbayeng
User Rank
Rookie
Counterweights ??
salbayeng   6/23/2014 5:05:03 AM
NO RATINGS
some crazy ideas with counterweights:

This only works on the Y axis: have a two head machine, going in opposite directions on either side of the gantry (could be a paste head one side, placing on the other) or camera on one side?, This probably needs a long "fast" counterbalanced Y axis and a slower X axis, so you would put components on a pickup grid above and below the PCB. Maybe 4"PCB + 4"grid top + 4"grid bottom = 12" Y travel, and 6" X travel = PCB width. 

Or you could go all out, and put the X motor underneath the workpiece (PCB+platen+parts+motor), have the heavy workpiece move sideways 1", while the light gantry moves 5" other way for 6" total . 

Also for the  Z axis , have you considered using bail bars (like the old fashioned pen lift on chart recorders) the "high position" has no accuracy requirement, while the "low position" will have a spring anyway, this would be adequate for placing, but might need better Z control for pasting.

You might consider making "double wishbone" suspensions for the placing nozzles from laser cut shim, this gives parallel motion and a soft spring (wouldn't be that hard to make it 4 nozzles spaced 1/2" apart  either) Need some thought how the component aligning fingers work with multi-head. 

Aeroengineer
User Rank
Blogger
Re: Counterweights ??
Aeroengineer   6/23/2014 10:48:38 AM
NO RATINGS
I like the thoughts that you have going on this.  This is one of the reasons that I have not rules out stepper motors and a belt system either as with the belt system, it would be very easy to do the counter balance approach by mounting the weight right onto the belt.  I have not looked into the steel wire method, though doable, usually for higher load parts.

 

The dual heads moving in different directions could be an interesting idea, and could make it very useful in that you would not have to dock heads.  Have to give this one some thought on a general methodology to integrate it.

 

Duane Benson
User Rank
Blogger
Re: Acceleration and nozzles
Duane Benson   6/26/2014 12:01:13 PM
NO RATINGS
re: "I tried a bit of googling, but couldn't find any actual PnP accelerations."

I've been told that some of the newest machines run so fast that the heads are only barely subsonic. The drive electronics has to accommodate elasticity of the mechanism and place the component right as the component swings back over the proper spot on the board.

Aeroengineer
User Rank
Blogger
Re: Desktop PnP--some inspiration
Aeroengineer   7/3/2014 6:25:12 PM
NO RATINGS
Hey, I just came across this link for a mechanical solder paste dipsenser that might be applicable to somthing that we might want to do, and while it is not very detailed, it is an interesting concept.  I think that it could be done in a much smaller package, but it is some food for thought.

 

http://letsmakerobots.com/content/solder-paste-dispenser

Aeroengineer
User Rank
Blogger
Build Area Size?
Aeroengineer   6/21/2014 1:47:59 PM
NO RATINGS
I am wondering, I have not heard too many comments about the build area size that people would be interested in for this type of machine.  I would love to hear your thoughts and why you think that it should be that size.

salbayeng
User Rank
Rookie
Re: Build Area Size?
salbayeng   6/21/2014 10:08:11 PM
NO RATINGS
G'day, 

There seems to be a popular panel size in use by PCB manufacturers of about 10.5 x 16" ,  So anything bigger would be a "special".

All of my bigger PCB's would fit 2, 3, 4, 5, or 6 onto a panel.

Note if you are using a stencil, you try to make a "set" of PCBs approximately the size of a sheet of paper i.e. 12" x 10"  to make the stencil economical.

But for the machine you are thinking of with paste dispensing, you can do a single PCB at a time (this possibly reduces the niumber of feeders in use too) 

A quick measure of what "big" PCB's I have lying around the lab I have 5x6  2.5x4 3x6 5x7 3x4(many) 1.5x5  all of these are revenue PCB's , So I could load 100% of my PCB's over the last 3 decades with a 5 x 7 work area.  It should also be borne in mind that any PCB longer than ~8" is likely to be flexed sufficiently to crack MLCC capacitors. This puts an upper reliability limit on PCB size. 

The exception to the size rule would be LED illumination strips (so maybe have a provision where you can have the pcb hanging outside the work area by removing some feeders)

So I would recommend an absolute minimum of a 6 x 8  or a more reasonable minimum of 8 x 10.

What will happen in practice is the dimensions will be limited by the feeders, (not the PCB size) (e.g a 8mm feeder might be 1/2" wide, so 20 of them for 10")

From a "number of feeders" perspective, it makes more sense to be able to pickup from the work area , i.e. use double sided tape to stick down 8" lengths of tape with bigger IC's , just next to the PCB. This requires some software support as you can't gaurantee you have stuck the tape down straight. Real PnP nachines do this anyway so they can pick from waffle packs. On the basis of using half of the work area for grid based pickup, I'd make the work area something like 12x10 or 16x10. 

You can refill the "grid" really quickly using a hand pickup pencil (made from a 1ml syringe) 

Just my ideas..

Aeroengineer
User Rank
Blogger
Re: Build Area Size?
Aeroengineer   6/21/2014 10:19:46 PM
NO RATINGS
This is good information.  Before I bias the result, I will let a few more respond to the size question, but you have given a pretty good argument for the sizes you recommend.

 

As to the issue of feeders, I am heavily biased against them for this type of design, and prefer something perhaps a bit of an upgrade to what you were saying with having an area off to the side.  I am thinking at a minimum a moving plate that you can place rows of cut tape on (indexed to an edge) and then the plate would move to the next row of components as it switched from one component type to another.  I also have a concept that is a bit of an upgrade from that, which would allow for more components to be stored (twice as many) and would not increase the footprint of the overall machine.  It would almost be a conveyor style that would bring a row of components into place right under the picking head, and then it would index to the next row once complete with that component type. 

salbayeng
User Rank
Rookie
Re: Build Area Size?
salbayeng   6/22/2014 12:46:41 AM
NO RATINGS
Ok.

Let's make the distinction between "grid pickup" and "feeders"

(a) Feeder: always picks up at the same X, Y coordinate for any component

(b) Grid pickup: picks up at the "next" coordinate 

So you are considering grid pickup (this entails some extra software, as it has to maintain a database of "next" for every strip of components, and these may be on 2 or more plates)

The second piece of the puzzle is the axes arrangement, there are basically two approaches 

(a) keep the work area stationary and move the head around in X,Y,Z

(b) move the work area in X , and the head in Y (much like those large plan plotters where the paper goes in and out and the pen goes side to side.

There are some significant advantages to moving the carriage in only one direction, as it can get pretty bulky by the time you add a camera, pickup nozzle, alignment plates, glue dispenser. You can also combine the two, so the carriage can move say 12" in Y and only 4" in X, but the workpiece (PCB + parts) is on a X conveyer so could be quite long.

You can make good use of the Z dimension too (which I think it what you are suggesting, when I hand place with tweezers, I use a cantilevered parts tray that sits about 1/4" above the pcb, less likely to bump already placed parts.)

If you are moving the workpiece (the grid and/or PCB) you need to be mindful of the X accelerations ,  the smaller parts are OK, but electrolytics, inductors and SM connectors have low paste area to mass+height ratio, so may need to limit acceleration to 0.2g at the end when the big bits get placed.

When you are thinking about plates with grids of parts , (I will call this a "grid plate" from now on, ) ensure the grid plates have sufficient weight to limit the acceleration induced during normal handling, e.g. the plates should be 1/8" minimum thickness aluminium. Or if made from PCB stock, glue a 1/16" sheet of steel on the bottom. 

Even if using "grid pickup" you should add a vibro feeder, these are simple and will take all the SO-xx & Dpak parts in tubes.

Aeroengineer
User Rank
Blogger
Re: Build Area Size?
Aeroengineer   6/22/2014 1:00:35 AM
NO RATINGS
Yes agreed that there are some tradeoffs as you suggested between a gantry system and a moving bed/workpiece system.  I currently prefer the gantry system for mounting the head to for both solder paste and the vacuum head.  I am not stuck on it, but for the reasons that you mentioned, I think that trying to have a moving bed may end up being more complex and producing an inferior result, including solder bridges on find pitched parts. 

 

Now for the area which contains the unplaced parts.  Here I think that a moving bed concept of some sort is the way to go.  The reason for this is that it will both simplify the operation, while reducing the foot print of the device.  As you mentioned, there will have to be some general indexing to know where to go and hunt for the next component, but then again a regular feeder needs to know how far to index the tape as well, so I do not really view this as an added complication.

 

The other thing (I am not sure that I mentioned it yet) is that I am in contact with one of the guys at TI in their C2000 group that does motor control.  I am looking at what the trade would be to do servo motors vs stepper motors.  The one thing that this would open up is that hobby brushless motors could be used in conjunction with a standard type of encoder.  This also would play to the fact that this type of machine does not need hi holding force, more that it needs to traverse quickly from one point to another.  This is now starting to play into the strengths of a servo motor over a stepper motor.  This is not even close to being decided, but I do want to do some investigations into it as this might be a way to get some higher speed out of the device that might not normally be there for this type of a machine.

salbayeng
User Rank
Rookie
Re: Build Area Size?
salbayeng   6/22/2014 5:30:18 AM
NO RATINGS
The postings have crossed in the ether!

I've half addressed the motor tradefoffs just now.

I've also looked at using the hobby BLDC motors for machine type applications, I've collected a few motors and a few ESC's  but got too busy to play with them.

Some notes re hobby BLDC motors:

(a) Most have the rotor on the OUTSIDE , known as an "outrunner", can be a bit awkward mechanically, so check which one you have.

(aa) Likewise most don't have two shafts so mounting an encoder on the back might be difficult.

(b) They spin very fast, denoted by the "k" number , so if k=3000rpm/volt , you get 30,000rpm for 10v. For our purposes a LOW k is desirable.  so Kv=500 is about 10,000 rpm at 20v , close to our requirements.  Note that the torque constant is the inverse of Kv, so halving the Kv , doubles the available torque per amp,  

(c) Some motors (designed for helicopters) have an integral gearboxes , so combined with a finer leadscrew, you could drive the motor in vector mode and get stepper type encoder-less positioning

(d) The motors are designed for sensorless operation,  making it difficult to generate much torque at near standstill.

(e) almost all BLDC's are 3 phase, so if you want to retrofit an encoder for commutation or vector control, it needs to have <some multiple of 3> lines per revolution (most are multiples of 4 or 25) 

 

Accuracy / resolution is another issue, consider you probably need 0.05mm to place a 0.65 pitch part, so with a 5mm pitch leadscrew you need to locate within 1/100 of a rotation, requiring perhaps a 100 line encoder = 400quad counts. 

