AB763 Showman - 700hz buzzzzzzz

[fdesalvo]

Hi, all.


Have a freshly recapped 60's Showman with new tubes that's evading my skills.  There's a 700hz buzz that's able to be modulated by the volumes, but is still present with them off.  There is no leaking in any of the coupling caps, nor is there stray dc on the circuit board.  Wire layout is per factory with shielded cable on the input grids (pin 2 of v1/2).  All buzzing ceases with PI pulled.  Disconnecting the PI input cap does nothing to abate buzz.    I converted this amp (for a customer) back to use with 6l6's, by installing a hammond 290fx replacement PT and choke.  It had a drake PT and choke pushing a quad of KT88s.  Had to make an artificial CT for the heaters and I grounded the orange internal shield for the PT along with the HT CT.  This amp is giving me fits I tell you! 

[pdf64]

Try disconnecting the feedback, eg lifting a leg of the 820R resistor - any effect on the issue?
What’s the V DC on the 100R feedback resistor, at the cathode end if the 22k tail resistor of the LTP, and at the LTP cathode.

[fdesalvo]

Try disconnecting the feedback, eg lifting a leg of the 820R resistor - any effect on the issue?
What’s the V DC on the 100R feedback resistor, at the cathode end if the 22k tail resistor of the LTP, and at the LTP cathode.

Noise increases with pulled FB resistor. 


Cathode end of 22K tail - 92.7vdc
100r - .3vdc
LTP cathode - 95.5vdc


TY!

[fdesalvo]

While looking over the PI circuit, I noticed the prior tech had a 100R resistor in place of the 470R on the PI cathodes.  After correcting that the voltages settled to where they should be +/- 10%.  The pervasive and annoying buzz persists.  I've also removed all pots and jacks and ensured solid path to ground via the brass plate.  Also clipped in another cap on the PI B+ tap and nada! 

[SILVERGUN]

Replaced PI tube?

[fdesalvo]

Yes. All known good tubes.

[HotBluePlates]

... buzz that's able to be modulated by the volumes, but is still present with them off. ...

This means "multiple buzzes."  Keep Volume controls all the way off while troubleshooting buzz later in the circuit.

... All buzzing ceases with PI pulled.  Disconnecting the PI input cap does nothing to abate buzz.  ...

Do you have any clip-on jumper wires?  Does the buzz stop when clipping the jumper to each side of the phase inverter's 1MΩ grid resistors?  That is, when you short-circuit those?



Separately "700Hz" kind of sounds like a choke and a capacitance oscillating.  Which then makes me wonder about the new Hammond choke.  What happens when you disconnect it, and install a resistor?  Or tack in the old choke?

[fdesalvo]

... buzz that's able to be modulated by the volumes, but is still present with them off. ...

This means "multiple buzzes."  Keep Volume controls all the way off while troubleshooting buzz later in the circuit.

... All buzzing ceases with PI pulled.  Disconnecting the PI input cap does nothing to abate buzz.  ...

Do you have any clip-on jumper wires?  Does the buzz stop when clipping the jumper to each side of the phase inverter's 1MΩ grid resistors?  That is, when you short-circuit those?



Separately "700Hz" kind sounds like a choke and a capacitance oscillating.  Which then makes me wonder about the new Hammond choke.  What happens when you disconnect it, and install a resistor?  Or tack in the old choke?

Hey, long time!  Hope all's well.  There was silence when jumping the two 1M PI resistors.  No change when subbing in a 1K resistor in the choke position.  TY for trying.

[fdesalvo]

At my wits end.  Open to any bizarre suggestions. I made another pair of shielded cables for the tone stack outputs.  No change. Can’t see what I’m missing.

[SILVERGUN]

Open to any bizarre suggestions

Maybe it's just you  https://www.quora.com/Do-enlightened-people-vibrate-at-a-frequency-of-700-Hz

[fdesalvo]

 

[HotBluePlates]

...  All buzzing ceases with PI pulled.  Disconnecting the PI input cap does nothing to abate buzz.  ...

...  There was silence when jumping the two 1M PI resistors. ...

When you remove the PI tube, the buzz stops.  And when you kill signal-input to the PI tube, the buzz stops.  I interpret these to mean the buzz is not due to the power supply, and not due to the tube itself.  This leaves the buzz as being a signal applied to the PI tube.

But when you unsoldered the PI cap the noise continued.  This may mean the PI tube is not receiving a 700Hz signal at its main input grid, but perhaps through the feedback loop (to the other grid) and/or its cathode.

When you disconnected the feedback loop, you noted the noise got louder.  The power section would be more-sensitive without the feedback, so this is an expected result.  Could this mean "noise applied to grid & cathode?"

What happens when unsoldering the 500pF cap AND the 820Ω feedback resistor?  We would hope the noise is stopped.

If it isn't, I notice the B+ lines emerge from the doghouse very near the PI components.  It sure would help to use a scope to verify no-700Hz on those (they shouldn't have any because of those new filter caps).  I would also be tempted to tack-in the old output transformer & rule out some oddity with the new part.

