Hello,
Would it be possible to provide a gerber file which shows the precise connector placement of the Alchitry interconnect connectors and mounting holes? Clearly, in order to be able to best mate a custom board, having precise dimensional reference for both the connectors and mounting hole would be advantageous.
As this is Open Source hardware, we are keen to be able to contribute to the library of component which could make use of this very innovative platform.
Much Thanks!
We have the lbr files https://cdn.sparkfun.com/assets/c/8/8/f … ements.lbr and https://cdn.sparkfun.com/assets/c/8/8/f … ements.lbr and the dimensional drawings http://www.4uconnector.com/online/objec … /19022.pdf http://www.4uconnector.com/online/objec … /19008.pdf which are what we use in-house
Perhaps the first message was not clear.
We are intending to create a daughter card for the board, and as such would need to have precise placement of each of the four mating connectors on the FPGA board so that ours can properly mate.
Hence the request for gerber of the PCB so we can assure we position our own connectors correctly.
Thank you.
The 2 types we have are the lbr linked above, which has precise measurements (open in freeware AutoDesk Eagle or similar…some work in-browser) and adjust grid settings/click on things to see measurements, or they have .iges files too here https://alchitry.com/boards/au
Hi Russell, It seems that we’re still not quite on the same page here…
The reason we are looking for the gerber of the board is that there are four connectors, and all of those four have to be in the correct location for another board to mate with this one. The gerber of the actual layer which has the connectors is the only 100% certain way to assure that we can extract the dimensional relationships between the four connectors on the board. So, we’re not asking for the footprint of the connectors, but trying to reliably place mating connectors so that when we fab our mating board, that it actually connects to the Alchitry board.
Using the Gerbers means that we’ll be working off of the exact dimensions the boards are fabbed with, which assures that if we adhere to the same placement (ie. our gerber and the Alchitry gerber line up perfectly) then we are assured that our board mates when we get it back. Without being able to verify this we risk wasting money fabbing a board which may have placement errors, and this would require a re-spin. Our project is not in a position to be able to do this by trial and error, and since this is open hardware, providing the gerbers (of just one layer) should be no big deal, since you guys are manufacturing the boards. These are fine pitch connectors, so being off by even a few mils will create mating issues.
In general, it’s ill-advised to use such connectors for this type of application, actually, as these connectors have no physical “pin” to force registration. The unconstrained connectors can drift during reflow, causing misalignment errors. We are trying to minimize the compounding of placement and component drift errors by starting with at least an identical placement of the connectors. We may further reduce pad dimensions to help constrain potential float/drift issues, but this is on top of assuring that the connectors themselves are placed in the exact same locations relative to each other on both boards.
Hence in order to assure all of this, we need the gerbers of the layer on which the connectors are placed so that we can in-turn place our connectors in the same positions relative to each other.
Thanks in advance for your help with this.
Was this ever resolved? I’m in the same situation. Currently designing a test card to mate to the Alchitry board so i require the gerbers to understand the exact placement of the 4 connectors in order to mate correctly.
Fulcio:
Hi Russell, It seems that we’re still not quite on the same page here…The reason we are looking for the gerber of the board is that there are four connectors, and all of those four have to be in the correct location for another board to mate with this one. The gerber of the actual layer which has the connectors is the only 100% certain way to assure that we can extract the dimensional relationships between the four connectors on the board. So, we’re not asking for the footprint of the connectors, but trying to reliably place mating connectors so that when we fab our mating board, that it actually connects to the Alchitry board.
Using the Gerbers means that we’ll be working off of the exact dimensions the boards are fabbed with, which assures that if we adhere to the same placement (ie. our gerber and the Alchitry gerber line up perfectly) then we are assured that our board mates when we get it back. Without being able to verify this we risk wasting money fabbing a board which may have placement errors, and this would require a re-spin. Our project is not in a position to be able to do this by trial and error, and since this is open hardware, providing the gerbers (of just one layer) should be no big deal, since you guys are manufacturing the boards. These are fine pitch connectors, so being off by even a few mils will create mating issues.
In general, it’s ill-advised to use such connectors for this type of application, actually, as these connectors have no physical “pin” to force registration. The unconstrained connectors can drift during reflow, causing misalignment errors. We are trying to minimize the compounding of placement and component drift errors by starting with at least an identical placement of the connectors. We may further reduce pad dimensions to help constrain potential float/drift issues, but this is on top of assuring that the connectors themselves are placed in the exact same locations relative to each other on both boards.
Hence in order to assure all of this, we need the gerbers of the layer on which the connectors are placed so that we can in-turn place our connectors in the same positions relative to each other.
Thanks in advance for your help with this.
I was looking for this info too. Apparently this guy figured it out. (https://github.com/SengerM/Alchitry_Cu_KiCAD). I get they don’t want to hand out Gerber files, but coordinates from part centers would work. It’s interesting you mention connector drift during reflow. I’ve had that problem with the new raspberry pi compute module 4 2x 100 pin connectors. I fear I’ll have that problem with these connectors too, what do you suggest to avoid this?
