You are running the bias off the regulated B+ rail so regulating it again doesn't really make sense.    Low drift resistors and you are golden. 

So I was fiddling with some things on mine today, putting in a new potentiometer, shortening a few cables, sleeving some of the wire runs with fiberglass to clean up the guts and as practice for cable management on the circlotron build. But when I tested the power supply before hooking everything back up, I got gremlins in that same damn +500V side again. No pops or sparks but it's outputting 182 VDC/402VAC at the +500 terminal. The trace tapped from the positive end of the HV rectifier bridge is 482VDC. These are new parts, but it should be the rectifier bridge since there shouldn't be any AC beyond that anyway right? Because the big caps behind that are for smoothing and the rest should be regulation?

 

I know the resistors are all good. I desoldered and checked them all on the +500 side last time I had a problem. Same with the axial diodes except just checked that they were still functional, and same with the caps, all take a charge. Terminal blocks all connect to traces and nothing in them connects to ground that isn't supposed to.

 

Anything else I might be missing?

 

I guess the lesson here is never continue to touch something when it was already working fine. I'm this close to just getting a new PSU board and transplanting the known good parts since all the soldering and desoldering especially on the sand locations is starting to make them a little rough, and that goes double for the solder pads that connect to traces on the top of the board.

For that amount of AC it has to be the rectifiers.  The soldering has to be pretty much perfect especially in the diff amp so you could be dealing with tin whiskers or something like that.  Also just a cold joint would cause this as a part has shifted when moving the board. 

1 hour ago, Tinkerer said:

Anything else I might be missing?

My experience with Phoenix connectors is that you can actually put in a wire, and tighten the screw with out getting a proper connection.
I usually check the connection from somewhere else on the boards than the actual connector.
Since you have been fiddeling with the wires I guess the Phoenix connector is the culprit here rather than something more serious.

I can second Birgir's and Soren's point based on my own experience - bad connection either caused by cold solder joints or connectors would be what I look into first. 

Add to the terminal block issue if they are fake ones off ebay.  Massive issues with that stuff. 

So this wound up being kind of funny. I checked everything, and I mean every part of that side. Used multimeter resistance mode on the transistors to double check them and compared values to the good side. Everything matched up. All the top trace stuff to ground went to ground. All terminals were fine with or without wires screwed in. The couple little bridge traces were fine. So left with nothing else, I went trace short hunting systematically, every point to every other point.

Then I finally found the culprit.

The last time I had replaced the 10m90s for that rail, the solder pad for one leg lifted up just a hair and I had pushed it back down and soldered everything up fine. But moving the board around must have snapped it somehow, or maybe it happened when I double checked the tightness of the screws on the heatsinks.

What it left me with was a shorted trace that you couldn't see at all under even magnification. Ridiculous. Took less than half a minute to remove the busted loose pad, score a bit of the trace down to copper, tin it, bend the cut lead over it and solder it properly.

Then everything fired up, just as it should.

Finished my HV its now all boxed up and running well. I used a Paul Holden transformer and its dead quiet, expensive but worth every penny.

Listened for hours and it ran just warm. Thanks to all for a utterly enjoyable Audio experience.  

Edited December 29, 201510 yr by iwik
Added proof of build

Well done Les, I agree Paul’s transformers are very stable and well made.

Before I get all carbonated, due to sudden parts saturation (thanks wink), I have another old school KGSSHV on the bench... it’s a beast compared to the mini I made a month or so ago, running a toasty 14mA.

 

 

 

Nice!

Sent from my SM-N9005 using Tapatalk

Ahhh an old school HV.    So much empty space on those PCB's... 

Thanks John

10 hours ago, johnwmclean said:

ell done Les, I agree Paul’s transformers are very stable and well made.

Before I get all carbonated, due to sudden parts saturation (thanks wink), I have another old school KGSSHV on the bench... it’s a beast compared to the mini I made a month or so ago, running a toasty 14mA.

 

are you using uisolated recifiers? 
and why no heatsinks on the 10m90's?

 

 

That board is 100% correct. 

those are the low voltage regulators. I still think that heatsinks were a good idea for those.

the 10m90s are in the middle and do have heatsinks

the rectifiers do look like they are as close as on the latest KGSShv PSU board ... why does it work in this situation?

You can put a nylon screw and three nuts through the tabs on the diodes to give them more structure.

1 hour ago, sorenb said:

the rectifiers do look like they are as close as on the latest KGSShv PSU board ... why does it work in this situation?

They are much closer on the newer boards as in they are touching.  For a normal KGSSHV there is no need for heatsinks on the LV stuff.  Doesn't generate enough heat for it. 

Hey John, what type of wires/interconnects do you use? They seem pretty robust  I'm always on the lookout for good wiring, presently I strip cores out of an Allen Bradley DeviceNet cable I took from a switchboard factory awhile back...

3 hours ago, Aive said:

presently I strip cores out of an Allen Bradley DeviceNet cable I took from a switchboard factory awhile back...

I had some quality Belden 19364 I was doing the same thing with 

The primary wiring is off-cuts from the transformer windings, I usually do this if I have enough, all the HV wiring is from Bulk Wire (600V Stranded UL 1015 20AWG), for the inputs I used Belden 83005 010100 22AWG silver plated copper, which is a nice high temp teflon wire and also rated 600V. The XLRs sockets are Neutrik DLX with gold contacts.

Thanks John - I love that teflon stuff, hides my terrible soldering skills

I waited 4 years for this. Completed the HV last year but waited another year before getting the can.

Heartfelt thanks to Kevin Gilmore, Spritzer (Birgir), Lil Knight, Victor Chew, Kerry, RiStaR, and the rest of Stax mafias. Thanks to KG for his design and guidance, Spritzer for helping me behind, Lil Knight for providing me the boards a long time ago, Kerry, RiStaR for their answers to my questions, and finally to local DIYers Mr. Victor Chew, thank you for teaching me, what you have taught me not only benefited me in making the amp, but also helped me in my job.

Next I will try to build the HV Carbon.

I have had caps and ICs exploding on me before when building the HV.

But this time, I plugged the 009 into the HV, and it exploded....

 

 

 

Into some amazing music.

 

Edited January 17, 201610 yr by kh90123

Congratulation on your build. You are lucky to have Victor in the same town. He also helped me a lot with my build. 

Congrats!  Enjoy

Actually I had one problem. I heard distortion coming out of it, then I bypass the volume pot and the distortion went away.

Seems like I might have a lemon volume pot. I will be getting a DAC with pre-amp out soon, and in the future I will go pot-less build, can keep cost lower by not needing any volume pot.

Perhaps it has got to do with how I wired up the pot. I am currently using SE input instead of balanced in. The pot I have was sold to me by someone, and from the way the resistance sweeps it really doesn't look like a log pot, it's a linear pot...

The 009 is rather revealing...it occurs to me that so many recordings has rather audible background hiss. And it also occurs to me that musicians also makes quite a bit of mistakes in recordings.

Edited January 17, 201610 yr by kh90123