Sounds like...?

[jordan86]

I have this weird intermittent issue with my AA1164. I wonder if some more experienced ears might have any insight as to what I can chase.

It’s a bad crackle that only seems to go away when I move the rectifier tube. Although most times it goes away by itself. It’s not affected by the AC signal. Like I can play over it and it stays the same. Adjusting the volume pot makes no difference either. Cranking the reverb pot makes it louder, and cranking the tremolo depth does cause it to fluctuate. I imagine it’s maybe something in the seating of the rectifier. At the end I reach back and move the rectifier and it stops.

[HotBluePlates]

... It’s a bad crackle ... I imagine it’s maybe something in the seating of the rectifier. ...

Ever heard someone weld before?  As in "arc-welding"?

Your noise kind sounds like arcing.  I think you already know your rectifier's socket doesn't grab the pins firmly.  And if those contacts are only barely-touching or just-apart, there could be arcing from the socket-pin to the rectifier-pin.

If you tell us the rectifier is a JJ GZ34, well...  JJ had a bunch of issues with too-skinny pins on octal tubes like a decade ago.

[jordan86]

It’s been a while since I took shop class HBP, but I do hear what you’re saying.

It does seem to happen more with the JJ’s and less with my Chinese 5AR4. The Chinese one is a tubestore preferred series that I got on sale. It’s their sort of premium quality hand picked line. Although, I think that clip is with a NOS Sylvania 5Y3.

So my solution is re-tensioning pins or replacing the socket? Would installing back up diodes on the rectifier socket matter? (I assume not)

[jordan86]

I guess it’s also worth asking...if I replace the socket, are backup diodes a worthwhile addition when doing the install?

[HotBluePlates]

So my solution is re-tensioning pins or replacing the socket? ...

Either, depending on wallet, interest & how loose the socket pins are.  I infer that you are confirming my guess that the socket is indeed loosely holding the tube.

...if I replace the socket, are backup diodes a worthwhile addition when doing the install?

Backup diodes are always good insurance against a rectifier tube shorting out.

They won't fix the noise here, but they will keep a filter cap from exploding the next time your rectifier tube short-circuits.

[jordan86]

Would this arcing be visible if I dropped the chassis to look at the socket on the inside?

[shooter]

sometimes, usually there will be a burn trail if it's a pin to pin, but it might only be a hair thick, so good eyes or magnification.
if it's socket to pin, could be harder to see.
put it in a completely blacked out room, fire it up, get it to "act up" while watching real close, the problem with real close, you might get a poke cuz your nose is closer than your eyes 


look inside the tube also, a mechanical fault/crack there can do the same

[jordan86]

Pulled the chassis and powered the amp up in my windowless walk in closet. Pitch black. No visual signs of arcing in the tube or at the pins inside the chassis. The fact that it happens with several rectifiers of different types and brands makes me think it is not likely the tube.

Definitely affected by the seating of the rectifier though. I can make it start and stop with the slightest angle change.

I checked the socket pins. Are these type able to be re-tensioned? I’ve only ever tweaked the old fender types with the two sort of “sleeves” that grab the pin.

[SILVERGUN]

I would replace that socket. Those pins are very easy to break because they have a section in the middle that you can't see that is only about 2-3mm of contact between what you see from the top and what you see from the bottom. Take a close look at pins #5 and 7 in your pic.
I broke one once and couldn't believe what I saw.
Just my 2 cents.
Also, squeeze that bear trap so it grips better when you put it back on.

[PRR]

Is the wire on pin 8 actually wet with solder? Or just touching lightly?

I can't tell. And when I can't tell, I reject it back to the assembler for complete solder removal, cleaning/scraping, flux, and solder under bright light.

[jordan86]

Also, squeeze that bear trap so it grips better when you put it back on.

Thanks, Silvergun. I'll try to tighten the bear trap in the interim. So would this be a better socket if I replace it and add backup diodes?
https://www.tubesandmore.com/products/socket-8-pin-octal-saddle-plate-black-bottom-mount

Is the wire on pin 8 actually wet with solder? Or just touching lightly?

Pin 8 is definitely soldered. It seemed a bit suspect bc it does look thin from the other side, but I gave it a firm tug and it seems well connected. It probably wouldn't hurt to go back through and reflow each pin before I completely abandon the socket.

[jordan86]

https://www.tubesandmore.com/products/diode-general-purpose-rectifier-1a-1n400x

Any reason to NOT just get the 1000V diodes, 1N4007? vs the 600V or 800V? I assume 600V a little too underspec'd to be safe.

[SILVERGUN]

Also, squeeze that bear trap so it grips better when you put it back on.

Thanks, Silvergun. I'll try to tighten the bear trap in the interim. So would this be a better socket if I replace it and add backup diodes?
https://www.tubesandmore.com/products/socket-8-pin-octal-saddle-plate-black-bottom-mount

Looks good to me

https://www.tubesandmore.com/products/diode-general-purpose-rectifier-1a-1n400x

Any reason to NOT just get the 1000V diodes, 1N4007? vs the 600V or 800V? I assume 600V a little too underspec'd to be safe.

Nope, buy em in bulk if you plan on sticking around

[HotBluePlates]

... So would this be a better socket if I replace it and add backup diodes? ...

You have not said the tube is loose in the socket.  Is the tube loose in the socket?

If no, then resolver all connections on the socket.  There's no point replacing the socket if it's holding the tube pins fine.  Arcing can also happen from pin to wire if it's not well-soldered at Pin 8.

[jordan86]

... So would this be a better socket if I replace it and add backup diodes? ...

You have not said the tube is loose in the socket.  Is the tube loose in the socket?

If no, then resolver all connections on the socket.  There's no point replacing the socket if it's holding the tube pins fine.  Arcing can also happen from pin to wire if it's not well-soldered at Pin 8.

I had this same thought after posting. If the socket is sloppy, then reflowing solder solves nothing. The socket is sloppy enough that I can move the tube a bit but it doesn’t move on it’s own. I will try tightening the shark jaw as well.

[HotBluePlates]

... The socket is sloppy enough that I can move the tube a bit but it doesn’t move on it’s own. ...

The tube being able to be moved by a hand is normal.  If the tube comes out of the socket with almost no effort, that's abnormal.

[PRR]

Any reason to NOT just get the 1000V diodes, 1N4007? vs the 600V or 800V? I assume 600V a little too underspec'd to be safe.

The diode has to stand twice the no-load B+. On smaller amps this tends to be 455V (brushing the 450V limit of e-caps), so you need a 910V part. Even if the cruising voltage is 350V.

If at all possible I would double that for line spikes. I have replaced a lot of 100V diodes in 72V (36V*2) applications. I had a microwave oven half-crippled by nearby lightning. But I guess you might not play through all lighting storms, while we left that over plugged 24/7.

Back in 1970 the 1N4006 was MUCH more expensive than 1N4001 (and there was no 1N4007). Today in lots of 10 or 100 they are all the same price so just get a bag of the '007 part.