A 200step stepper with a 5mm pitch leadscrew gets you 0.025 without microstepping. Same 200step motor with a 1/2" diam pulley gets you 0.18mm without microstepping, 4:1 microstepping is achievable so 0.05mm resolution is then possible.

Apart from the gear backlash , a 38mm standard "robotic" gearmotor, with 3:1 gearbox and 100 line encoder is in the right speed/resolution range with a 5mm leadscrew (but might fall apart from all the load cycling) 

 

 

Aeroengineer
User Rank
Blogger
Re: Build Area Size?
Aeroengineer   6/22/2014 2:11:37 PM
NO RATINGS
I do have some familiarity of the hobby bldc motors, though the information that you post is great background for those that do not have that background.

 

The thing that I love about the hobby brushless motors is that there is such a large varity at very cheap prices.  The other fact is that their power output per volume is much higher than the steppers.  Though their downfall, as you mentioned is that they would need to be controlled in a closed loop fashion.  The great thing about this, though is that you can reduce the size of the motor as you are not as concerned with slipping at high torque if you have the right motor control solution that can handle that. 

 

The standard hobby controllers would be out of the question, but I have been speaking with TI about their Instaspin Motion control solution.  I am not saying that this is the direction to go, but it does provide an itneresting option that may be more ideal for this type of setup.

 

As to the actual motors that I would be looking at, I am actually thinking that outrunners are going to be the desired option if this was the path to go down.  They can be found in small sizes at lower kV (in the 1200-2500 range) and would not be that difficult to adapt for this type of operation. 

 

On the topic of encoders, I would like to go with perhaps a hall based, non contact encoder.  the biggest issue is going to be update rate, but there may be some solutions out there that could deal with this.  It is something that I really need to iinvestigate much more.  Standard contact encoders are rather large, but as you said, there is not a large need for high line count.  I am sure that there is a technical solution, but is it cost effective?  That will be the question.  This is something that I would like to have play out in a post as it will provide some valuable learning.  I think that this application will be right on the cusp of both solutions.  I think that there will be value found in both and that both could be made to work.  It will then turn to what is the cost of each solution, and do they make technical sense from an engineering perspective, not just a that is a cool thing perspective. 

salbayeng
User Rank
Rookie
Motion solutions
salbayeng   6/22/2014 9:30:00 PM
NO RATINGS
Trying to "roll your own" insta-spin motion solution might be turning a molehill into a mountain. I'd suggest using one of the hobbyist CNC-router 3axis motion drive packages that already included the insta-spin.  

Also noticed this in the instaspin technote re sensorless control:

"Position applications require a mechanical sensor in order to precisely identify the motor angle at zero and very low speeds. The FAST Software Encoder may provide redundancy in position control applications; this can be used as a safety feature in case the mechanical encoder fails. "

I've used magnetic encoders before, Austria Microsystems were the leader in this field about 5 years ago. You need SPI type as I2C too slow for motion control, the AS5040 is typical , 10 bits "absolute" ,  you can cascade multiple devices , so 48clocks = all 3 axes in 6bytes, (some encoders also do multiturn absolute, can save you a couple of lines of code)  . Just glue a magnet on the rear shaft (nudge it gently while slowly rotating to centre it). Austria also have ring magnet solutions too. One issue with the outrunners is the magnetic flux of the magnets might cause positioning errors. 

You can write your own trapezoidal move algorithms with 1kHz update rate on a 20MHz CPU that will give adequate positioning control. To get 9PWM's you will find PIC's may be a better option than ATMEGA's, and of course the TMS320 range have many PWM's.

When it come to encoders, you need about 4 to 16 times more resolution than your required accuracy. This is necessary to be able to accurately measure the speed right at the end of the move.

 

Aeroengineer
User Rank
Blogger
Re: Motion solutions
Aeroengineer   6/22/2014 9:47:14 PM
NO RATINGS
With 10 bit total resolution (assuming 8 bit effective) on a .25" lead, this gives .001 placement tolerance measuring capability.  Though on a 24 pole 22 tooth motor, assuming only stopping on a pole, this would give about .01" placement/step.  This is not to say that it cannotbe improved by going to a lower lead, lead screw.

 

One thing that I need to look at with brushless servo motors, is how steps are performed, or if it can hold in intermediate placement between teeth.  These are some of the details that I would like to look into before writing it off.

 

As to the challenge of rolling our own, yes it may be more complex, but if one does not at least look at the idea, then you will not know if the pain is worth it, and then the product will be no better than any one else out there.

 

As a designer, I like to throw any marginally reasonable idea out there and then chase down why it will or will not work. 

 

salbayeng
User Rank
Rookie
Re: Motion solutions: motors- a mammoth post
salbayeng   6/23/2014 2:12:20 AM
NO RATINGS
re "One thing that I need to look at with brushless servo motors, is how steps are performed, or if it can hold in intermediate placement between teeth.  These are some of the details that I would like to look into before writing it off. "

All motors are, from an academic perspective, identical and can be represented in an equivalent d-q model (direct and quadrature). First consider a motor frame with two poles consisting of a north magnet and a south magnet, and the rotor is a steel bar with a coil of wire wrapped around it. Obviously when you put current in the coil , it will point either up or down (i.e. zero or 180degrees = direct) , and the available torque will be zero. When you rotate the coil to left or right (with current applied) the torque is maximum (i.e. 90 or 270degrees=quadrature) at this point the torque is proportional to current, the torque is sinusiodal with rotation.

If you rotate the rotor with power off it will detent twice per revolution. This effect is known as "saliency" or "cogging", it is deliberately enhanced with stepper motors due to the poles being tooth shaped, a good servo motor on the other hand has negligible cogging (using a skewed rotor or stator) , the average BLDC (not designed for servo work) will have some slight cogging. 

Each pole may have more than one slot to place wire in, so by rearranging the wire in the slots you can achieve a non-constant magnetic field across the pole face. This affects cogging, and also the back emf. Back-emf is caused by moving the coil through the magnetic field , so the waveform of the back emf  (imagine the motor is a generator) could be sinusoidal (synchronous AC, most servomotors) or trapezoidal (most BLDC, some servomotors, most brushed motors) . Like wise the torque vs angle curve can have differing wave shapes. A trapezoidal back-EMF suits simple motor drivers, where you only have to turn transistors on or off. Note that the difference between the back-EMF, and the voltage you apply to the motor creates torque, so if the waveform is wrong it creates torque ripple (A stepper motor has really high torque ripple)

The minimum requirement to cause a motor to turn is two coils at  right angles, so as to create a magnetic field alignment between 0 and 360degrees (it's just sine and cosine) you can do the same with 3 phases, 4 phases, 5 phases etc in the end all you have is a magnetic field strength and an angle.  So the term "FOC", field orientated control, merely means you are focussed on controlling the field direction (aka "vector") (and amplitude). 

Voltage vector control means you apply voltages to the windings, and hope the motor aligns with the rotor, a stepper motor is an obvious application of this, you apply 5v to any of 4 windings and it points there, the vector space for this motor only has 4 points per electrical rotation (but there are 200 electrical rotations per mechanical one) . More sophisticated VFD's change the voltage with frequency. The vector space for BLDC motors is 12 points per revolution. The vector space for sinusoidally wound motors is essentially infinite.

Most servo controllers use current mode control, they force a current in the winding to generate a restoring torque (to drive position error to zero), note that the phase angle of the current is 90deg for maximum torque (unlike zero degrees for  voltage control). Note that the power dissipated in the windings varies with load,  so you can "borrow" upto 10 times rated torque, and let the motor cool down later at constant speed (can't really do this with voltage control).

All of the above have effects on the positioning accuracy of a motion system, typically the positioning error is the sum of {saliency effects} and  {effective stiffness x (friction + static)force}.  
  • Stepper motor (with microstepping) high saliency means that applying 25% + 75% winding currents won't put you at 25% of the pole spacing , and the stiffness is highest on a pole and softer between.
  • BLDC motor (trapezoidal winding) with microstep, also won't be linear between poles , but better than a stepper
  • A servomotor with Voltage control , has no angle dependant accuracy, and the stiffness is constant (but softer than an equivalent stepper)
  • A servomotor (closed loop current control) has the highest stiffness and best accuracy.

So getting back to your original query, the "steps" are actually commands to the drive, and it works out what to do from there, for example the drive might be set up for 1024steps per mechanical revolution, so:
  • stepper motor this means each of the 200 poles is carved up to ~5 steps , but really only accurate to +/-2 steps. (slow , fairly accurate)
  • BLDC, 6 pole pairs , each of the 12staes withing each pole are carved up (interpolated) to ~14steps, maybe accurate to +/-5steps. (very fast, but sloppy)
  • Servo (closed loop) 4 pole pairs + 4096qc encoder , each of the 4 poles carved up to 256 points, accuracy +/- 0.5step (fast and accurate) 

So a BLDC does not have "teeth", but it does have "poles", while "steps" are commands to a drive. All (magnet) motor types will "hold" at intermediate positions, but the accuracy and stiffness while holding varies across types. Closed loop control dramatically improves holding accuracy and stiffness. Closed loop control can double available accelerations, but no effect on limiting speed.

 

 

Aeroengineer
User Rank
Blogger
Re: Motion solutions: motors- a mammoth post
Aeroengineer   6/23/2014 10:40:35 AM
NO RATINGS
Good info. I will read this in more detail when I get back to the hotel tonight.  I will see what I can digest and then ask a few more questions.

Aeroengineer
User Rank
Blogger
Re: Motion solutions: motors- a mammoth post
Aeroengineer   6/24/2014 12:25:48 AM
NO RATINGS
"Closed loop control dramatically improves holding accuracy and stiffness. Closed loop control can double available accelerations, but no effect on limiting speed."

 

Though the speed will be present by the very nature that you are using a bldc motor over a stepper.  Though the fact of adding closed loop to the bldc motor will not change its nature in speed, but only increase its accuracy and low speed qualities.

 

On the though of going closed loop and needing an encoder.  I just saw this product announcement on a competing website.

 

http://www.eeweb.com/news/high-accuracy-magnetic-position-sensors

 

 

salbayeng
User Rank
Rookie
Re: Motion solutions: motors- a mammoth post
salbayeng   6/24/2014 3:00:17 AM
NO RATINGS
Yep , You have the gist of the motor comparison. Each sort of motor has its weaknesses and strengths.

