Shunt regulator readings [resolved]

[Bingtou]

I've finally soldered and installed the shunt regulator, and run the voltage test. At least the previous problem I posted about, with the voltage taking minutes to bleed off, has resolved as Paul suggested it would.

I ran the test with a step-up transformer to bring the input voltage from the mains from 100V up to 110V. I did this also in the previous testing stage, re-running my earlier tests to get voltages closer to the manual's instructions, and that worked out.

The shunt regulator readings are odd though: the voltage climbs above 300V in both cases, but drops to 144V for the wire from D1. The wire from D5 goes to 222V, so close enough to the 225 specified in the manual.

The four LEDs are on and steady, and the tube heater glows as expected.

 

[Paul Birkeland]

How are your Kreg voltages?

 

[Bingtou]

Is that still measured with T23 for the black alligator clip?

I got 4.9V at the Kreg on the A side, and 1.86V on the B side.

 

[Paul Birkeland]

Yeah, so that's one side that is doing what it's supposed to and one that is not.  Can you post some photos of that area of the board? 

I definitely would not move past this stage until the regulator board is all the way up and running.

 

[Bingtou]

I have 3 photos so far, this is the best I can do for the moment. Lifting the board and getting decent shots of the D terminal is going to be a lot more fiddly, I don't think I'm going to manage that this weekend. So far, I've looked at solder joints and installed parts and haven't noticed anything; also the wires seem to be to at (and to) the correct terminals so far, but I will have to go back over the entire set of connections and parts for that whole section to be sure. Again, with the time I have left this weekend, I'll probably have to put that off until I can really set aside a few hours. In the meantime, these are the shots of the top of the board.

 

[Bingtou]

Also, I assume you mean that the problem is on the B side of the board? I have taken some additional closeups of that area, though I'm not sure if they show much that isn't visible in the first three photos I posted. I'll add them if you think it might help, but just wanted to confirm first that you're talking about B, not A.

 

[Paul Birkeland]

Some things I would look at:

1.  Check that the wires going from Kreg to pins 3 and 8 on the 9 pin socket are going to the correct places.  The same goes for the wires that attach to pins 1 and 6 on the 9 pin socket.  When I have switched a pair in the past, I have experienced strange behavior like this where half the regulator will work and the other half will not.  If one of the 220 ohm resistors going to the D socket is damaged or poorly connected, it could also cause odd problems like this.

2.  Check that the 2N2907 transistors are in place and the 2N2222 transistors have not been accidentally substituted.  I don't think this is an issue for you, but I'd look.

3.  Check that the TL431 regulators are properly placed and not MPS4250 instead.  I also do not believe this is your issue, as I believe this makes the Kreg voltage take off pretty high.

4.  Check that the R1 resistors are the correct values. 

5.  Flip the PC board over and be sure everything is well soldered. 

6.  Set your meter to beep for continuity and be sure there are no shorts between any pair of legs on the TL431 regulator on that side. Repeat this test for the 2N2907 and MJE5731A.

-PB

 

[Deke609]

It might be fine on the underside, but from the topside it looks like the circled connection point in the attached could use some solder.

 

[Bingtou]

Thanks Derek, I believe I checked that the other day to make sure the connection is good, though I may add some extra on top to be sure. I think all the connections on this board were soldered from the underside at the time of stuffing the board, and I did look pretty closely at those connections at the time. They appeared fine, but it's not apparent from the top.

Paul, thanks for your assistance and patience, I'll take a look at all those points you mentioned. I've already had a look (visual check so far) for any signs of shorting and there were none. I'm going to look closely at the 220 ohm resistors to start with, because soldering those in was no picnic.

 

[Bingtou]

I've done all the checks I'll have time for this weekend, unfortunately further checking will probably have to wait until next weekend.

Resoldered one of the 220 ohm resistors, just to be sure there was a clean joint and good connections.

Continuity tested the legs on 431 regulators (both sides), no beep. Continuity tested legs on both MJE5731A transistors, no beep. I haven't done this test for the 2N2907A's, because I need unrestricted access to the bottom of the board for that, which can't happen until I undo some wire connections. It doesn't just flip over...

Confirmed that both R1 resistors are 37.4 ohm. Confirmed that 431's are definitely 431's.

