20100123

the junkman prevails

My outfit was spring-cleaning not long ago, and a TrippLite PR40 was destined for the heap.

It's a 40-ampere DC power supply. Well, it was, but it wouldn't supply any DC recently, and it was part of a suite of electronics that had been forklift-upgraded anyway. The proponent of the forklift upgrade didn't ask for it to be turned in.

Your humble narrator chose to snag it instead of letting it go to the heap, because I need a supply for other amateur radio gear that comes along.

This supply had no output voltage at all. I opened it and took the regulated DC off of the final board. Bingo, 22VDC. The crowbar circuit on the final board was shorting the output to ground, as designed.

TrippLite does not offer schematic diagrams for its products. It offered to take this PR40 back and offer me a sweet deal on a refurbished PR40. Total with shipping (this units weighs about 30 pounds) would have been around $200.

I asked around on qrz.com and was told that, likely, there are 4 NPN power transistors in parallel that clamp raw dc down to 13.8VDC, and one of them is smoked. If I wanted to go the incremental route, first change the IC that was applying the bias to those 4 transistors. For comparable power supplies, like an Astron, there's an IC just to do that. The TrippLite does not use that part, but 3 of the 4 opamps on a quad IC instead.

The quad opamp came for about $10 shipped. That didn't do it, replacing the power transistors would come next. I parked this project in the man cave.

The project came back around on the lost-cause rotation, so I sought out a good price on 2N3771 in TO3 cans on eBay. Somebody took my Make An Offer offer, 4 of them plus shipping for under $10.

I put them in tonight. With their outputs still away from the final board, dammit the output voltage was still 22V.

I touched up the joints on the opamp pins and tried again. Still nope. As I started to pack the tools away, it occurred to me that the IC on the final board might not be able to bias the transistors to cut the raw dc down to the desired 13.8V, if the outputs of the transistors are not coupled to the final board.

With the regulator outputs connected to the final board, the output voltage settled at 13.86 or so VDC.

I assembled the supply as far back to factory as possible, and found that I needed a new way to insulate the output post from the cabinet. The original supply used a stack of plastic washers, which disintegrated as I disassembled the supply the first time.

I improvised a pair of insulators from scrap lexan, one inside and one outside the cabinet, with a dab of silicone caulk to keep the red post centered in its hole through the cabinet.

I'll go back inside in a few days and put a rubber O-ring around the red post to better center it in the hole so it doesn't short out.

So now we have a 40A regulated DC supply for about $20. And we also can hook on to the raw DC in this supply for DC of higher voltage than the pickup truck would provide, the next time there's some anodizing to do.

No comments: