Had you checked the Gerber files with a Gerber viewer before sending the board to be manufactured?
On the other hand, a few years ago I used the Batch PCB service from SparkFun to make a board that had a filled outline (copper pour). I specified the RS-274X file format, with G36/G37 enabled. My Gerber viewer (ViewMate from Pentalogix) and the Batch PCB design checker both showed the copper pour was OK. But the board that came back only had the thin line around the outside edge of the copper pour. Luckily I had also used thick traces (this was the important GND net) to connect the pins, so the circuit worked. Maybe the board shop thought that the copper pour and the traces weren't both needed, so deleted the pour? Or they just accidentally deleted the copper pour? Who knows. When I sent an e-mail to Batch PCB, I never got a response. By the way, another copper pour on the other side of the same board was OK.
I am currently working on a board design that I plan to send to OSH Park. When I used ViewMate to view the Easy-PC v17 generated Gerber files, I was surprised (but shouldn't have been) to find that they did not totally match my design.
I had used 0 width lines to specify various filled outlines (yes, the Gerber specification explicitly says they are acceptable for creating G36 regions), but Easy-PC had changed them to 1 mil lines, without warning me. This caused all areas to grow by 1/2 mil on each side, which caused some clearances to be less than required.
Also, I had used some rounded rectangle pads. Easy-PC didn't use arcs for the corners, but instead generated 2 or 3 line segments to approximate the arc, which looks ugly. The Gerber specification has an example of how to write a macro to generate rounded rectangle pads -- it's actually easier than using line segments.
Not using enough resolution for the Gerber coordinates can also make the manufactured board vary from the design due to rounding errors. Ucamco (who produce the official Gerber file specification) actually suggest using 6 decimal places for inches, and 5 for millimetres.
While I agree that there are pros and cons to every shop that you use, I did not go into this blindly. I understood that there was no check of the boards, I did not pay for that, but at the same time, I was able to get quality boards boards for a price much cheaper than it would have cost me for larger boards. Judging by the fact that there are few if any complaints about the processes that are used by OSHPark, I would suggest that their vendor has their process well under control and that if designed within the tolerances of their process you should have no yield problems. On the other hand, if you cut into the design allowable by 33% and not expect to have any issues, then you are fooling yourself. In that I do not blame them or their process.
I agree that there could be errors such as this that could be found via many methods, and that file conversions can lead to errors, in this case it was faster to encounter the error and learn from it. As you can see, the design is quite small and I was able to salvage all the expensive components. I may have felt differently about a larger board.
The do give great service at a very reasonable cost. There are very few companies that offer an a la carte method of creation of proto designs with these design allowables. As you can see, they have had great service.
Well my fist thought would be why the CAD SW designer had a separate switch for copper pours, clearance is clearance, and why you would not want to do a clearance check I don't know. I use OrCAD and Altium and I'm not sure about Altium, but OrCAD had a DRC button that checks everything. You can set the rules for pours differently (as you should) but DRC for clearances is DRC for all clearances. Sounds like designed by monkeys.
My second thought is why your PCB maker doesn't do a DRC check against their capabilities before they start the job, it seems totally silly unless they don't warrant their product and so don't care.
My final thought is why they didn't do a bare board test? My supplier does BBT on all boards at no extra cost, I might be loading $1000 worth of parts onto what is only a $5. bare board, why would I want to use an untested part? Would you build the whole board from untested parts from the manufacturer? I doubt they would sell untested parts because their reputation is important.
I don't know, maybe I'm an OCD sufferer, but for me the cost of repairing production boards would far outweigh the cost of a supplier that does DRC & BBT.
I agree with your first point. The software I use is not my favorite, but until I find something else, I have stuck with it, though I am looking.
On the second point, it is hard to complain when I went into it knowing that they did not do a check on these things. If I had paid for it, that would be one thing. For me, I was able to get all 15 boards I ordered for less than $5 shipped. If I had used another service, it would have cost me at least $50 because I would have had to go to a service that could provide better design allowables.
For that, I am happy with the service that I have received and have learned a valuable lesson.
Well that is of course fair enough, I just wonder whether the the savings are worth it when one considers that our time on earth is rather finite so that time becomes one of our most valuable commodities
Ah yes the opportunity cost of things in life. In this case, I lost perhaps 2-3 hours. To pay for a service that offered similar design limits for a proto board, I would have had to work a few extra hours, so it was probably a wash. Now, if I had my family with me while I was doing this, I would not have worried aobut it and spent time with them, but they were about 9 hours away when I was doing this, so it would have been this or perhaps watching an episode of NCIS.
I've built quite a few boards at OSH Park recently and I've generally been happy with the results. While they don't run a DRC for you, they do provide layer by layer and composite checkplots that make common errors obvious.
