Fender CBS Pro Reverb Master vol - :Normal channl reverb bleed thru

[Mars Amp Repair]

Hey folks,
Just adding to the library of weird issues. I have a 74 - 75 CBS SF Pro reverb Master vol. The customer was complaining that when using the normal channel, he hears reverb in it. I tested & for sure it does.
I tend to think this is not an anomaly with this particular unit, but how they all are.
I found that the bleed thru is finding it's way from normal channel to the Reverb driver tube grid via V2B cathode which is shared with V2b cathode resistor & bypass cap.
If this is ever an issue, you can remove the link between the 2 cathodes, add bypass components for V1B 1.5K and 22uf cap to the same ground as V1B cathode.
On the Vibrato cathode of V2B, change the cathode resistor from 820ohms to 1.5K as that stage is no longer sharing current with V2B cathode.
Hope this helps for anyone who actually had a cust complain about this.
Glen

[Chuck H]

I much prefer the aesthetic of separating the cathode circuits. That said...

The shared cathode is a common design element of ALL the Fender BF and SF designs that don't exhibit this problem. I think it's more likely that the shared cathode capacitor was failing/aging and demonstrating higher than ideal impedance. Just replacing that capacitor probably would have solved the matter. You mentioned replacing the resistors when you separated the cathode circuits but you only mention adding the other bypass cap for the V1B circuit. If the capacitor from the stock shared circuit is still in there as part of the V2B cathode circuit it's still a compromised capacitor and should have been replaced.

[Mars Amp Repair]

I much prefer the aesthetic of separating the cathode circuits. That said...

The shared cathode is a common design element of ALL the Fender BF and SF designs that don't exhibit this problem. I think it's more likely that the shared cathode capacitor was failing/aging and demonstrating higher than ideal impedance. Just replacing that capacitor probably would have solved the matter. You mentioned replacing the resistors when you separated the cathode circuits but you only mention adding the other bypass cap for the V1B circuit. If the capacitor from the stock shared circuit is still in there as part of the V2B cathode circuit it's still a compromised capacitor and should have been replaced.

So ok, I neglected to mention that this was a complete re-cap overhaul, so all bypass caps have been replaced. Of course, the first item to check would be the common bypass cap if they hadn't already been replaced. I guess I assumed too much with the entry, but thanx for pointing that out.

Yes, there are other many other models that have the same sharing V2B cathode network, however I have no idea if anyone ever noticed this issue or complained about it. So, I have a choice: 1) To spend hours trying to figure out why this might be the only one with the issue, or 2) I can spend my time trying to mitigate the issue for this particular unit after investigating all the possible avenues as to why this design isn't working.

So, I signal traced with a high-gain audio signal tracer as well as scope to determine that there was no other path (even thru shared ground points or the power supply) for the bleed-thru to get from the normal channel to the Vibrato channel. It was clear that the shared cathode was the only path, and of course removing the link, proved the results of the diagnosis rather quickly.

Thanx again for the info, so I could clarify for others that might have misunderstood.
Glen - Technician for 12th Fret Music, Boise, ID

[Chuck H]

Well good that cap was replaced after all in case that was the actual issue

If you ever have a mind to test if that cap was the actual cause...

Next time you have a BF or SF reverb amp in you could temporarily solder a pot in series with the capacitor in the V1B/V2B shared cathode circuit. Then you dial in some series resistance on that capacitor. If the reverb bleed effect is similar to what you heard with this amp then it probably was a bad cap.

[Mars Amp Repair]

Um,
Since all the electrolytic caps were original, the overhaul portion of this repair was of course performed before the symptom with the reverb bleed (or any other symptom) was addressed to assure it was not caused by weak decoupling power supply caps or weak cathode bypass caps.
It would have been counter productive to repair the amp in any other order. You don't want to be troubleshooting an issue that might be a part that you're going to replace anyway. That would be very silly (:- )

[Chuck H]

Well it wasn't clear. It's a pretty common MO to repair accute problems and get an amp working before doing any major overhauls. That way you are assured that any issue you find isn't a mistake in your work. What you say makes as much sense as far as I care. So just to be clear...

You're saying the amp exhibited the reverb bleed with the new capacitor installed and THEN you split the cathodes? Please clarify this.

Because you opened with the customer complaint and then confirming that complaint. Did you wait until after the re cap to confirm the customer complaint? Is it possible you confirmed the problem first, then did the re cap and instigated the cathode separation during that process? Therefor never having tested for the problem with a new capacitor on a shared cathode?

I don't want to play this game any more than you do but my asperger's is getting the better of me.

