Shunt regulator readings [resolved]

[Bingtou]

I've connected D7 and D2 to each other by wire, rerun the test, and I still get 144 VDC on the D2 wire.

I can get the shunt board off and photograph the underside for you, but if it's what it takes to get rid of this issue, I'm about ready to order a new board and put all new components onto it.

 

[Paul Birkeland]

Putting a new board with new components on it may or may not resolve your issue.  If you can take the board off and photograph the bottom, then also photograph the D socket, that could be helpful.

 

[Bingtou]

Putting a new board with new components on it may or may not resolve your issue.  If you can take the board off and photograph the bottom, then also photograph the D socket, that could be helpful.

If I can do that before the weekend runs out, I will. Don't laugh at my soldering. I'm also going to retest the power supply board, just to confirm that I have the right readings from there.

I will just mention that, as I said several posts ago, I resoldered the D socket connections, including reinstalling the resistors, to see if it helped. The only one that I left in place at that time was the STP from the C socket.

I have looked at the C socket itself and all connections there seem to be as specified. I will put up a photo of that as well. Also, just to be sure, I disconnected the black wires at the T23 end, trimmed and restripped the ends, and soldered them back in place.

This process has been a little frustrating, because I don't feel great about heating some of the connection points or component leads multiple times or for too long. Also, I have broken connections sometimes when bending wires out of the way (broken off the end of the wire) - where practical, I have been disconnecting them in preference to bending them.

 

[Bingtou]

Also, I assume this is nothing, but could you take a look at the 2N2907A transistors? They're visible in the "shunt4.jpg" photo I posted on June 16th. The case of the transistor on the left (A-side) is pretty corroded. I'm assuming it's not a problem - can you confirm that?
(The thing that looks like a bead of moisture was just some dried adhesive or something - I've already removed it.)

 

[Paul Birkeland]

Some corrosion on the metal case of a 2N2907 isn't particularly concerning. 

 

[Bingtou]

Hi Paul,

Here are some photos of the underside of the board. This is the best I can do without taking the whole thing off. I'm nervous both about yanking the board around while in place, and about soldering and resoldering the same locations. Things are getting a bit more raggedy than when I first finished soldering up this board.

These are all with the B side of the board on the left of the picture. There's a detail of the far corner because I didn't think the original shots showed it clearly enough.

I also included a shot of the C socket.

 

[Paul Birkeland]

Can you get one in focus of the D socket itself? 

Where did the white wire come from, and what's it being substituted for?

 

[Bingtou]

Where did the white wire come from

Sorry, I had intended to mention that yesterday, and it slipped my mind. At the risk of answering too literally, it came from a Crack kit, left over from a build I did a few years ago. The white wire was handy as a contrast to the red and black wires that the Eros uses.

The terminal was getting crowded and difficult to work with, and more than once now as I've spent more and more time on redoing bits on that board or parts of that socket section, I've nicked things with the soldering iron. No damage beyond cosmetic, but it feels bad and it looks bad. So to run the wire from D2 to D7 as you suggested, I attached it round the outside of the socket instead of crossing through it. It's not that I'm not going to know which wire I used, but the white makes it conspicuous as I figured I'd be removing it later if it didn't test out. And as I mentioned a couple of posts back, unfortunately the tests were the same as the previous results.

I'll try and get photos of the D socket up tonight.

 

[Bingtou]

Here are some photos of the D socket. I've got a feeling you might be going to ask about the 4 and 5 pin, which may not be clearly visible in the pictures. I did check that connection when I first tested this section. It looked good. However, I also undid it and resoldered - I think I even cut new wire for that part. From having looked at those pins many times, and the wires and the solder joints, I believe those connections are not at fault.

These shots are from various angles, as you can see from the numbering, but I couldn't get around by D4/D5 while the shunt board is still attached.

 

[Paul Birkeland]

Ah, I see now that it's your D2 to D7 jumper.

I know this isn't the most helpful suggestion, but honestly I would go back over each individual step in building and installing that PC board.  Most of the resistors on it can be measured after they are installed, so you can check them.  Also pay very, very close attention to where the wires go from each individual solder pad on that PC board down to the D socket.  You are correct that there are many wires of the same color, and we have added blue and white wires to the Eros 2 (not yet available for tape) to help with this situation.

If all else fails, I would suggest sending it in for repair.

-PB

 

[Bingtou]

I know this isn't the most helpful suggestion, but honestly I would go back over each individual step in building and installing that PC board.

Actually, it probably is the most helpful suggestion. But I want to make sure I understand correctly - go over each step means to check, or to physically rebuild? If you mean checking each component, in order of installation, I will do that first.

And if that doesn't go anywhere, I am actually ready to rebuild this section from scratch, assuming it wouldn't cost a fortune. I know you said that a new board wouldn't necessarily solve the problem, but what about a new board and replacing most of the components? I assume that any resistors which check out can be considered undamaged, but I did wonder about transistors. (And I don't know how to check them either.)

If, for example, I can just replace some or all of those components for the shunt regulator and install them on a new board, using new wire for all the connections, I'd rather spend a little money (within reason) on that, because as long as the power supply section shows no problems when I retest it, the rebuild of the board with new components ought to resolve the issue. Can I safely assume it's nothing to do with the tube? Or the socket?

I'm far from considering sending it in, because it's not even one-third built yet, so there aren't that many steps to retrace. Concerning new components, if I do that it would be more convenient for me to put together an order through Mouser or Digikey in Japan - are there any particular pitfalls versus getting replacement parts from Bottlehead?

 

[Paul Birkeland]

I would recheck all of your building steps, not actually rebuild everything from scratch. 

While there is a 1 in a million chance that you have a bad component, the more likely scenario is that there is a bad solder joint or a mistake in the build. 

 

[Bingtou]

After several months of delay - various reasons, including some other projects on the go - I got back to the Eros and rebuilt the shunt regulator board, with mostly (actually, I think all) new components. Another few weeks of delay followed that, and I ran the test again. This time it all checks out on both sides, rising over 300 V and then dropping to exactly 200V. A big relief.

Thanks so much again Paul for your assistance. Replacing the parts may not have been the ideal solution from a learning point of view, but I wanted to move this build along. I'm not sure how soon I will complete it, but I can at least move to the next stage now.