I see you have found the Austria Micro Systems = AMS website !  ,  these guys were relatively unknown about 6years ago, we had to go through all sorts of weird purchasing arrangemengts to get the stuff from Austria, now you can buy from your favorite distributer. 

Selsyn: just a kind of Synchro ,  check them on wiki , too hard to explain without pictures. They were antecedents of the resolver. Quite common on 1940's ships and aircraft for relaying the position of say the rudder to the bridge (some still in use on DC10 era aircraft for fuel gauges). You just wire two together, put AC on each end , when you turn one shaft , the other turns, it's bidirectional, great fun for kids of all ages. The other neat thing from back then was the amplidyne: before the days of high power servo amplifiers, you could, for example, attach a synchro to the hour hand of a wind up clock, and a 1000ton radio telescope would follow the sun across the sky, just like leading a bull with a ring through the nose. 

Belts: Do you mean toothed belts like a timing belt? or multigroove like an automotive fan belt. Belts have the desirable property of adding damping, most stepper motors without micro-step drives are unstable if a high inertia load is directly coupled to the shaft (That's why you see so many timing belts on stepper driven items). Timing belts are more accurate than you might think (if they are tight enough) and used on most PnP machines of 10yrs vintage.

Aeroengineer
User Rank
Blogger
Re: Motion solutions: motors- a mammoth post
Aeroengineer   6/24/2014 7:56:08 PM
NO RATINGS
I had actually known about the AMS site for some time.  They have a great app note about using RFID tags under water in salt water.  It was pretty impressive.  They also have an interesting line of noise canceling chips.

 

The thing that impressed me about this chip is that it seems to be a full quadrature encoder replacement that is programmable.  This is something that is pretty ideal.  The cost was not too awful, and the other thing about it was that it essentially applies a predictive value to compensate for lag while maintaining pretty high angularity accuracy and precission.  It seems like a chip that I need to look more into.  I like that it is a drop in replacement for an encoder and that it can be programmed with different scaling factors.  It can also output the valueas a digital word value, or as just a quaderature style output.

 

On belts, I am specifically referring to the toothed timing belt style.  These are not bad, and are used in a lot of systems.  It is not so much the belt that I am opposed to, just that they are sometimes notas clean to integrate.  They are effective solutions.

salbayeng
User Rank
Rookie
AS5xxx encoders
salbayeng   6/24/2014 8:40:05 PM
NO RATINGS
 

The AS5030 encoders were used to retrofit cheap satellite actuators for the heliostat field at CSIRO , each mirror has two actuators, as bought they have a 4pole-pair ring magnet and a reed switch, we take these out, glue in a small magnet, and mount an adapter PCB using the reed switch screwholes: 

http://csirosolarblog.com/category/solar-thermal/solar-towers/solar-field-1/

With an actuator BoM cost of $25ea, we could point the 200lb+ mirror assemblies to a few percent of the suns diameter even with strong winds.

I also worked up a design for angle sensing through a fibreglass bulkhead in a marine situation with the same family. There is some really neat stuff in this encoder family like UVW to simulate hall switches, as well as the quad outputs you mentioned. (And the zebra stripe magnets for linear position sensors).

 

Aeroengineer
User Rank
Blogger
Re: AS5xxx encoders
Aeroengineer   6/24/2014 8:50:35 PM
NO RATINGS
It sounds like you too have had a lot of fun in your day.  With your frequet references to astronomy, I am guessing that you work in a rather specific field.  Though it seems too that you may have some not just ground based hardware, but space based hardware to your name as well.

 

We are going to do a poll about the build area size.  I was going to pretty much use a cut down list from Seeed Studio as it seems to offer a rather extensive set of size options. 

 

http://www.seeedstudio.com/service/index.php?r=site/pcbService

 

Do you have any other size options that you do not see here that you would like to see on the list for inclusion?

salbayeng
User Rank
Rookie
PCB Size poll
salbayeng   6/24/2014 9:29:02 PM
NO RATINGS
OK, If I exclude two huge LED lighting PCB's.

And include all my boards over last 3 decades, and include 4 colleagues I have worked with. All of these individual PCB's will fit in the 15cm x 20cm size.

(From an economic perspective, being able to load 20cmx25cm sub panels is better for production runs, but this is not the question being asked). 

So that's 5 votes from Oz for 15cm x 20cm  (you might want to discount the votes a little, as we are unlikely to be buyers) But I guess it's statistics you are after . 

Note that 16cm is a common extrusion width for enclosures. And that by number alone, half of the PCB's would be under the 10cmx15cm. (also revenue earned from PCB's is related to area)

salbayeng
User Rank
Rookie
UNC hardware rocks!
salbayeng   6/24/2014 9:06:49 PM
NO RATINGS
Some may say I'm a big fan of metric, but you can't beat UNF & UNC for quality hardware. I've just spotted on the solar blog http://csirosolarblog.com/page/6/  , about 3/4 way down

26.Each heliostat and its components are held together with 55 bolts – for a total 24,805 bolts in Solar Field 2 alone.

Most of these we airfreighted in from McMaster-Carr, including fasteners you just can't buy over here: 10-32UNC Stainless Hitensile thread forming hex head screws to precisely align the "origami" frames. 3/8UNCx1/2 flangehead ZP bolts with matching serrated nuts. 3/4"UNF x2" ZP for the bearings. And AN526C1032R8 for attaching actuator brackets. 

Aeroengineer
User Rank
Blogger
Re: UNC hardware rocks!
Aeroengineer   6/24/2014 9:38:48 PM
NO RATINGS
I have come to peace with both systems of measurement.  Each has their advantages. As you mentioned, you can get a whole lot of common hardware in Inch.  Many scientific calculations end up being easier to do in metric except when it comes to derived units, then it can be just as hard because you never remember what the derived units actually are in their proper combination of units.

 

McMaster is one of the greatest stores ever.  Here in the states, they have great service and a lot of things at a reasonable price.  There are a handful of other sources I use, but they are a big one.

salbayeng
User Rank
Rookie
Re: Motion solutions: motors- a mammoth post
salbayeng   6/29/2014 5:26:34 AM
NO RATINGS
Had time to grab two servo motors from the shed and compare with the outrunner types.

I'm looking at nominally 200W size motors in a NEMA23 frame (60mm square) , and for the outrunner I grabbed a big 55mm outer diameter unit with a Kv of 700, prior experience these three are probably in the ballpark for your PnP project.

So I spun the motors with a battery drill and recorded the waveforms, and additionally as I had no data for the outrunner, I did a torque vs angle measurement (with kitchen scales & digital protractor) (I was going  to clamp it in the lathe and use a strain gage , but the tiny amount of torque would be lost in the noise with a 1kg gage).
Manufacturer Type Detail   Link
Yaskawa USAREM USAREM-01C   https://www.yaskawa.com/pycprd/lookup/getfile/workspace%26SpacesStore%26432b65d9-d7cd-4235-a438-7d0f805e9110%26TSE-S800-4.1E.pdf
Hathaway (Allied) QS02300 QS02300-A   http://www.alliedmotion.com/Data/Documents/Quantum_Selection_Guide_Hath.pdf
Import C5055 C5055C-KV700   ebay: "C5055 outrunner"

So after much measuring and calculation the following table results:
Type USAREM QS02300 C5055
Polepairs 4 3 7
Vnom (Volt) 200 24 10
Icont  (Amp) 1 9 50
Pnom (W) 200 216 500
no load RPM 3411 3137 3925
Hz @ 3000 RPM 200 150 350
RPM max 4500 18000   20000?
Rm  (ohms) 22.7 0.25 0.012
L/R  (ms) 1.6ms 0.94ms 0.65ms
Kt (Nm/A) 0.353 0.038 0.011
Kt (oz.in/A) 50.13 5.40 1.56
Kb (V/rad/s) 0.353 0.038 0.011
Kb (V/krpm) 29.41 4.00 1.43
Kv  (rpm/V) 34 250 700
Tcont (Nm) 0.353 0.342 0.55
Tcont (oz.in) 50.126 48.564 78.1
Tpeak (Nm) 0.955 3.9 0.825
Tpeak (oz.in) 135.61 553.8 117.15
Wt 1050g 600g 290g
Frame Size 60mm 60mm 55mm
Cogging (Nm) 0 0.003 0.028
Cogging (oz.in) 0 0.426 3.976

Dynamic stiffness measured at 50% Inom
Ks (Nm/deg) 0.0121 0.0088 0.0330
Ks (oz.in/deg) 1.72 1.25 4.69
Ploss @ 50% Inom 5.68 5.06 7.50

I'll discuss in next post

 

 

 

 

salbayeng
User Rank
Rookie
Re: Motion solutions: motors- a mammoth post
salbayeng   6/29/2014 6:28:39 AM
NO RATINGS
oops just lost a bunch of typing.

OK three types of motors used (found in my shed) , experience suggests they are of right size to do the PnP job. All ~ 60mm size (NEMA23), The servomotors are typicaly of what pops up on ebay regularly.
  • Yaskawa : old fashioined , nice sinusoidal windings, needed an inefficient driver as you need 200v DC to make a 140V sinewave, very low distortion on back emf, zero cogging. about 300% torque overdrive available
  • Allied: newer style BLDC type, a bit coggy, moderate flux distortion, designed for trapezoidal drive, has a lot of reserve overdrive capacity with 1000% torque and 600% on speed.
  • no-name outrunner , very little data on the web, these numbers mostly from my measurements, very non-sinusoidal back-emf (3-5%) and a lot of cogging, a lot of asymettry as the motor is spun over a full revolution (the magnets are glued on haphazardly, and the airgap is non-uniform)

The motors have different nominal voltages (200,24,10) and different nominal currents (1,9,50) although all deliver nominally 3000rpm and 0.4Nm torque, so the electrical differences are just that some have more turns of thinner wire (if they were transformers they would have different turns ratios). These differences are reflected in the Kt and Kv values (which are analogous to turns ratios).

An important consideration for a robot is the dynamic stiffness, this determines how soft the drive is at high frequencies  (i.e. above the servo bandwidth). This also determines how well the position is held, when holding at a position (most servo controllers will drop to half current, and hold the last flux vector (i.e. this is open loop). So I've shown this as  for example Ks=4.69oz.in/deg This actually looks better than the real servo motors (mostly because the outrunner has twice as many magnets), but it should be considered in terms of the cogging torque of +/- 4oz.in , so the actual position error of the outrunner motor, due to some external disturbance torque of say 4oz.in could be around 1 degree plus or minus another degree (due to cogging) so position error is somewhere between zero and 2 degrees. 

