Test Points

Since connectors are expensive and bulky, we designed circular contacts into the PCB layout for testing the circuit and flashing the MSP430. I made a simple "bed of nails" test jig with spring-loaded contacts to mate with these pads.

The jig is made from perspex, glued with a plastic solvent cement. Two grooved blocks hold the back of the robot PCB. The front is supported on a perspex column. The test pins are push fits in the arm which swings down and locks, compressing the sprung contacts against the board. The lock is on the far side, not shown in the photo. The transparent perspex allows IR to pass for communication tests. The wheels are free to spin to enable testing of the drive circuits and programming. A ribbon cable carries the test signals to a breakout board on the back of the jig, which contains a row of cage clamps and a 14-pin MSP430 programming connector with the standard TI pinout.

If you build a similar jig, make the arm longer. On this one the radius of the arc is too short, so instead of coming straight down onto the pads, the test points drag along the surface of the pad for 0.5mm or so, scratching the copper. The test pins are very sharp; round-ended ones would probably be better. Alternatively; come up with a mechanism where the bed of nails is lowered straight down rather than in a circular path.

This test jig is great for quickly flashing the robots and testing the PCBs during assembly. However, if you have the PCBs assembled commercially, and usually program the robots over IR (although I believe this feature requires a little more development), you may not need something so complex. The test pads are placed on a 2.54mm grid, so you could make a very simple bed of nails from stripboard or similar, with no precision machining. That said, this jig was made in one evening mostly using simple hand tools. The holes for the test pins were drilled on a milling machine for precision, but could be made in a pillar drill if marked out carefully.

categories [ ]