The bare boards came in yesterday for my tiny laser driver, and people started assembling the first one, and I had no way to test it. The real test set boards won't be in until next week.
So we hacked this up, out of scraps of FR4.
The laser connections are usually via wires that the customer solders to the board, but we made the holes a tad oversize so we could use 0-80 screws to make the connections to spacers during testing. I wouldn't trust pogos at 200 amps.
Phil is right, superglue works fine on copperclad, seems better than epoxy. I scuff it up a bit first with an emery board. Walgreens has all that.
It all worked first time, thanks to massive amounts of review and checking. This uses the boost-doubler power supply topology, which now needs a bit of tweaking for the application but is all understood and tunable. It ran overnight at 200 amps into a 0.1 ohm dummy load, 4 kilowatts of pulsed power.
Things are getting weird when 0805 resistors and 2-56 screws start looking big.
One assembly problem we had was that the tiny header connector pins wouldn't fit into the board. The pins are 21 mills diagonal, and the holes were spec'd as 25 mils, but the connectors had to be pressed in. I specified 2 oz copper, and we're guessing that the board house didn't properly account for whatever heavy-duty plating they used.
And we forgot to radius the corners of the board. I hate boards with sharp square corners.