While holding @ 50% Inom , all motors will dissipate 1/4 of "nominal" I2R losses, around 5W for the traditional servo motors, and 7.5W for the outrunner, most outrunners spin rapidly, and suck a lot of air through the windings, but in a hold situation the outrunner may overheat. (Possibly could hold at 33%, but cogging torques get proportionally more significant)

From a dynamic perspective, the motors are similar, with L/R time constant of 1.6,0.94,0.65ms so you could perform servo updates at 1kHz,1kHz,2kHz and do the PWM at 10kHz,10kHz,20kHz. Because of the irregular flux patterns of the outrunner , you will need to back off the servo gain to ensure stability, but you will probably end up with a similar GBW product as the "real" servomotors.

In summary , it looks like the outrunner could produce similar positioning outcomes to servomotors, at  less than half weight and 1/10 of cost.

So you just need to plug in say 78oz.in as torque and 6000rpm , with some leadscrew pitch number , and the putative mass of the head to see what trajectories are possible with the outrunner.  (Note, as is typical with most servo drives, the RPM is generally limited by external mechanical effects and bearing life, not by electrical constraints) 

Aeroengineer
User Rank
Blogger
Re: Motion solutions: motors- a mammoth post
Aeroengineer   6/30/2014 4:46:48 PM
NO RATINGS
Sorry for the late reply.  I was able to travel back home, and I have been spending the weekend with the family.  This is really great information for the reason that it really opens up the design space of the motor selection.  I think that this combined with the hall effect, non-contact encoder from AMS that operates at high speed could make a great combination. 

 

I was able to do some basic researching over the weekend about using bldc in closed loop control for positioning.  It does seem like a promising thing to continue to look into as it could open up the speed profile for what we are trying to do. 

 

As you have mentioned before, though there are a few other reasons to continue to look at stepper motors.  Some of those include the large installed user base that already has experience with this type of technology.  At the same time, though, if you continue to do what other are doing, you will result in a machine that does just that, the same as everyone else is doing.  This could be a differentiating factor.

 

I have a few days off this week.  I am going to see if I can get a meeting with the guy at TI to talk through some of the things of their solution, and what it would take to implement in this type of design.  They do have a nice dev kit that I might see if they will allow us to use for a while as we really try and evaluate this option.  This would allow us to really dig a little more into this and see if it is an option.

TonyTib
User Rank
CEO
Re: Motion solutions: motors- a mammoth post
TonyTib   7/1/2014 1:25:34 PM
NO RATINGS
Just to give an idea of costs:

QB02300 / QS02300 motor is >$500 each, without encoder (I recently looked into them, because I was speccing out a nice NEMA23 servo with 80,000 count/rev encoder).

Typical modular or kit encoders run >$60, although CUI's AMT capactive encoders are ~$25 (and require a dual shaft motor).  BTW, most encoders have pretty low maximum speeds (e.g. 15000 RPM for AMT, 8000 for most Accucoders).

If you want cheap and good repeatability, you may want to think about software corrections (map the particular motor and encoder).  Hand-gluing a magnet on a carelessly built hobby motor looks like a recipe for cheap, but not precise and repeatable.


BTW, I have a lot of respect for our ME's, since they build fixtures that can measure a part with a repeatability of 2 microns -- and that's removing the part and putting it back on, again (if the part stays clamped, repeatability is more like 0.2 microns).

Aeroengineer
User Rank
Blogger
Re: Motion solutions: motors- a mammoth post
Aeroengineer   7/2/2014 8:51:39 PM
NO RATINGS
Tony,

Sorry for the late reply.  I have been traveling for the last month, and get to spend this week with my family.  I am catching up on my responses.

 

As to the issues that you highlight with the dual shaft motors, this assumes that you are going to mount the encoder on the shaft of the motor.  There are many other places that it can be mounted.  Some of these places are at the other end of a lead screw, on the opposite side of a pulley, or on another pulley that is not being driven by the motor.  Each of these locations have their advantages and disadvantages.

 

The issue that you bring up of speed, that is a potential issue, but most likely, we will not be running in those speeds as we will be limited by other things such as the PV ratio of a lead nut, critical shaft speed of a lead screw, belt stretch, etc.  It is something, though to take into account if we get there.

 

As to the magnets being placed in a hobby motor in incorrect locations, this could be an issue, and one that we will have to look into, but if we do a back of the envelope calc and assume that we have a 5° alignment error, and a pitch on a lead screw with a lead of .05", this would lead to a position error of ±.0007".  This would still be an acceptable placement error.  I will let Bob aka salbayeng comment on my math ;)

TonyTib
User Rank
CEO
Re: Motion solutions: motors- a mammoth post
TonyTib   7/3/2014 6:57:31 PM
NO RATINGS
There's no reason to apologize for taking your time; family and job always come first.  So some more notes:

If you're using an encoder for commutation and/or speed control, it's probably best to have it mounted on the motor (or use dual loop feedback).  True, I've used Newport linear stages ($$$$) with an internal belt where the sole encoder was mounted after the belt -- but those used brush motors.

We also had issues with getting a circular encoder wheel mounted exactly in the center for a protoype stage (one reason it remained prototype).  That project had much tighter specs (IIRC, we were looking for ~1,000,000 counts per 360 deg), but little details like centering do matter.

If you're really going for affordability, you pretty much have to go steppers and custom drivers, since even affordable commercial drivers (such as Gecko Drives) will eat a lot of your budget.

ST does have some advanced stepper drive chips such as the dSpin models; I'm not sure how they compare to Trinamic.  Allegro Microsystems and TI are other "usual suspects" for integrated stepper drives, but as far as I know, neither one currently offers anything comparable to the more advanced Trinamic parts.  However, TI is also considering doing instaSPIN for steppers, but I have no idea when or if it will appear.  BTW, Trinamic was at ESC--I mean EE!!!LIVE!!!!--2014.

One possible approach would be to use sensorless control (such as instaSPIN) to control hobby motor speed and commutation, and use a cheap linear encoder for position (such as a US Digital EM1 + encoder strip; cheap but low res (for a linear, but probably good enough for this project) and looks hard to mount).

For overall control, an old PC is one approach.  I'd recommend taking a look at the BeagleBone Black, which is $55 for Rev C, runs Linux (including MachineKit, which is based on LinuxCNC), runs OpenCV, has good connectivity (Ethernet, USB, etc), good real time control (PRU-ICSS, which people are using for stepper control), 3 encoder inputs, a bunch of PWMs, an active community, some 3-axis (or more) stepper capes already created, and some attempts at using it for Pick and Place.

Aeroengineer
User Rank
Blogger
Re: Motion solutions: motors- a mammoth post
Aeroengineer   7/3/2014 7:39:00 PM
NO RATINGS
Yes, the BeagleBone black is high on the list of parts that we anticipate using.  We were able to speak to Jason Kridner at EELive and he said that he would be willing to help get us pointed in the right direction.

 

As to the encoders, I really think that it will depend on the overall stiffness of the system.  If you are mounting to the opposite end of a 3/8" leadscrew that is only 10 inches long, then at the maximum capacity of a plastic leadnut (which we would not be running at) the deflection due to torque would be .8°.  This would be a 350 lb load.  If we were to say that the gantry head weighed .5 lbs and we were to be running at a max accel of 10g's, then the deflection would be .01°.  For an encoder to be able to register this, it would need to be over 30,000 counts per revolution.

 

As for going with the Geko drives, that is probably out of the range of what we are able to handle.  Most likely we will have to leverage some other solutions that are already out there.

 

We did just get some information from Trinamic on a new dev kit that they have.  This is something that we may want to look at.  I know that the TI solution is a full closed loop control for BLDC motors, essentially creating a servo motor.  I have a feeling that we are going to have a lot to learn in this area. 

 

 

salbayeng
User Rank
Rookie
Re: Motion solutions: motors- a mammoth post
salbayeng   7/8/2014 5:06:31 AM
NO RATINGS
Hi Adam, looked at the numbers in your last post, call this AC1,  put them in a spreadsheet , errors looked good, but very high motor speed and low torque utilisation. So toned it down a bit ,  AC2 used a coarse leadscrew of 0.5", and only 2g accel, and trapezoidal speed profile. So now have manageable motor speeds but larger errors, would need a second 1024count encoder somewhere?. (I've randomly called the error budget 0.1mm)

Overview:
All cases are 0.5lb payload, with 10" travel. Error budget = 0.1mm total
Case AC1:
10g triangular profile with 20tpi leadscrew = 0.050" lead (1.27mm)
Use an outrunner motor , say 5% error due to angle estimation, (=18% of budget)
An AS50xx mag encoder requires 4.9% of budget.
And 0.001" of backlash is 25% of budget.
Maximum speed=16fps (5m/s) and ~100msec travel time.
Big problem is motor RPM = 236,000rpm
And torque is a miniscule 4.5mNm (about 1% of motor's spec)
Basically we are doing 200mph in first gear.

Alternate case AC2:
has 10x longer lead (0.5") , and trapezoidal velocity profile with 2g max.
Has some error problems, e.g. 152% of budget if we want to hold using 12step mode.
Magnetic encoder 49% of budget.
Total travel time = 240ms
Max motor=7500rpm
Torque=.01Nm (2% of motor spec)
OK now we are doing 60MPH in 2nd gear

I guess what all this says is we need to match the motor and gearing to the payload and error requirements. Most commercial PnP machines have very coarse gearing (consider a typical 1"diameter toothed belt pulley has a 3.142" lead) 

 

Aeroengineer
User Rank
Blogger
Re: Motion solutions: motors- a mammoth post
Aeroengineer   7/12/2014 2:38:40 PM
NO RATINGS
As an announcement to all those that are reading, Bob (aka salbayeng) has been invited to participate in an official capacity with this project.  I think that all of us have been able to learn from his very detailed posts.

 

He sent me his detailed calcs in the spreadsheet that he pulled these numbers from, and it is top notch work.  We should look forward to seeing more from him as he has indicated that he is willing to do some guest posts for us.






Aeroengineer
User Rank
Blogger
Re: Motion solutions: motors- a mammoth post
Aeroengineer   7/12/2014 2:48:53 PM
NO RATINGS
Bob,

 

I see that you are mentioning the C5030 and C5055 but have you taken a look at the C5047 device?  The only reason that I mention it is that it has not only a serial digital out, but it also has a standard quaderature interface.  In other words, it acts as a drop in encoder replacement.  This could simplify any roll our own efforts as we may be able to completely leaverage a dev kit from one of these motor control solutions with little modifications to their code.  This could make the transition to a BLDC servo solution very palitable.

salbayeng
User Rank
Rookie
re: CS5047 encoder
salbayeng   7/12/2014 10:48:41 PM
NO RATINGS
Hmm I thought the CS5055  had quad  outputs, but it doesn't, I was in a bit of a hurry so randomly plucked any 14bit datasheet.

