Need heip with peculiar AB763 issue

[markmalin]

I'm building one of my Bad Dog amplifiers (see schematic below).  It's a single channel AB763 Vibrolux w/out vibrato, and with a Dumble ODS style overdrive.  Originally when I first powered it up, it was drawing too much current - turns out I had not cut the leads off a couple resistors, and the "C" "rail" in the power supply dropping string was shorting to the Bias point between the 220K resistors on the power tube grids.     I corrected that, fired it up and it plays.  Here are the voltages - everything seems to look normal to me, but I'd appreciate your input - but please see the issue below - that's what's driving me nuts.  Ok, voltages:


V1a  193v
V1b  197v
V2a  172v (probably low for a Dumble OD)
V2b  183v (ditto)
V3a  411v
V3b  411v
V4a  184v
V4b  193v
PIa   192v
PIb   199v
KT66 1  415
KT66 2  415


Power supply voltages:
A 417v
B 415v
C 333v
D 282v


bias -47.6
I can only get the power tubes to bias at 28ma, but I need to change the 22k resistor to ground on the bias pot - but I don't think that's an issue.


So here's my issue.  Amp sounds fine, but if I play a note HARD with the volume up past 2, I get a nasty distortion.  So I tested it with the Master Volume down and the channel volume at 10.  Nasty distortion as the note dies out if I play a note hard.


I stuck a chopstick in there and if I poked the wire on the input to V1a (the coax going from the two 68k resistors at the input jack to V1 pin 2) the nasty distortion was worse.  I figured "fine, the cable's bad or the resistors are flakey".  I replaced both 68k resistors and re-did the coax.  Much to my frustration...SAME issue.


I am attaching links to two videos.  In these videos I have the Volume on 10, Treble at 7, mid at 10, bass at 3, and Master Volume about 1 or 2.  OD switched OUT, OD trim off, OD drive and level controls off, reverb off.  I pluck a string HARD on my les paul and if I poke around on this coax, the nasty distortion comes and goes.  Since I already replaced the 68k's and the coax, what the heck is going on?  Can this possibly be parasitic?  I'm completely stumped.

This one, the volume is about 3 or 4, master about 4 or 4
https://youtu.be/2YtHJIBdE74
this one the Volume is at 10, and Master at 1 or 1
https://youtu.be/ebl4nLPZ5-Q


Really need some help.
Thanks, guys!


Humbly,
Mark

[markmalin]

here's a pdf of the schematic


Other stuff I've tried
 - I swapped out all the tubes with a known good set from another amp
 - I tried a different speaker

[markmalin]

and a PNG - sorry for the multiple replies, just want to be sure people can read this

[ac427v]

I see that V4a is in series with V4b and they share a cathode resistor. I have always separated the cathodes in that scenario based on previous discussions here. Could that be causing problems?
Craig

[markmalin]

I see that V4a is in series with V4b and they share a cathode resistor. I have always separated the cathodes in that scenario based on previous discussions here. Could that be causing problems?
Craig

I know Gerald Weber always said to separate them for more control.  But why would that effect the input to V1a?  If I push on that coax it will go in and out of the distortion problem.


Mark

[sluckey]

What happens if you replace the coax with a simple wire?

[markmalin]

What happens if you replace the coax with a simple wire?

I was thinking of that...  I know the original Fenders didn't use coax.  I'll try that this evening.  (this is one of those days I wish I could call in sick to work and figure the blasted amp out!)

[macula56]

What does your NFB connect to? Doesn't that usually connect to the 22K/100ohm junction on the PI?

[markmalin]

What does your NFB connect to? Doesn't that usually connect to the 22K/100ohm junction on the PI?

That's a mistake in the schematic - I forgot to draw in that node to the 100 ohm resistor on the PI.  It's there on the amp, though.

[macula56]

Cool. Nice schematic. Love to hear it when you get it fixed.

[markmalin]

What happens if you replace the coax with a simple wire?

yup.  That did it, sluckey.  My buddy Bob G explained to me the details of shielded wire causing a ground loop, since the jacks were not isolated.  Man, lesson learned.  Stick to the way Leo did it.


Thanks guys!  And Thanks Bob :)


Mark

[robrob]

Mark, congrats on finding and fixing your problem. Could you give us the details on how the shielded wire was connected and grounded (did you ground both ends of the wire)?