[Mars Amp Repair]

OK Chuck:
1) Amp came in exhibiting the reverb Bleed-thru from the normal channel to the Vibrato channel.
2) The amp also needed a complete re-capping.
3) Tested all tubes as I do customarily, and replaced any that were noisy or just checked bad.
a. Submitted estimate to customer.
b. Was not concerned about any major costs involved with the Bleed-Thru issue as Fenders of this vintage are pretty straight-forward to diagnose. As well, I've seen tons of old Fenders to the point of actually being pretty bored with them, unless I run into an intriguing issue like this one had.
3) Performed the recapping & amp still had the bleed thru issue.
4) Proceeded to troubleshoot the Bleed-Thru issue.
5) Used my high-gain audio signal tracer to find how the Normal channel audio was getting thru to the Reverb driver including possible ground sharing or just bad grounds.
a. The only path I found was the common cathode which did have the channel 1 signal on it.
b. So, the signal was being injected thru the V2B cathode, and being amplified by V2B and then routed to the reverb driver grid of V3. the 12AT7 Reverb transformer driver tube.
Here's the schemo to help.

[g1]

The interesting part for me here is whether the issue is common to this particular model. I've never heard this complaint, but how many people use the normal channel anyway.
If it is due to the shared cathode design, they should all do it.
If not, why this particular amp, and why did splitting the cathode cure it? I'm pretty sure Glen will have tried a tube swap, but if not, would this fix have cured a tube with excess crosstalk between the 2 triodes?

[Helmholtz]

, but how many people use the normal channel anyway.

I think that's an essential point why the effect might go unnoticed.
But if you're using the normal channel why on earth would you want to turn up the reverb (recovery) control?
Only explanation I could think of would be that owner needs to switch between channels using an A/B box.

[Mars Amp Repair]

Hey Hemholtz,
As mentioned, all tubes checked and certainly I initially tried tubes.
As for crosstalk between sections of triodes, V1b is normal channel and V2b is the vibrato channel, so no chance for triode crosstalk.
Also, the bleed thru is amplified in v2b Vibrato channel (thru via the cathode as an input) and fed to the Reverb driver which has it's triodes in parallel.
Why this customer couldn't just turn the Reverb down, I don't know. Perhaps he was feeding another signal in the normal channel while also using the Vibrato channel with an instrument.
I believe I mentioned earlier that I really didn't think this was any particular anomaly with just this amp. I was presented with a symptom, and pursued a solution.
It may never come up again, but the trouble shooting went relatively quick with the high gain audio tracer, and the remedy was simple and didn't change the character of the amp..
So there Ya have it.

[Chuck H]

Ok... Thank you Glen for hanging with me and the concise reporting. As is my wont I ran some simulations

As it happens the stock Fender circuit with the shared cathode at V1B and V2B can put as much as half a volt of signal to the grid of the reverb driver tube with a new spec bypass capacitor. I'm sure the simulation was imperfect but I don't think it was flawed. So how much for an aging capacitor? This made me wonder why we haven't heard more about this phenomenon but then I considered Gregs post aboot both channels not typically being in use at the same time.

EDIT: No that wasn't a typo Greg. I'm poking at you for being Canadian. Actually it WAS a typo and I was gong to correct it but I thought this was funnier

[Helmholtz]

Hey Hemholtz,
As mentioned, all tubes checked and certainly I initially tried tubes.
As for crosstalk between sections of triodes, V1b is normal channel and V2b is the vibrato channel, so no chance for triode crosstalk.
Also, the bleed thru is amplified in v2b Vibrato channel (thru via the cathode as an input) and fed to the Reverb driver which has it's triodes in parallel.

Ok, but the points addressed above were not raised by me.

I believe I mentioned earlier that I really didn't think this was any particular anomaly with just this amp. .

That may well be.and hardly anyone would notice the issue.
I think your explanation with the cathode coupling makes sense and the fact that separating the cathodes worked confirms it.
But with a good, low ESR common cathode cap only low frequencies should appear at the cathodes.

[Chuck H]

But with a good, low ESR common cathode cap only low frequencies should appear at the cathodes.

This was something I considered too. But the reverb driver sees whatever that 500pf can pass and that's where I tested in my simulations. It's certainly enough to shake the springs a little.

There was another thread just a couple of weeks ago with a similar issue and I think it was determined that the speaker and cabinet vibration were shaking the tank springs enough to cause bleed with the reverb channel not in use (but with the reverb control not shut off). I don't remember if this was a Fender amp but even this scenario seems plausable.

[Helmholtz]

As it happens the stock Fender circuit with the shared cathode at V1B and V2B can put as much as half a volt of signal to the grid of the reverb driver tube with a new spec bypass capacitor.

Yes, that seems possible. .
Idle current of V2b is 1mA, so it can swing a peak signal current of 1mA.
The impedance from cathode to ground is 15 Ohm at 400Hz.
So signal voltage at the cathode can be up to 15mV peak (@400Hz).
Multiplying by the voltage gain of maybe 50 gives a plate signal of 0.75Vp.
After around 3dB attenuation by the 500p cap the 12AT7 grid signal would be close to 0.5Vp.
Maybe even somewhat more at lower frequencies.

A shared 250µ cathode cap would lower the unwanted signal by a factor 10 or 20dB.

[Mars Amp Repair]

Wow...Quite an interesting in depth discussion this whole thing has created. So much to learn for everyone.