Pogo Pin Misalignment

[etothejtheta]

If this is the wrong place to post, forgive me. It looks right to me.

I have a major problem that I'm in urgent need of rectifying. I built a test fixture to test two radio circuit boards. The user interface to one of the UUT boards breaks out signals, power and ground into an array of 80 pads to be contacted by pogo pins and delivered to the test fixture board. Upon receipt of the test fixture boards I began testing it. I found that I was getting a number of dicontinuities between the UUT board and the test fixture board. I immediately suspected pogo-pin to pad misalignment. Both PCB files were checked. I found that with odd columns aligned to dead center, there is a 20 mil horizontal differential between all even columns and respective pogo pins. An x-ray of the boards confirmed this finding. Attempting to wire from the pogo pins to the test fixture board has proved futile. My question is this; has anyone out there experienced a similar issue and come up with any good solution? I have suggested having a flex made with matching pad patterns on each side and some type of metal nub mounted to the side that hits the test board pads. This has been shot down as possibly causing other mechanical stack-up issues. Does anyone have anything? Please help.

 

[VEBill]

Once upon a time, we did a great deal of Circuit Card Assy (CCA) testing using In-Circuit Testing (ICT).

Our fixture were built from the same Gerber (or similar, whatever it was at that time) file, so that pin-alignment was *never* an issue. Not once that I can recall, with hundreds of different CCA designs and perhaps 100k pogo pins total in hundreds of fixtures.

Note: For ICT we had a pogo pin on EVERY node.

Our pogo pins were primarily on 0.100-inch spacing. They pogo pins had toothed cups that landed directly on the soldered pins (not special pads). In our case, a 0.020-inch misalignment would probably be almost (but not quite) good enough.

Can you rework the test fixture?

Are the PCBs made to spec? Is there time to go back to the PCB house?

 

[MacGyverS2000]

Seems like the easiest "fix" is to rework the testbed assembly. Make a precision jig to hold the pogo pins in place while they're being soldered, if necessary... this can be nothing more than a small block of aluminum with holes drill into it via CNC.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[VEBill]

Our ICT fixture have wire-wrap pogo pin sockets, with an easily replacable, spring loaded pogo pins that slid into the fixed socket. Everything is gold plated.

The substrate that held the pogo pin sockets is simply a thick fibreglass panel, just like a blank PCB except much thicker. As mentioned earlier, it was drilled with the same file as the PCB.

 

[etothejtheta]

Thanks for the feedback, everyone.

I'm still not sure how/why this happened in the first place. Going out-of-house for the layout didn't seem to work so well. Anyway, management has decided to do another board spin to correct the problem (in-house layout this time). In the meantime with minimal wiring from pogo pins to test board pads, we can bring up the UUTs and displays and communicate to the boards over USB. This at least allows us to load the latest code into the boards until the new test fixture boards come in. I'm still trying to get one unit up well enough to test RF decks. Thanks again for all your suggestions.