[markmalin]

Mark, congrats on finding and fixing your problem. Could you give us the details on how the shielded wire was connected and grounded (did you ground both ends of the wire)?

Thanks.  I guess I'm so used to building dumble clones that I figured, use shielded wire for V1a and V1b grid inputs, but what I missed was when I build those I use isolated Cliff jacks.  When I use shielded wire I always only terminate the shield at one end...so there's no ground loop...and in this case, at the ground buss.  After submitting this thread I also ping'd my buddy Bob, a retired EE with years of amp repair experience and sent him a link to the thread.  His reply was "sluckey nailed it", but he also explained why.  Here's a clip from his email.  I find this fascinating, and also make myself a little vulnerable here because as many amps as I've built, I am still learning :)


"You've got a ground loop created by the grounded INPUT jack to the chassis (metal washer, right?) and the shielded ground of your coax. Your coax shield is not at the same ground reference as your guitar signal. Pushing the coax against the chassis seems to clear things up. By doing this, you're creating a teeny "capacitance" that is filtering the distortion. This clearly indicates that there's a difference of potential (i.e ground loop) between the chassis ground plane and the shielded braid.  If you lift the coax shield, the distortion should not occur. If you insist on using coax, best place to ground the shield is at the cathode resistor of V1a, not to the chassis. But, you'll need anodized washers to isolate the INPUT jacks from the chassis.
 
On my blackface Twin Reverb, I used the Fender scheme of twisting the red and yellow solid core pushback wires from the INPUT jack and the BRIGHT switch. All INPUT jacks are grounded to the chassis with metal washers. Works for me :-)."


Humbly (and a little more humble today)
Mark

[Willabe]

When I use shielded wire I always only terminate the shield at one end...so there's no ground loop...and in this case, at the ground buss.

From your friend;

"You've got a ground loop created by the grounded INPUT jack to the chassis (metal washer, right?) and the shielded ground of your coax. Your coax shield is not at the same ground reference as your guitar signal."

"If you insist on using coax, best place to ground the shield is at the cathode resistor of V1a, not to the chassis. But, you'll need anodized washers to isolate the INPUT jacks from the chassis."

Was your ground buss wired to the input jacks ground lug? If yes, did you also ground the buss wire to the chassis at another point/place too?

If your ground buss was wired to the input jacks ground lug and you soldered the coax's shield to the ground buss wire, it would be very close to the jacks ground lug, like 1/2" or less? That shouldn't cause a ground loop.

The preamp ground buss wire, if soldered directly to the input jack ground lug, along with the coax shield will all be grounded together at 1 point to the chassis through the jacks metal frame/threaded sleeve. So no ground loop. 

Either way, they would have to be at the same ground potential. As long as you didn't ground the buss to the chassis at a second place.

[gtrplayr1976]

I have a similar issue with a Blues Junior I'm working on. Any idea what could cause that similar noise ?

[markmalin]

Was your ground buss wired to the input jacks ground lug? If yes, did you also ground the buss wire to the chassis at another point/place too?

If your ground buss was wired to the input jacks ground lug and you soldered the coax's shield to the ground buss wire, it would be very close to the jacks ground lug, like 1/2" or less? That shouldn't cause a ground loop.

The preamp ground buss wire, if soldered directly to the input jack ground lug, along with the coax shield will all be grounded together at 1 point to the chassis through the jacks metal frame/threaded sleeve. So no ground loop. 

Either way, they would have to be at the same ground potential. As long as you didn't ground the buss to the chassis at a second place.

The ground buss is tied to a single point near the input jacks.  The jacks are attached the the chassis, so not isolated, but also have a ground wire that's tied to the same lug (point) the ground buss is attached to the chassis with.  The coax shield was wired to the ground wire on the input jack


hope that made sense

[robrob]

"The jacks are attached the the chassis, so not isolated, but also have a ground wire that's tied to the same lug (point) the ground buss is attached to the chassis with."

There's your ground loop. That will literally act as a loop antenna. Either isolate the jacks or remove the ground wire.

[markmalin]

"The jacks are attached the the chassis, so not isolated, but also have a ground wire that's tied to the same lug (point) the ground buss is attached to the chassis with."

There's your ground loop. That will literally act as a loop antenna. Either isolate the jacks or remove the ground wire.

Yah, I've wondered about that....