On retest, I still get the same voltages for the 2 red wires. When I get time to come back to this, I'm ready to take off the board and check all underside solder joints, look for shorting, and then redo all soldering of D1 terminals and wires, in case the problem is there. I've looked at it and all seems fine (no obvious shorts), but going back and redoing it step by step may be the best course now.

(Incidentally, I did get a better close shot of the B side of the board, which I attach here. But I don't expect it will reveal anything important.)

 

[Bingtou]

I've unsoldered the connecting wires/resistors from the D terminal to the board so that I could fully remove it and carefully inspect the underside. No sign of soldering problems or short circuit. That's as far as I'm going to get midweek, but when I can come back to this, I'll inspect all components on the top of the board one by one; if they seem to be in order, I'll take off and resolder all the wires/resistors on terminal D and then reattach them to the board. Then retest.

In the meantime, I've noticed that I'm going to run out of red wire. If I want to order some a few metres of the stuff (and some black) from somewhere, which I may do just to make sure that I always have some spare, what do I need? I measured the kit wire at 0.7mm diameter for the core.

I am assuming I need to look for:

22AWG solid core copper wire, 300V, insulating material non-crucial (Teflon seems a bit more challenging to find than PVC). Would that do it?

 

[Paul Birkeland]

You could contact replacementparts(at)bottlehead(dot)com if you'd like to order more of the wire we use.  I'm not sure that it's available by the foot from any vendor.  Otherwise any 20-22AWG solid core (never stranded) wire that's tinned under the jacket should work OK.

 

[Bingtou]

Thanks.

 

[Bingtou]

Okay, after trying to remedy my build and testing again, I've got exactly the same incorrect results as before. The red wire from D1 rises above 300V then drops down to 145V. The D6 wire reads normally, over 300V then dropping to 222V. (Also retested Kreg voltages, again around 4.7 on the A side, and 1.87 on the B side. On A, it takes quite a time to rise all the way to 4.7, over a minute I think, while on B, it reaches 1.87 quickly and stays very steady on exactly that value.)

The 4 LEDs light, and the tube appears to be glowing normally.

Here's what I've done in the last couple of days before this retest:

Removed the shunt regulator board and inspected the underside for poor soldering or short circuits. Also checked any soldering on the top of the board.

Checked each component on the board to make sure the right components were installed. Checked all resistor values with the multimeter.

Unsoldered the shunt regulator board and started again with the wire connections - from the PC power supply board, and for each terminal pin on the D terminal.


I did not resolder at the C, B, or A terminals, but I did visually reinspect them and everything seems to be where it should be and correctly soldered.

I replaced the shielded twisted pair from the power supply board to D terminal, in case the original one was damaged when I stripped the ends.

Mainly what I've tried to do is go step by step from the power supply board to the shunt board to the D terminal, to make sure that all solder connections are sound, all wires and resistors are in the right place, and there are no short circuits.

What should I be looking at next as a possible cause of this problem? Perhaps I overheated a transistor on the shunt board?

 

[Paul Birkeland]

There's a solder pad by BA and BB marked with an X, what voltages do you see at each of those?

 

[Bingtou]

The X pad reads 2.72V on the B half of the board, and 2.48 on the A half. The multimeter settled rock steady on those two values.

 

[Paul Birkeland]

Would it be possible to see some bottom photos of that shunt regulator board? 

 

[Paul Birkeland]

Another thing you might also try is wiring D2 to D7.  If one of the resistors that is responsible for grounding the grids isn't doing its job, tying them together will ensure that neither grid can drift up past ground potential.

 

[Bingtou]

Another thing you might also try is wiring D2 to D7.  If one of the resistors that is responsible for grounding the grids isn't doing its job, tying them together will ensure that neither grid can drift up past ground potential.

I take it you mean just running a wire between those two terminals and soldering them in? And doesn't need to be insulated?

Would it be possible to see some bottom photos of that shunt regulator board?

Damn. Wish I'd taken the shot when I had the board off. Sure, I can get a photo, but I'll need to remove a few wires and take the board off again. It might be a week or so before I can get to that.

In the meantime, I really appreciate the time and trouble you've put in to help out with this.

 

[Paul Birkeland]

Yes, just connect D2 and D7 with a piece of wire.