Ok I was just looking at fundamental resolution issues, so trying to estimate the "ENOB" for the whole class of CS50xx encoders, I concluded that all these seem to have the same magnetic front end, which is where almost all the errors lie. 

With a bit of luck one should be able to get ~1.4deg accuracy ~ 7bits,  If one mapped a full rotation as one poster suggested earlier, then 0.5deg accuracy after correction might be achievable so a tad more than 8bits. 

The 14 bit resolution is still highly desirable, as it improves the servo drive stability = more accuracy and faster, this is because the critical feature of the positioning algorithm is getting the speed accurate at low speeds as you close in on the final position. (In the old days we would use a tacho-generator, and this provides an infinite resolution velocity signal for the servo drive.) 

Given you have say 256 possible target positions per revolution (i.e. the top 8 bits of our 14bit angle value) , let's say we are targeting position 200. Every time the controller scans the position value (lets say every ms) it will read the following values for the top 8 bits. .... 190 194 197 199 200   , if we then look at the velocity (i.e. differences between values) it goes .... 4 3 2 1  (as 8 bits) or   320 256 192 128 64 (as 14bits)  , As you can see on the very last step there are only 64 counts, so if we see 63 or 65 , the servo drive can adjust the braking to accurately stop on location. (Contrast this with: If we only had 8 bits available to generate a velocity, the very last step might have values of 0 or 1 or 2, so the velocity term is very chunky, we would need to drastically reduce the velocity  gain to ensure stability, and the whole drive would be slower and less accurate)

It should be noted that sampling faster isn't that much more helpful, if done at 0.5mS, the delta counts read:160 144 128 112 96 80 64 48 32 16  so your velocity is more chunky (on the last step) than before even if you do get more measurements, it's one of those GBW problems  (i.e. gain x speed = constant = resolution)

 

Aeroengineer
User Rank
Blogger
re: CS5047 encoder
Aeroengineer   7/12/2014 10:58:07 PM
NO RATINGS
The CS5047 still has the 14 bit output, and 11 bit quaderature output, so I think that it does not discount it.

 

I do understand though as our dVel/dt will have a higher quantitization error with a smaller number of bits as you slow down.  I can imagine that can lead to some hunt and pecking if not properly tuned.  Though I can imagine that if the motor system is properly tuned, that once you get into the low end, that you may essentially be running almost open loop in an effort to avoid that.

 

I am working on a modular concept that will let us test different motor/lead combinations.

 

 

salbayeng
User Rank
Rookie
re: CS5047 encoder
salbayeng   7/13/2014 12:26:42 AM
NO RATINGS
@Adam. I'm kind of biased to reading the encoder directly from the AS50xx chip, although keeping the quad outputs is beneficial as it allows plug and play with other servo-motors / optical encoders and most motor-controllers.

I've used the LS7366 quadrature to SPI chip, and it's good as it has a 64bit counter , HP also make a 4 channel quad to spi? / parallel chip too.

Typically the controller algorithm would be attempting to force a certain velocity profile as the motor slews to a stop, so if you can force a (14bit) target velocity profile of 256 192 128 64 0 , with an accuracy of 16 counts at each step, this will result in a final worst case position error of 64counts (in 14bits) which gives you 8 bit positioning accuracy. Note that a velocity accuracy of 16 in 64 is a fairly modest requirement. Of course once the carriage has stopped, the velocity is zero, and can't be used for control (it would limit cycle if you tried).

As you have said, one would normally switch to open loop once in position, this means running the motor in "stepper motor" mode, (with a fixed current flowing in all three windings) , normally you would also switch to a lower motor current to let it cool down (1/2 current = 1/4 heat), the problem with a "coggy" motor is that it may move slightly when the current drops.

Another approach is to switch to a low gain position holding algorithm, this is severely limited in torque ability, so that it provides no more than say 75% of the stiction torque, so if the carriage gets bumped enough to unstick it, it will move in the direction that reduces the position error.

Aeroengineer
User Rank
Blogger
re: CS5047 encoder
Aeroengineer   7/13/2014 12:28:48 AM
NO RATINGS
There is no issue as this chip has both SPI and Quad out.

Aeroengineer
User Rank
Blogger
re: CS5047 encoder
Aeroengineer   7/13/2014 1:20:27 AM
NO RATINGS
I guess that there is another way than by going completly open loop, and that would be a quasi open loop mode that would rely on a model of the motor and go open loop in velocity, but stay closed loop in the displacement.  In this it would control velocity based upon this model estimating the needs, and then once the move is made, checking with the position side of things to see if the target was hit within a certain accuracy.

salbayeng
User Rank
Rookie
re: CS5047 encoder
salbayeng   7/13/2014 9:24:46 AM
NO RATINGS
I like to think of it as funnelling down to the end position, if you are aware you are in the top part of the funnel then you need more braking, if you are in the bottom part you need less. 

It's a bit like staying on a glidescope for an ILS landing , as this also funnels you into touchdown. 

The "quasi open loop model " you describe can be integrated into the trapezoidal trajectory. So if you consider that maintaining a target velocity that is some constant value above the measured velocity results in a constant acceleration (and a target velocity below actual results in deceleration). Then you take the trapezoidal trajectory , and add a constant amount to the upslope, and subtract a constant amount from the downslope. So the resultant motor motion (from using the open loop modified trapezoid , without feedback ) will be very close to that required to get the exact move position. Now all the closed loop has to do is mop up the difference. The "model" is basically just accounting for the extra force required to accelerate the mass of the carriage, so the constant offset is proportional to mass and acceleration. You do this all the time when driving and don't notice, you need to squeeze the pedal more when the car is heavier to keep up with the traffic when the lights change.

I would normally call these model terms "feedforward" ,  most of them can be pre-calculated before the move. You can also calculate them on the fly. You might end up with 75%control effort from the feed-forward terms and 25% from closed loop.  The other feedforward model term to add is to shift the aim point slightly, you will find that the control algorithm/ settings that produce the most repeatable, smooth stable trajectories will consistently overshoot or undershoot, so simply shift the aim point temporarily, then when you are almost at the destination, set it back again. No different to aiming a dart high so it hits the bullseye. Expect to use maybe 4 or 5 parameters at most to describe some simple feedforward terms.

Aeroengineer
User Rank
Blogger
re: CS5047 encoder
Aeroengineer   7/13/2014 3:11:45 PM
NO RATINGS
I am familiar with the feedforward terms from when I designed a high cycle fatigue rig for simulating flight loads for composite rotor blades.  It was a hydrauliclly controlled system that used a Delta Motion controller.  That was a fun rig.  I understood the mechanical needs of the rig, but had no clue about the hydraulics.  We hired a company to help us, and they failed miserably.  We ended up learning a lot about hydraulic design and control in a very short time (about 2 months).  The rig ended up being very successful. 

 

We were able to run at just over 1Hz with about 6" deflections on a fully reversing ~1700lb load.  In total there were 4 axis of control and the highest one had 17,000lbs of load.  The limiting factor ended up being that the feedback control was based upon load cells and not displacement.  Because of the mass of the system, this caused it to get unstable at higher speeds of operation.  Unfortuantly my boss would not let me get actuators that had linear encoders in them.  So because he did not want to spend about $2000 more, we ended up spending $20,000+ more because we could not up the cycle rate.

TonyTib
User Rank
CEO
Re: Motion solutions: motors- a mammoth post
TonyTib   7/1/2014 6:21:04 PM
NO RATINGS
Of course, if you really need acceleration, speed, and precision, the way to go is linear motor, but they aren't cheap.

Another approach is to use steppers with advanced drives, which can extend the torque curve or detect stalling.  For example, Trinamic sells chips and drives which detect stalling, and stepper drives from Copley, Nanotec, and Quicksilver can all drive steppers using servo drive techniques (and require position feedback, and cost quite a bit, although Nanotec sells integrated NEMA24 stepper + driver + controller for <$250).

Aeroengineer
User Rank
Blogger
Re: Motion solutions: motors- a mammoth post
Aeroengineer   7/2/2014 8:54:02 PM
NO RATINGS
I just got an email about the Trinamic parts.  I will have to look into their solution.  I fully expect that we will do a full trade and see what will work best within our constraints.

Aeroengineer
User Rank
Blogger
Re: Motion solutions: motors- a mammoth post
Aeroengineer   6/30/2014 4:51:08 PM
NO RATINGS
Would you mind if I put in a request to one of the powers that be that we be able to contact one another directly?

 

That or if you look me up in LinkedIn under Adam Carlson working at GE Aviation.

antedeluvian
User Rank
Blogger
Oversized I think
antedeluvian   6/22/2014 3:27:06 PM
NO RATINGS
Adam

I think 11" x 17" is overkill. The widest board we make (if you are aiming this at in-house prototyping rather that hobbyist) is 100mm wide by about ~9". If you are aiming at the hobbyist, Indon't think you would need much bigger than a Arduino shield.

Aeroengineer
User Rank
Blogger
Re: Oversized I think
Aeroengineer   6/22/2014 8:22:45 PM
NO RATINGS
Thanks for the input AD.  I think that I am leaning towards a max build area of about 8"x 10" at the largest.  Perhaps even a little smaller.

 

What are people's thoughts about this size range? Still too big.   I need to go and look at what Seeed Studio is offering for their sizes.  That might give a good calibration for this.  What other popular services are people using that offer their services in certain standard board sizes?

salbayeng
User Rank
Rookie
Re: Oversized I think
salbayeng   6/23/2014 5:16:35 AM
NO RATINGS
Hi antedeluvian  ,

I think the 11" x 17" refers to the size of the machine sitting on the desktop. 

The actual max pcb might only be 4" x 6" . (I think all the hobbyist proto shop sizes run in multiples of 50mm)

The amount of travel needed to get to the PCB and the parts, depends on how the parts are picked up , could be double the PCB area.

 

Aeroengineer
User Rank
Blogger
Re: Oversized I think
Aeroengineer   6/23/2014 10:39:07 AM
NO RATINGS
Correct, this was the overall footprint of the device, not the build area.