I did have a problem with too small annular rings when I got aggressive with my design rules. Dan Sheadel responded quickly confirming that I met their design rules and refunding my payment for the mis-drilled PCBs. While I agree with Etmax about the risks of using untested PCBs, I've been able to prototype complicated designs (ie. 0.8mm pitch BGAs) on PCBs costing me only ~$70 which is hard to ignore.
I agree with what you have said here about their value proposition. I just wish that I could find a similar service that would allow prototyping with .4mm BGA in a similar a la carte fashion. I cannot justify a prototype for this chip that I want to use with a board size of .38" by .38" and have to pay at least a $100 or more to do it.
I did a 0.5mm pitch BGA deisgn at OSH Park, but I had to cheat on the escape routing since their width/clearance rules just aren't small enough. I you find a similar service with finer tolerances than OSH Park be sure to post about it!
Thanks for sharing. I will have to take a look, do you detail in your post how you "cheated"? If not, I would love to hear about it. I am itching to try a board with the Freescale KL02 chip. I have thought of a few ways that I might try as you said, cheating the allowables, though not sure if they would work. For a one off, it might work.
The BGA that I routed was a Xilinx CPLD and only had two concentric rows of pads. I escaped the inner pads by routing thru (connected to) a pad on the nearby outer row. This cheat worked because I was only building the design to test my ability to solder BGAs. I used two separate loads of CPLD code where one or the other connected pads was tristated.
I've seen similar tricks work on large pin count BGAs where only a few of the balls connected to the die. Inner pads were routed thru external pads that were not connected to the part's die. Unfortunately it's a such a rare use case that it's almost useless.
Ah yes. I had though about perhaps not connecting a few of the outer connections to allow for some of the inner connections. This still could work if I did not need all the connections, I was just not that hard up on using the chip, but is something that I may yet consider if I really get to the point that I want to use it. Thanks for sharing!
I use a service that does 2 x 8"x10" of PCB material for $300 plus shipping. They do 7/7 track & spacing with 8thou vias and 0.2mm annular rings. I just put a dozen or so designs together on the panel. Includes BBT and DRC and I can get gold flash for an additional $50. $300-$350 might sound like a lot but 2 panels of 8"x 10" with as many designs as I want if I do the panelisation myself it's hard to beat for an all inclusive. Comes with both sides silk and mask. At the end of the day we have only so much time. We can always get more money but never more time.
I appreciate the comment, though with my last 10 projects, I could fit all of them on your board and still have a lot of space left. I would still be out $200+ on the cost of the boards and I would not have been able to have them produced there because I have parts that require the 6/6 spacing that is offered.
I do get the point though that time is precious, but at the same time, it I spent all my time worrying about what I should be doing with my time, I would not have learned many valuable lessons of life by either the failures I had to solve, and hence take more time, or the successes that I have had because I would be worried I might waste my time in failure. There are costs associated with everything, like right now. I could be making dinner, showering, sleeping, or enjoying this sharing of differing viewpoints. All cost time and a lost chance to do something in the moment, but I decided on the last option as it gives me a chance to interact with someone as well as understand a different viewpoint.
By the way, which is the service you use. Always good to know of vendors that others would recommend.
Well one can't argue with that, Utimitely everything is horses for courses :-) I tend to do 20-30 designs per year all requiring small proto runs and seem to work some 18 hours most days which makes me I guess more time conscious than most when wanting to find time for family. I also very much value the time spent here sharing with others so I don't under estimate that value either. Usually 10-20 proto services per year. Sometimes a friend wants a board and I add it to the panel and depending on size they pay only a few $.
The company is called Entech, This link gives the full range of specs:
Thanks for the info. I am sorry that you have to spend so much of your time away from your family. Thankfully, I get to spend much more time away from work. I hope that you can slow down a bit. That is not fun to have to work that much.
If you want to play with the KL02 microcontroller, the evaluation board only costs U$12.96 at DigiKey.
The 3x3mm QFN-16 package wouldn't fit on your 0.38"x0.38" board?
Like you, I design very small boards. My current project is only 0.29" x 0.31", and has parts on both sides. It's easy and fun to design these projects, but what a pain to build and test and debug something that tiny!
I agree about how much easier board building has gotten. I'd never use some of these quick and dirty techniques when I'm on a strict timeline, but it has really opened up pathways to build quick, 'usually' reliable prototypes.
Yeah, I find the small boards a nice challenge. I love trying to fit everything on them that needs to get there. What is your current project that you are doing that is that small? I would love to hear about it.
On the KL02, I am just exploring being able to use it. Once I find a way to do it, then I will find what to do with it ;)
I've been using OSHpark (previously Dorkbot/Laen PCB) for several years. The price and turn around time is pretty darn good. Being 'down under', affordable fab houses are non existant. I couldn't find anyone who could come close with my design's requirements - 4 layer, 5/5mil track/spaces, 4mil annulars and 10mil holes.
Their prices are unbeatable if you want lots of boards too.