(FIXED!) BK Precision 1740 Repair

[staze]

Hi All,

Picked up a 1740 to repair, and while I got the schematics from BK, they don't have any kind of BOM available. =/

Anyone have one of these units that they can confirm the transistor model/type for Q2 (the only TO-220 with a heatsink)?

Previous owner swapped in a NTE331, which is a TIP3055 equivalent (I think). Pulling it I noticed the heatsink was loose, and the gain was only about 50 (on the pretty low side of NTE's spec). I just want to double check, if I can, before I go buy a replacement.

Thanks!

[staze]

For reference, I'm getting a full 105V across the output and the current meter pegs. Which I'm assuming means the feedback circuit is bad. If I unplug the 3055 transistors, I get relay clicking, and the current meter doesn't peg, but resistors R1 and R3 (which are 10k, not 4.7k in the schematic) both end up dropping a full 110V and heat up... fast.

Open to help. I've swapped out a few things that the previous owner had done, including C3 being toast.

[oldway]

This power supply seems to me quite simple and should not pose many problems to be repaired. 
It seems to me that the former owner did not know much about electronics and did just about anything.

With 105V of no-load voltage on C2, Q2, Q3 and Q4 should have a Vce voltage of at least 150V.
2N3055 are obviously not adequate.
In TO3, I would choose the MJ15003 of ON for Q3 and Q4 and in TO220, MJE1503 for Q2

[staze]

Oddly, the schematic only shows 2 3055’s where there are actually 4.

What I’m not sure about is where to start. Yes, Vce across those transistors is probably obnoxiously high. But obviously something has failed to bring me to this state. Shorted cap/transistor/etc.

Where would be a good place to start looking? Am I wrong that unhooking the 3055’s would be reasonable to isolate them? Or maybe that’s the intent (to have 100v across those transistors).  Looking at the table, and confirming with my unit, C2 is actually rated for 100V, at 6,000uf.

[floobydust]

Note that the control-circuit common (IC2) triangle-gnd symbol connects to output (+) at P6.
The entire +/-13V rails float with the output(+). This is a bit confusing and if you get this, much easier to troubleshoot. I would keep my multimeter GND on the output (+) terminal when measuring the control circuit voltages.

Q2 is typically a TO-220 transistor like TIP31C, 2SD313. This power supply design has no current-limiting resistor for Q2 so it is a weakness.
Check all the pass-transistors and their drivers. They usually fail short C-E.

The relays select transformer taps to provide the lowest necessary DC voltage for the pass-transistors, to lessen heat dissipation.
The table at sch bottom says K1 and K2 on gives 75V max. at full load.
If the PSU is stuck at full output voltage, K1 and K2 are on as they should be.

[staze]

So leave everything hook up during testing, or would it make sense to pull anything?

Given the 100v dropped across those resistors, and how hot they get (dissipating 1w), would you suggest a fan or anything blowing on the board?

I am using an isolation transformer, but obviously that 100vdc is going to hurt if I complete the circuit.

The replacement I used for Q2 probably isn’t going to survive. It’s a 2N6487, which seems to only be rated for 60v. :/ But I’ll check things out. The NTE331 (which looks like a TIP31C) May be fine, just has crap hfe.

Thanks. I’ll start looking at transistors.

[floobydust]

R1 is a bleeder resistor for the raw DC and should get hot, especially 100VDC/10k is 1W.
R2 is a dummy (minimum) load for the regulated side and will also get hot at 100V output.
I think 60VDC is the supply's max. so this is more than usual.
But why is C14 only 25V... What are the capacitor voltage ratings, like C2 and C14? Watch they are OK with 100V.

If you unplug the pass-transistors, the output should go to zero volts, relays K1 K2 off to give around 30VDC instead of 100V across the big cap C2. This allows some measurements.

I suspect a shorted pass-transistor and possibly open emitter resistor (RA4,5 RB4,5 etc.) as the ammeter should not peg with no load.

I would leave them unplugged and measure the voltages on each pin on IC2, to see if it is OK. If you post those, that can help.
I'm a bit worried C3 failed, that could mean the +13V rail got over-voltaged.

[staze]

Ah hahahahahaha... thou shalt check voltages.

So, the caps are all 100V rated. And the RA4,5 and RB4,5 all are correct readings.

Checked the IC2 pins, and things just didn't seem right. The power rails were pegged at 16.5V or so. So, looking at the schematic, I noticed D2 is what derives the negative power rail for IC2. It's supposed to be a 4.3V Zener. What has the previous owner soldered in... a 1N4006. DOH.

Swapped it out for a zener that was close enough (measured it at 4.27V), and fired things back up. Bam! It works! Went through the calibration process in the manual, and everything checks out. I've got some replacement, matched 3055's on the way that I plan to swap out (what's there are 3 ancient RCAs (hfe of 30's to 50) and a new ST (hfe of 250!!!)).

But otherwise, it seems to work great. Total outlay in parts... about $3. I had the zener, the cap, a replacement LM324 and 723 in my parts bin. Just had to buy the Q2. If I include the 3055's on the way (which Radioshack is blowing out for $1.50 each), I'm out $9.

Thank you floodydust!!!