One of the substantial changes for the TESLA 3000 was the incorporation of the DR2/TESLA power supply right onto the MPB. The DR2/TESLA power supply was an amazing accomplishment, able to operate full power at reasonable efficiency from 38Vdc to 300Vdc. A custom buck switchmode pre-regulator was used to bring the input voltage down into the range that a Lambda 48V half-brick DC-to-DC converter could handle – which was 36V to 72V nominal. Below about 54V input, the pre-regulator would lock into full-on, passing through with minimal drop, giving optimal low end operation.
When testing the original DR2/TESLA power supply a decade earlier, we had an elaborate procedure that required a sequential voltage bump, checking critical operation parameters as the test proceeded. For instance, first going to a low voltage just above where the pre-regulator would start switching, and verifying that it was switching at the proper frequency, and that the output voltage was reasonable and within range. Without ensuring that the pre-regulator was working properly, a high voltage pass-through would destroy not only the DC-to-DC converter, but also the large reservoir capacitors that we had at its input. In fact, the reservoir capacitors were dangerous, exploding when overstressed!
Somewhere along the way, the cautious approach was forgotten. The contract manufacturer making the TESLA 3000 MPBs was experiencing explosive failures in testing. Their staff was afraid of the board. They would put a box over the board before testing.
Elecsys put its experience with GPIB to good use. We obtained more GPIB equipment and a Xantrex XHR300-3.5 power supply, which had a serial interface. I wrote the first draft of a PERL program to direct the fully automated testing of the TESLA 3000 MPB power supply, then had Nishant Dhruve complete the work.