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?
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.
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.
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.
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.
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' ;-)
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.
image may need to be opened in a new tab, due to the width constraints of this blog ;-)
Replay available now: A handful of emerging network technologies are competing to be the preferred wide-area connection for the Internet of Things. All claim lower costs and power use than cellular but none have wide deployment yet. Listen in as proponents of leading contenders make their case to be the metro or national IoT network of the future. Rick Merritt, EE Times Silicon Valley Bureau Chief, moderators this discussion. Join in and ask his guests questions.