Max The Magnificent
User Rank
Blogger
Trading speed for size and accuracy
Max The Magnificent   6/23/2014 10:29:02 AM
NO RATINGS
Personally, I'd be happy to trade speed for increased PCB size and increased placement accuracy.

FYI I've already cleared a corner of my desk in anticipation of this unit :-)

Aeroengineer
User Rank
Blogger
Re: Trading speed for size and accuracy
Aeroengineer   6/23/2014 10:38:22 AM
NO RATINGS
What size work area are you thinking of Max?  It seems that we are talking as low as 4"x 6" up to 8"x 10".

Max The Magnificent
User Rank
Blogger
Re: Trading speed for size and accuracy
Max The Magnificent   6/23/2014 11:01:53 AM
NO RATINGS
@Aeroengineer: What size work area are you thinking of Max?  It seems that we are talking as low as 4"x 6" up to 8"x 10".

Oooh -- that's a hard one -- obviously I would like to have the capability of going as big as possible -- on the other hand most of my boards would be quite small -- it would be rade indeed for me to be making an 8 x 10" board ... and then I think about making an 8x8x8 LED cube...

How about going up to 6" x 8"?

Aeroengineer
User Rank
Blogger
Re: Trading speed for size and accuracy
Aeroengineer   6/23/2014 11:05:46 AM
NO RATINGS
I think that this may be a good thing to do a poll on.  What do you think about that Max?

Max The Magnificent
User Rank
Blogger
Re: Trading speed for size and accuracy
Max The Magnificent   6/23/2014 11:32:41 AM
NO RATINGS
@Aerospaceengineer: I think that this may be a good thing to do a poll on.  What do you think about that Max?

Let me ask those who don the undergarments of authority and stride the corridors of power -- those magnificient creatures whose names I am not fit to utter...

Aeroengineer
User Rank
Blogger
Re: Trading speed for size and accuracy
Aeroengineer   6/24/2014 12:02:43 AM
NO RATINGS
I have it on good authority that his mighty lordship, Lord Mannion would not consider such an act an act of treason, but would laude such a valient effort to bring forth this mighty, yet mystical machine.

Aeroengineer
User Rank
Blogger
Re: Trading speed for size and accuracy
Aeroengineer   6/26/2014 11:04:33 AM
NO RATINGS
Max,

Here are the sizes I would recommend for the poll cm listed first, and rough inch sizes listed in parenthesis:

 

10x10 (4x4)

15x15 (6x6)

20x20 (8x8)

20x25 (10x12)

30x30 (12x12)

 

Duane Benson
User Rank
Blogger
Re: Trading speed for size and accuracy
Duane Benson   6/26/2014 11:46:40 AM
I would vote for either 15x15 (6x6) or 20x20 (8x8). I think that would cover a large percentage of hobby boards. It would be a good compromise between desk space and capabilities.

I might be a little biased because of where I work - you might also say I've got extra insight because of where I work - but I say that if we're going to make it a personal device, we should really design it around personal use. Convenience and the ability to fit it into small workshops should outweigh the rarely used need for larger boards.

If boards are large and sparsely populated, bigger components and hand placement is an easy option. If the boards are bigger and densely packed, then it's better off going through a commercial shop.

Aeroengineer
User Rank
Blogger
Re: Trading speed for size and accuracy
Aeroengineer   6/26/2014 11:50:44 AM
NO RATINGS
Duane,

 

There is going to be a flash poll that will be posted here soon.  Make sure to add your votes for the size.  The commentary, though is equally valuable as to understanding why we want to do a certain size.

cabatear
User Rank
Rookie
Desktop Pick & Place: the big picture
cabatear   6/26/2014 10:06:32 AM
NO RATINGS
I've read through all of the various posts and comments so far, and learned quite a lot!  But I gotta say:  a fully automatic desktop pick and place with the kinds of parameters you're talking about is a pretty ambitious way to go about addressing the need to populate a few prototypes once in awhile.  While there would certainly be a "cool!" factor having such a machine, perhaps a more practical approach would be to simply improve the method of manually populating boards so that it is less tedious and more efficient.  I researched this challenge a couple of years ago and came across a great product from Abacom called EZpick, which we now use.  In fact we use it for low-volume manufacturing as well, in combination with low-cost laser-cut mylar stencils from Pololu, and a retrofitted toaster oven for reflow.

Carl

Aeroengineer
User Rank
Blogger
Re: Desktop Pick & Place: the big picture
Aeroengineer   6/26/2014 10:19:15 AM
NO RATINGS
Carl,

 

I appreciate you input, and yes it is ambitious.  Here are some of my personal motivations behind it.  I usually travel a lot.  I also have a handful of hobby boards that I produce.  The problem is that when I am home I do not want to spend the time building a bunch of boards by hand when I could be with my family.

 

So, I looked into what it would cost to have these boards assembled for say 100 or 1000, and the cost was on the order of $10-20 per board.  The boards only have 10 components on them, and they can sell for about $15, and have $5 in BOM costs.  This makes it financially not viable. 

 

But if I have a machine that costs me $500, and in my first round of 100 boards, this means that the cost to me was only $5.  This is an acceptable cost, and gets paid for in the first round of production.  At 20 boards, it is at least break even with having it outsourced.  So from a cost perspective, it makes sense.

 

Now that the cost perspective makes sense, the time also falls in line.

 

For a person that only does 1 small board per year with under 20 components, then this may not be for them, but if they do a single board with 200 components, then there is value in having a machine that is more likely to place the correct components more accurately.

Aeroengineer
User Rank
Blogger
Re: Desktop Pick & Place: the big picture
Aeroengineer   6/26/2014 10:21:59 AM
NO RATINGS
On a side note, I would love to hear any other comments that you might have on improving the throughput for hand assembling SMT boards.  I am sure that there are valuable lessons that can be learned while we are contemplating these next steps.

zeeglen
User Rank
Blogger
Re: Desktop Pick & Place: the big picture
zeeglen   6/26/2014 10:46:03 AM
NO RATINGS
@Aeroengineer improving the throughput for hand assembling SMT boards.

Design in the largest components that fit on the board, 1206 are a lot easier to manually handle than 0402. If ICs come in more than one package size use the widest lead pitch available.  Use the widest copper etch possible for physical robustness.

Aeroengineer
User Rank
Blogger
Re: Desktop Pick & Place: the big picture
Aeroengineer   6/26/2014 10:51:22 AM
NO RATINGS
That is great advice if it will work for the board needs.  A lot of my boards are very space constrained, and 0402s are the smallest I feel comfortable working with without having a lot of rework.

 

I have seen a manual pick and place machine over the last few days that has looked pretty interesting for doing smaller components.  That seems like an interesting concept, and perhaps something that could be part of the modular approach that we are looking to use in this project.

salbayeng
User Rank
Rookie
Re: Desktop Pick & Place: the big picture
salbayeng   6/26/2014 11:07:54 PM
NO RATINGS
I'm with ZeeGlen here, on hand assembly, I use 0805 parts on a 100mil pitch, with 20mil tracks. This sort of spacing means its easier to jam in extra parts should you need them, and you can cut and scratch a 20mil track to retrofit a component. 

The other tip for hand/proto assembly is to make the pads longer than needed, (i.e. hanging outside the component further) this gives you a reserve solder volume, and allows hand touchup with an iron or solder wick later. You can also put the paste on as a stripe for 0.5mm pitch parts, so that the stripe is outside the actual component leads, it will then suck up whatever it needs , the excess solder forms balls OUTSIDE the part. The longer pads mean you have somewhere to solder some wirewrap to if you need to reroute pins.

Another tip is to have some alcohol based flux is a squeeze bottle with a 28g tip, then you can apply a little drop of magic when needed. (i.e. when reworking, or with the paste striping technique).

Also ALL the signal pins (e.g. on CPU's) should be routed through a via somewhere even if it's not currently in use, much easier to do mods later (or retask the pcb) and it looks better, as the mod wiring is under the PCB .

And where two adjacent pins are connected together, join them outside the chip (direct connections tend to look like shorted pins after soldering, and if you change your mind later, you can't fix it). 

 

Duane Benson
User Rank
Blogger
Optional vision
Duane Benson   6/26/2014 12:49:23 PM
NO RATINGS
Vision might be a good feature to make optional. I super high volume machines, the manufacturing pre-engineering will ensure that all components are in a known orientation. In mid volumes, or low-volume, high mix, as they describe it, it's more appropriate to have a vision system to ensure that everything is oriented properly.

With a small home-use device like this, vision would still be a very nice feature, but I could certainly see it being optional. The user could just take a bit of extra time to make sure that the machine knows to orientation of each component.

Having things like vision, past dispensing and reflow as options will give more options in the trade-off between time and money.

Aeroengineer
User Rank
Blogger
Re: Optional vision
Aeroengineer   6/26/2014 1:45:11 PM
NO RATINGS
I agree.  We might be able to have a very basic model, as basic as it has all the slides and the vac system, and it will allow for a manual pick and place, all the way up to the full blown system where you put in the components and press a button, and you end up with a finished board.

Max The Magnificent
User Rank
Blogger
Any chance offf...
Max The Magnificent   7/12/2014 2:08:04 PM
NO RATINGS
Hi Adam -- did you see this Desktop PCB Printer project? Actually, it's more  of a desktop PCB etcher.

But the point is, there's a lot of talk about the  problem of drillig holes in PCBs. If your pick-and-place  machine were modular, would it be possible tow swap the  "place" head for a "drill" head and actually drill through-hold PCBs?

Aeroengineer
User Rank
Blogger
Re: Any chance offf...
Aeroengineer   7/12/2014 2:21:05 PM
NO RATINGS
It is something that could be looked at.  The biggest challenge would be to ensure that we were able to keep the stiffness high enough for drilling. 

 

It would be a later option, but something that we can look at.

David Ashton
User Rank
Blogger
Re: Any chance offf...
David Ashton   7/12/2014 7:32:28 PM
NO RATINGS
@Adam - you're being subjected to the dreaded "Feature Creep" from the sales guys :-)

Aeroengineer
User Rank
Blogger
Re: Any chance offf...
Aeroengineer   7/12/2014 7:39:38 PM
NO RATINGS
Hah, yes and no.  Max drives a hard bargin. 

 

Though I do have to admit that it is inline with some of the goals that have been set out for this device.  Having interchangable heads has always been one of the concepts.  It allows for you to have a solder dispensing head and a vacuum head, and then to be able to mount a camera if needed.  So adding one more head, that is not a huge deal, and because this is something that could most assuridly be done with a high speed brushless motor from the hobby side of things, then it is not all that difficult.  Designing the spindle so that it is accurate will be the greatest trick.