[fdesalvo]

HBP: thanks for the analysis. Great bit of detective work to kickstart my imagination. I’ll try totally isolating the inputs of the pi and will report back on my findings. Can you tell me what those 1m resistors are doing at the pi?  I can’t discern their functions. Fwiw, the ot is the original unit and wasn’t changed during the kt88 conversion or restoration back to 6l6s.

[sluckey]

Here's some short, easy reading about that PI.

     http://www.aikenamps.com/index.php/the-long-tail-pair

     http://www.valvewizard.co.uk/acltp.html

[HotBluePlates]

... Fwiw, the ot is the original unit and wasn’t changed during the kt88 conversion or restoration back to 6l6s.

I'm sorry!  I misread your first post, which noted a new power transformer.

... Can you tell me what those 1m resistors are doing at the pi?  I can’t discern their functions. ...

Sluckey posted good links.  But quickly:

     The 470Ω resistor is a cathode bias resistor.  Pretend this is a "normal stage" and that resistor connects to ground.

     Every tube stage needs a reference for its grid.  A "normal stage" would have a resistor running from grid to ground.

     The 1MΩ resistor is the grid reference; it runs from grid to the "ground side" of the 470Ω resistor.

     If you want to measure the bias voltage of the long-tail, it is best to measure the volts at the "ground side" of the 470Ω, and then measure volts again at the "cathode side" of the 470Ω.

     One side of the long-tail is the "amp input" while the other side is the "feedback input" via the coupling cap from the feedback resistors to the grid.

[fdesalvo]

Thank you so much!  Merry Christmas to all, by the way. Very informative.


Ok so after disconnecting the 500pf input and feedback resistors, the noise persists, but even louder.

Ground side of the 470R- 93.4vdc.  Cathode side is 95.9vdc.

Curiously, with either volume control is turned up from zero to 2, the noise decreased greatly. Then once the volumes are brought up past 2, the noise increases along with the vol pots. Here’s what my app is capturing -

[fdesalvo]

Is it possible for the trem intensity circuit to introduce noise downstream of the pi input cap?

[sluckey]

Is it possible for the trem intensity circuit to introduce noise downstream of the pi input cap?

Tremolo was eliminated when you disconnected the 500pF input cap.

[fdesalvo]

Doh!     How about noisy 220k bias feed or those 1.5k grid resistors?  My cartoon brain is out of ideas. Can’t imagine anything else between that input cap and the 6l6 grids.


In all fairness, the chassis looks like Keith Richard’s lungs. The pot shells are still rather nasty and I removed a ton of old flux and smoke schmutz from the sockets, which are spotless now. The insides of the pots should be spotless, but who can say.

[PRR]

Why 700Hz???? That's not a normal power frequency. It seems very odd for a Fender's internal howl since the TMB tonestack dips 700Hz. I remember CRT computer monitors singing 90Hz, and wi-fi erratic beat....

[fdesalvo]

I wish I knew why. Can’t even have normal malfunctions at the house 

[acheld]

I've seen some weird noise from LEDs, and wall switches controlling LED lights.   Just a thought.  700Hz is a very strange signal.   I suppose an oscillation could produce it, though I've never seen it.

[tdvt]

I've seen some weird noise from LEDs, and wall switches controlling LED lights.   Just a thought.  700Hz is a very strange signal.   I suppose an oscillation could produce it, though I've never seen it.

Probably not the issue, but I was working on my Princeton Reverb last month, chasing all kinds of noises as it hadn't really been used much, much less serviced, over the last 20 years.

In several instances, just when I thought it was dead silent, I would get an intermittent noise & mid-frequency hum. So I would go back to looking for more microphonic wires, components, board leaks, etc.
Felt pretty clueless when I realized it was only happening when my cell phone was on the bench.....

[fdesalvo]

Thanks for weighing in gang. Yah, I’ve become very familiar with the sound of cellular noise lol. I keep my phone away from the bench while debugging.

[bbmade]

At my wits end.  Open to any bizarre suggestions. I made another pair of shielded cables for the tone stack outputs.  No change. Can’t see what I’m missing.

I’ve found myself in a couple of situations that seemed impossible to track and solve.

One time Sluckey saved me noticing a connection in a picture I uploaded.

Now, when I find myself in a situation like that I look at every component starting at the input and work my way through.  I verify the component value and also the correct location.  Anything that has been “modded” gets documented.

I find Microsoft One Note with a tablet is the best for this.  I can write with a stylus just like it’s a piece of paper.  I can use different colors to tracing things, highlighting things etc.  I also document voltages as I go too.

The Twin Reverb I was working on had been modded and there was a resistor hooked up to the wrong eyelet.  Not sure if I did that or if it was done in the original mod and then my making changes to their mods exposed it.  Either way, Sluckey saved my a*$.

[fdesalvo]

Yah brother. I think SLuckey has save quite a few keisters around here - mine included lol.