Reverb makes a fog horn

[G._Hoffman]

Some medical issues have kept me from working on amps for a while (who knew you actually needed to sleep when you were in bed to be productive?), but I'm getting around to finishing up some old projects, and I'm just finishing up the electronics for a Weber 6A40 kit (their Super Reverb kit), with some construction improvements - following Doug's grounding plan, some better tube sockets, etc - and reverting a couple changes they made to an AB763 schematic (why would you use a normal pot for adjusting bias?).  But when I turn up the reverb, as soon as it gets to 8, it makes a fog horn sound.  Any hints on what could be causing it?  I needed to make some tweaks to the tremolo, but other than those two things, it sounds killer through my bench test speaker.  Really looking forward to getting it finished for real.

[separateness]

My first suspicion would be the 220k grid resistor after the Reverb pot, then the 3.3M/10p mixing network.

[G._Hoffman]

My first suspicion would be the 220k grid resistor after the Reverb pot, then the 3.3M/10p mixing network.

I'll have to wait until tomorrow when everyone else isn't sleeping, but I'll give those a try.
It is the only pot I didn't replace in the kit (they really were cheap parts).

[separateness]

In that case I would also investigate that pot.

[TIMBO]

Try swopping the green and black wires on the reverb driver transformer.

[tubeswell]

Oscillation

[pdf64]

Yes, oscillation, maybe due to a bad decoupling cap or cap connection at V4 cathode or the D HT node.

[G._Hoffman]

Yes, oscillation, maybe due to a bad decoupling cap or cap connection at V4 cathode or the D HT node.

Connections all look solid, and the caps all read good on the ESR meter, and within spec.
Honestly, though, just playing through this thing, even through a really wrong speaker (200W Bag End - very clean and efficient), I do get the hype for the AB763 circuit.  It sounds really good.  I gotta get the cabinet made so I can put it through the 4 Weber ALNICO 10" I have for it.  (Is it just me, or is it weird that they make pretty nice speakers, but their kits use pretty junky parts?  Some kinda disconnect there.)

[G._Hoffman]

My first suspicion would be the 220k grid resistor after the Reverb pot, then the 3.3M/10p mixing network.

Those all seem fine, from what my meters can tell me.  That 10pF is beyond the sensitivity of any of my equipment (I mean, I've got a capacitance meter which SAYS it can read it, but moving my hand near the meter makes it change by almost 8pF), and I don't have any replacements atm. 

[separateness]

How mechanically isolated are the speaker and the reverb pan in your testing conditions? I presume it doesn't happen with the pan disconnected? What if you unplug the input to the reverb pan?

[G._Hoffman]

Try swopping the green and black wires on the reverb driver transformer.

Did not change the behavior. 

[G._Hoffman]

How mechanically isolated are the speaker and the reverb pan in your testing conditions?

Not great.  The pan is just sitting on the bench, and the speaker is under the bench.

I presume it doesn't happen with the pan disconnected? What if you unplug the input to the reverb pan?

No change, whether either or both tank connections are disconnected.

[G._Hoffman]

I did also try swapping tubes around, which didn't help.

[tubeswell]

Gutshot?

[shooter]

looking at this for reference;
https://el34world.com/charts/Schematics/files/Fender/Fender_65_super_reverb.pdf


stabbing at dark
disconnect the NFB n re-try


what little logic I used, if "V4" is like this schematic with a shared common cathode R of the recovery side reverb, things might be sneaking around the coupling caps ??

[G._Hoffman]

Gutshot?

Not super proud of the spaghetti.  I was so low energy when I started on this (sleep disorders suck), I just didn't have it in me to do it "right".  And those giant copper ground busses were a PITA, and not something I'll repeat, though they do seem to work well.

[tubeswell]

I’d start by shortening as many of the wires as possible. Eliminate long parallel runs of wires (unless they are equal signal amplitude in 180 degree opposite phase).