If once everything else is done, and it works without modification, then we will go for it.  That is after we have gotten all the other base model things working.

David Ashton
User Rank
Blogger
Re: Any chance offf...
David Ashton   7/12/2014 8:03:52 PM
NO RATINGS
@adam - I wonder if you could get a light (ie not heavy)  laser that would make holes in PCB?   itf so it would be lighter than a motorised drill.  But I don't know anything about the capabilities of laser diodes.  And you might have problems from carbonising the edges of the PCB, hence making a semi-plated-thru hole....  Do you or anyone else know more about lasers?   I had an idea to use an x-y plotter mechanism with a UV laser to print-expose pre-coated PCB without going to the hassle of having a positive transparency, I'm not sure if that's what the Bread-Box PCB maker guys are doing....

Aeroengineer
User Rank
Blogger
Re: Any chance offf...
Aeroengineer   7/12/2014 8:10:11 PM
NO RATINGS
The biggest issue is that copper is very effective reflecting many frequencies of lasers, though one thing that I have seen as an effective solution a while back was where they painted the dual layer board with black spray paint.  They then used a very cheap laser that was pulled from a blueray player to etch the paint.  Once this was done, it was placed in a dip tank and etched.

 

One might be able to take this a step further and after you have etched the board, then remove the original paint mask.  Then apply solder mask over the entire board, and etch the doler mask to expose the pads.  This could be a rather interesting thing.  I am not sure that it would work, but would be interesting to see.

 

The challenge that our machine would have is that we currently get to cheat in the sense that the path is not important, we just need to get from point A to point B.  Doing laser etching, we would have to do a specific path.  This would get trickier, but I am sure not impossible.

 

I have gotten to the point that for all the boards that are under a few square inches that it is just too easy to get the from OSHPark. 

David Ashton
User Rank
Blogger
Re: Any chance offf...
David Ashton   7/12/2014 8:18:59 PM
NO RATINGS
@Adam....hmmm...interesting.  I was thinking of using a laser to drill holes in the board material of etched boards (ie drilling the holes in the pads thru the PCB.  If your positioning was good enough you wouldn't have to worry about reflecting off the copper.  but you'd still need a fair bit of power I think?

Aeroengineer
User Rank
Blogger
Re: Any chance offf...
Aeroengineer   7/12/2014 8:23:28 PM
NO RATINGS
Yeah. most likely more than a simple blueray (I may have said bluetooth in the last post, and that would be because I am reading a bluetooth tutorial) laser could supply.  Not having played with this too much, I cannot say for certain.  I would imagine that you could get enough power to vaporize the resin in the FR4.

David Ashton
User Rank
Blogger
Re: Any chance offf...
David Ashton   7/12/2014 9:34:05 PM
NO RATINGS
@Adam - yeah you did say blueray (Do I sense an article about bluetooth coming up ?? :-)  and yes it would need something more powerful than a blueray laser I think, but I have no idea what is available.   Anyone else care to chime in on this?

Aeroengineer
User Rank
Blogger
Re: Any chance offf...
Aeroengineer   7/12/2014 9:41:45 PM
NO RATINGS
David,

I have looked into this a bit (as you might have been able to tell, I have a wide range of interests).  The power required is not all that high, though it is the frequency/wavelength of the laser.  I forget now which was best.  The other thing is the method of pulsing the laser to help vaporize the material.  I have looked into it a bit on what would need to be done to create a laser cutter.   Though at this point not too much beyond the theory stage, and enough to know that I have other more productive things that I can spend my money on.

 

As to the bluetooth, it is one of those outside curiosities that I will eventually tackle.  I am rather disappointed that it is really hard to find a good multibutton bluetooth mouse.  I have the Logitech Anywhere Mx mouse, but I have to use their pain in the butt dongle.  This is not normally an issue, but I have been away from home, and operating using my Dell Venue 8 Pro tablet.  It only has one usb port, and cannot charge and use the USB port at the same time unless you hack it.  The short term solution is to hack it, but perhaps in the future, I might look at creating my own mouse.  That would be a fun project.  I could do the eletronics for it, and then 3D print the shell for it. 

salbayeng
User Rank
Rookie
Re: Any chance of...
salbayeng   7/12/2014 11:19:14 PM
NO RATINGS
@Dave , Adam.

We looked at using a laser to cut PCB ~ 15 years ago to make some precision magnetometer parts.

(a) Your typical industrial laser requires an assist gas , so it effectively operates as as a plasma cutter, the laser ablates some material, forms a plasma, blows this plasma through the material, so with steel it behaves like an oxy cutter and actually burns the material. Without the assist gas, you can generally only cut thin perspex and simlar

(b) So you first need to start your hole (hard part), then move it along (easy part)

(c) Copper is nearly impossible to laser cut due to reflectivity and thermal conductivity, if the surface is blackened somehow then it is possible to get enough heat absorption. Problem then is the high conductivity and low melting point, so you either get nothing or a massive crater as copper doesn't ablate very well. Just burning off the unwanted copper and leaving the PCB is nearly impossible, as you need extra laser power to guarantee the copper goes away, which burns the PCB.

(d) The fibreglass PCB doesn't cut well either , it produces toxic gases, and the glass and resin have completely different absorption, melting, vaporising and charring properties, The fibreglass laying direction also makes the laser beam go wavy. You end up with a very ragged, badly charred edge, and the PCB swells at the edge. (We had to revert to water jet and CNC routing)

(e) Remember the average industrial CO2 laser (at 10micron) uses copper for the internal reflectors!

 

Aeroengineer
User Rank
Blogger
Re: Any chance of...
Aeroengineer   7/12/2014 11:40:28 PM
NO RATINGS
In another application we had some just standard G10 cut via a laser, and it did char a bit, but was not too bad.  Like you said, a waterjet cutter or router tends to be better. 

 

Here is the reference to the guy that was doing a laser etching of a paint mask, and then doing an acid etch of the resulting board.

salbayeng
User Rank
Rookie
Re: Any chance of...
salbayeng   7/13/2014 12:44:19 AM
NO RATINGS
@Adam , I saw the hackaday article about a month ago. 

Seemed to me to more difficult than the toner-transfer method, it's only viable if you happen to have a laser cutter/etcher lying around. 

I did a bit of messing around with a 500mW IR laser diode last year, trying to see if I could cut a stencil with it, but it would only melt (not properly vaporise) thin black plastic , wouldn't touch clear mylar :(  

Another mate bought vinyl plotter and uses it as stencil cutter, So I get him to do the one off proto stencils for me. 

Not withstanding all the above, there is no reason not to mount a laser on your PnP machine at some later stage. Although you need to think about laser power and wavelength.

David Ashton
User Rank
Blogger
Re: Any chance of...
David Ashton   7/13/2014 12:50:19 AM
NO RATINGS
@Salbayeng, thanks for that very detailed analysis.

Ah well, it was a nice thought :-)

flightsfan
User Rank
Rookie
Re: Any chance of...
flightsfan   8/15/2014 1:30:37 PM
NO RATINGS
@salbayeng

Funny... LPKF already has a commercially available PCB laser etching system available so the theory is sound. Their system is capable of etching copper off just about all standard materials including FR4 and Rogers.

 

http://www.lpkfusa.com/protomat/pl_s.html

salbayeng
User Rank
Rookie
LABS pcb machine
salbayeng   8/21/2014 12:40:49 AM
NO RATINGS
@flightsfan

Well so they have, I would have guessed you needed ablation to be able to blow away the copper, and that's what they have done, I'm guessing it's a NdYAG operating at 1064nm , these pulsed laser traditionally operated slowly, < 100 shots / second, too slow to make a PCB. 

Somehow they have managed to get 10-100 thousand shots / second, that's going to be an expensive piece of kit. $50k??

I have some cost effective NdYag lasers in the cupboard, that can manage one shot per second, thats a 50um diameter crater every second, take a couple of years to make one PCB.

 

salbayeng
User Rank
Rookie
Re: Any chance offf...
salbayeng   7/12/2014 9:43:40 PM
NO RATINGS
Max, Re doing drilling on a PnP machines.

 A couple of good reasons not to do this:

(a) Drilling is a dirty operation, you need to do it a seperate room preferably, otherwise you will contaminate the PnP machine with bits of swarf.

(b) Drilling machines normally have a flat base, with a sheet of MDF, then the PCB clamped on top of this. This gives a good burr free hole on the bottom. PnP machines need to be able to load double sided boards, to do this, the PCB needs to be floating, to enable placing the second side with lumpy components already on the bottom. 

(c) The drilling machine probably would want to be able to do V grooving as well, as would require a slow travel speed with lots of force and rigidity. Best really to use a CNC router for the drilling/grooving routing job.

It been 40 years since I've drilled a PCB (apart from reworking/mods) , It's quicker and cheaper just to get nice drilled and masked PCB's from the multitude of vendors out there. 

 

Max The Magnificent
User Rank
Blogger
Re: Any chance offf...
Max The Magnificent   7/14/2014 11:26:56 AM
NO RATINGS
@salbayeng: A couple of good reasons not to do this...

All good points!

mikeuk
User Rank
Rookie
XY Arm
mikeuk   7/19/2014 7:23:42 AM
NO RATINGS
I worked on a similar PnP design many years ago, and used a grid to store components, that worked fine.  IMO, you should choose a gantry/carriage design before comparing motors etc.  There's some designs of XY arms where all the motors are static, so the arm is very lightweight. For low volume PnP, the capital cost of the system is more important than speed, so I think steppers+belt are OK (all low-end 3D printers use steppers).

Aeroengineer
User Rank
Blogger
Re: XY Arm
Aeroengineer   7/22/2014 9:11:34 PM
NO RATINGS
Sorry for the late reply.  I was traveling and prefered to spend the time with the family over being on the computer ;)  I appreciate your thoughts and would love to heare some of your experiences that you had in looking at your machine.

 

As to looking at motors, this actually plays into the cost, size and weight considerations.  If one can use hobby motors, this would open up a lot of doors as well as keep cost low.  Hobby BLDC motors tend to have higher power to weight ratios than other motors, though, there are some other considerations that make them not as desireable. 

 

Once again, thanks for sharing, and I hope that you share more about your experiences.

srkh
User Rank
Rookie
Desktop pick and place carnage
srkh   7/25/2014 9:46:28 AM
NO RATINGS
I was heading up Altium hardware development in 08 when a push was made for Altium to pursue this objective. The idea was to tie in the source CAD data and the 3D PCB visualization technology with a desktop pick and place machine to make assembly a one press operation, complete with a 3D simulation of the boards assembly progress (being a PCB in 3D with components appearing as they were placed). A dry run could be simulated, showing the order and path in which components were positioned onto the board prior to commencing final pnp assembly.

The backlash we received from pick and place manufacturers was horrific - far from interested in any partnership. They clearly have software, profitable software, that a desktop PNP would make redundant. They take CAD data and make new saleable "manufacturing IP". Dont expect an easy ride to desktop pick and place, or for machines to turn up any time soon. Not without a few aggressive acquisitions first.

Altium dropped the desktop pick and place project, but 3D moved onto becoming standard.

Steve

Aeroengineer
User Rank
Blogger
Re: Desktop pick and place carnage
Aeroengineer   7/29/2014 9:08:13 PM
NO RATINGS
Steve,

 

Your story is very interesting.  I would be interested to hear some of the design concepts that you were looking at for your machine.  It is interesting to see what is trending on the poll right now.  There seems to be a neck and neck race for either a machine that can do 4"x 4" boards and and 8"x 8".  Can you share more about what you guys were doing?

Loser99
User Rank
Freelancer
Stop reinventing the wheel.
Loser99   7/31/2014 1:03:26 PM
NO RATINGS
Just buy a chinese one for 2 or 3K. like in the video below.

No way are you going to design one for under that price.

Even the cheapest of the cheap limited 3-D printers go for 800 bucks or so and this is way more complicated than a 3d printer.  If it can be done then someone else probably already did it, so buy their solution.

 

http://dangerousprototypes.com/2013/02/21/workshop-video-56-tm220a-table-top-pick-and-place-overview/

 

 

Aeroengineer
User Rank
Blogger
Re: Stop reinventing the wheel.
Aeroengineer   7/31/2014 2:11:43 PM
NO RATINGS
Your viewpoint is interesting, though somewhat counterproductive to learning and advancement.  For example, the people that are making 3D printers for under $300 decided to buck this trend.  At first did they succeed in a low cost device, no, but it cost less than the alternatives. 

 

I am not trying to be rude in my response, but there are times that you can actually do better because you are optimizing for a different solution.  The solution you mentione uses feeders.  Each one of these feeders adds significant cost due to increased need of hardware and control to the solution.  The other issue is that these are designed to have much larger build areas.  This once again increases cost and complexity.

 

This is not to say that we will be successful (me being assigned away from home for the last two months without access to the tools and desktop computer is really hampering my progress, but that is soon to be over).  Even if we are not successful, we will be able to post information about each of the parts that goes into a device like this.  Focusing on both the electronics as well as the mechanical integration.  This to me is just as valuable as having the end result.  I can say that my participation in this project so far has already brought significant learning and I have been able to meet people that I would not have before learned.

Aeroengineer
User Rank
Blogger
Re: Stop reinventing the wheel.
Aeroengineer   7/31/2014 2:16:31 PM
NO RATINGS
Also, here are the links to two low cost 3D printers well below your $800 value.

 

http://www.engadget.com/2014/04/07/the-micro-3d-printer/

http://3dprint.com/3755/new-matter-mod-t-indiegogo/

 

There were others that I saw just doing a quick google search.  I know that there are even SLA machines that can be had (which are usually more expensive than FDM) for under the $800 mark you mentioned.

 

Loser99
User Rank
Freelancer
Re: Stop reinventing the wheel.
Loser99   7/31/2014 3:03:20 PM
NO RATINGS
Those are not real products yet from companies that will not survive, or end in failure.

I can design a 3d printer that will cost me 1000 dollars then sell it for 200 dollars too.  Better yet, I should start a kickstarter campaign then take the money and live in Mexico.

 

 

Aeroengineer
User Rank
Blogger
Re: Stop reinventing the wheel.
Aeroengineer   7/31/2014 4:40:08 PM
NO RATINGS
I would be interested in hearing where you are getting your numbers are from as far as costs.  What are your assumptions? Volume? Profit margin?  I cannot think of any small companies that would be able to have ~$3.5 million in orders and be able to essentially suffer a ~$17 million loss.  In order to do so, this group would have needed to have some pretty heavy VC investing as well as a business plan that somehow locks in users long term to recoup that investment.  So far, I see the exact opposite.  This group is pursuing a high volume, low profit margin plan.  There is no Amazon Kindle effect here.

 

Once again, I am not trying to be rude, but at the same time, your claims are not passing the initial smell test do to lack of supporting details in your analysis.  There is a lot to learn from people, and perhaps you have certain experiences that can be shared.  So far though, you have manifest unsupported claims without any real substantiation.

Loser99
User Rank
Freelancer
Re: Stop reinventing the wheel.
Loser99   7/31/2014 5:02:42 PM
NO RATINGS
 

http://money.cnn.com/2012/09/27/technology/startups/kickstarter-products/

 

www.reddit.com/r/kickstarter/comments/1j6ubm/complete_list_of_funded_kickstarter_projects_that/

 

 

Aeroengineer
User Rank
Blogger
Re: Stop reinventing the wheel.
Aeroengineer   7/31/2014 5:23:57 PM
NO RATINGS
Ok, so you have shown that there are a small percentage of products that fail, but then I could counter your point with the fact that there are a larger percentage of products that succeed.  So the point you have made is negated.  Just to start, I will list a few that have become successes: Red Pitaya, Peble Smart Watch, Arduino Screwblock Proto Shield, Hexairbot, Oculus Rift, etc.  These are just the ones that I can name off the top of my head without doing any web searches.

 

Though back to the point of debate, please show me on 3D printers that I listed will cost $1000 to make but the vendor will sell them for $300.  This is an important part of debate and learning.  Substantiating your point.  I have no problem having a difference of opinion, but in order to convince me of your point, you will need to do so with specific, applicable analysis and substantiation.  I would estimate that to do this properly that it would actually be around 400-800 words.  The gauntlet is laid down my friend, I hope you accept the challenge and help us learn by a more thorough analysis.

Loser99
User Rank
Freelancer
Aeroengineer
User Rank
Blogger
Re: Stop reinventing the wheel.
Aeroengineer   7/31/2014 11:10:32 PM
NO RATINGS
I am sad to see that you have not taken up the challenge.  I really am.  Good healthy debate is ok, it helps everyone from time to time to reevaluate positions.  Unfortunately good debaterequires effort and not just three sentence zingers.

 

The first link there from Engadget actually encapsulates a lot with what we are trying to do with this project.  We are looking to go smaller than is traditional.  This allows us to reduce the size of the components as well as power requirements.  This in turn reduces cost.  I would invite you to read that story as it even serves as a good look at concepts such as engineering trades. 

Loser99
User Rank
Freelancer
Re: Stop reinventing the wheel.
Loser99   7/31/2014 11:18:54 PM
Both products in the links don't exist yet.

If they do successfully finish them, they won't make any money. I bet they lose money on them.  They are 800 dollar products that they are selling for 200 dollars and I bet they will suck.  No examples of their output.

 

 

Duane Benson
User Rank
Blogger
Re: Stop reinventing the wheel.
Duane Benson   7/31/2014 7:47:42 PM
NO RATINGS
re: "If it can be done then someone else probably already did it, so buy their solution."

I kind of think that much of what engineers do is reinvent the wheel. In fact, I don't think technology can really advance if we don't spend a lot of time reinventing the wheel. I like my kelvar belted radial tires with grippy all-season tread.

Aeroengineer
User Rank
Blogger
Re: Stop reinventing the wheel.
Aeroengineer   7/31/2014 11:12:06 PM
NO RATINGS
A nice way of putting it Duane!  Sometimes you have to go back and look at what you have already done at some point later to see if you can do it better.  For one, you should now have your lessons learned on what you liked and what you think that you can do better.

Loser99
User Rank
Freelancer
Another Failed "100 dollar" 3-D printer
Loser99   8/1/2014 11:30:42 AM
NO RATINGS
http://techcrunch.com/2014/07/29/mota-99-3d-printer-too-good-to-be-true/?ncid=rss&cps=gravity

 

 

flightsfan
User Rank
Rookie
Close already
flightsfan   8/15/2014 1:12:06 PM
NO RATINGS
We are already getting close to this as a reality. There is a manual pick and place design up on thingiverse already that would lend itself to automation fairly easily. With the addition of motors and driver software the design could readily be adapted. I was already considering doing it however I have yet to encounter the mythical "free time". Realistically the design could be built for under $300 including all motors and drivers. It would not include the ability to reflow them but that could be added later. The first step would be home building a automatic pick and place.

 

http://www.thingiverse.com/thing:385567

mikeuk
User Rank
Rookie
Squink on Kickstarter
mikeuk   9/10/2014 7:45:09 AM
NO RATINGS
Has anyone seen the Squink project on Kickstarter?  They hit their funding target in August. IMO, single sided boards is a big disadvantage, and I'm not sure the paste is ultra-reliable for production.  

Aeroengineer
User Rank
Blogger
Re: Squink on Kickstarter
Aeroengineer   9/12/2014 10:55:41 PM
NO RATINGS
The fact that there is interest in this type of tech from other groups is fantastic.



EE Life
Frankenstein's Fix, Teardowns, Sideshows, Design Contests, Reader Content & More
Max Maxfield

Steve Wozniak Reacts to Latest iPhone
Max Maxfield
9 comments
Funnily enough, just a few days ago as I pen these words, I was chatting with my wife (Gina the Gorgeous) when she informed me that -- as a kid -- she had never played at making a ...

EDN Staff

11 Summer Vacation Spots for Engineers
EDN Staff
20 comments
This collection of places from technology history, museums, and modern marvels is a roadmap for an engineering adventure that will take you around the world. Here are just a few spots ...

Glen Chenier

Engineers Solve Analog/Digital Problem, Invent Creative Expletives
Glen Chenier
15 comments
- An analog engineer and a digital engineer join forces, use their respective skills, and pull a few bunnies out of a hat to troubleshoot a system with which they are completely ...

Larry Desjardin

Engineers Should Study Finance: 5 Reasons Why
Larry Desjardin
46 comments
I'm a big proponent of engineers learning financial basics. Why? Because engineers are making decisions all the time, in multiple ways. Having a good financial understanding guides these ...

Flash Poll
Top Comments of the Week
Like Us on Facebook
EE Times on Twitter
EE Times Twitter Feed

Datasheets.com Parts Search

185 million searchable parts
(please enter a part number or hit search to begin)