Reverb fart on scope

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I've been experimenting with my '71 Fender Deluxe Reverb amp. I added a Larmar type post PI master volume control. This allows me to set the "volume" control high to get overdriven sounds from the PI's clipping, while keeping volume low. It's a common "mod."

Now that I'm setting the drive (volume ) control high I'm noticing a "fart" sound during the initial attack. More so on lower notes and only with the Reverb control above 1 (on.) I changed the associated tubes and I think I've isolated it to the signal coming out of the reverb drive tube (bottom left in the schematic.)

The point "A" trace on the scope (see scheme) stays essentially sinusoidal at all volume levels whereas the "B" trace shows what happens at higher volume settings. The pic was taken with the volume control around 7. The bottom (A) grid signal is not inverted and the tope (B) plate is inverted on the scope.

I have never scoped this area of an amp so I don't know if this is normal. In stock form, w/o the post PI attenuation that I have added (not on the scheme,) the volume out the speaker would be loud when reaching this point of distortion on the "B" signal. And the 6v6 outputs would be distorting from being overdriven. So this reverb fart would normally not be as noticeable with all that volume and output clipping going on. With the post PI attenuation I'm noticing it and would like to resolve it.

I tried disconnecting the reverb tank on the secondary of the tranny on point "B" and doing this made the "B" signal much stronger in amplitude. No surprise since the load was gone. I did not turn the volume high since it would have gone off my scope's scale and it doesn't seem to be an important mode to test. I noticed the problem is more pronounced with lower frequency input signals. Higher frequencies cause the distortion in the wave to mitigate. I had tried changing the 500pf cap before point "A" to 250pf but that did not help so I put the 500pf back in. The cathode of the tube in question is reading about 8.6 volts at idle which is very close to the scheme.

Has anyone ever looked at the signal at this point in a Fender? Do they all do this with the volume control set above 6 or so? What is that shape of a trace called? Ringing? Any common root causes?
 

Attachments

  • DR-Scheme.jpg
    DR-Scheme.jpg
    128.1 KB · Views: 144
  • Reverb-SFDR.JPG
    Reverb-SFDR.JPG
    23.6 KB · Views: 149
Last edited:
Looks like typical single-ended transformer coupled stage clipping to me. One half wave (usually the bottom one) clips much earlier than the other one and the severe clipping coupled to using an inductive load also introduces a resonant peak at the point where clipping begins. That peak, if high enough in voltage, can also arc through the sockets or the transformer windings, BTW. Sorta self-destruction mechanism of cranked tube amps.

If the frequencies are very, very low you might also see components of the famous "transformer saturation", which on scope basically looks like chopping off pieces of the waveform and skewing some other parts. Since that is the plate signal I'd rule saturation out, though.

Yep, both distortion mechanisms will pretty much sound like farting or an off-tune radio. That's the revered SE tube amp clipping in all its glory. Wait till you scope the output of something like an overdriven Champ amp....

The solution: Don't overdrive the reverb driver stage.
 
Last edited:
Hey teemuk

I have a '74 (Vibro) Champ and I have seen the clipping on the bottom of the current wave of the 6v6. But it's nice and flat. I've seen this described as "Cold clipping." I remember a discussion on the '65-'78 Champ's output clipping. Increasing the idle, assuming one had a reason to do so, is not good since it already idles "too hot."

Looks like the 12at7 is idling at 4ma total (2ma each triode.) I think a 12at7 can safely idle at significantly higher current. So would it make sense to try a lower cathode resistor to bias the tube more towards the middle of the swing?

I think this is an example:

http://www.el34world.com/charts/Schematics/files/fender/twin_reverb_sf_100_schem.pdf

Looks like the 100watt Twin on the above link idled this tube at 12.8ma total. And the Fender part#022921 of the reverb tranny is the same on that Twin as on my DR. So I guess my tranny can take some more idle current? When considering the voltage drops it looks like the Twin dissipates 0.04watts and mine 0.02watts thru the primary at idle on the schemes.

Or maybe my clipping is not the tube reaching cutoff? I am not knowledgeable about how transformer coupling differs from capacitive coupling on scopes.

And that Twin scheme shows a 560pf cap across the primary of the tranny. Might that help me?
 
If you significantly double the idle current,
you may run into transformer saturation, another undesirable effect.

The simplest solution is this:
Add a second tranny in parallel to handle the current.
Since this is a small tranny, it should be cheap (and probably is the smallest they could get away with).

Alternately, you could get a larger tranny with the same impedance,
which would probably be better.

This increased headroom should compensate for any
increased distortion due to a lower load on the tube.
But if the load isn't too low, the sound might actually be sweetened
by a little harmonic distortion.

Of course the better solution here might be the obvious:
If you want tons of clean reverb, add it via a high quality reverb unit
(nowadays they make units that rival and imitate studio units),
by splitting your guitar signal and feeding the pedal to a 2nd channel.

Ideally for recording, you will probably want to add quality reverb
at the recording console instead, or during digital mixdown.
 
I tried a 500pf cap across the primary like on the Twin. I caused ringing at the edge of the horizontal area of the wave so I removed it. I then replaced the cathode resistor with a pot and dialed in from 500-2200R. Definite improvement in reducing the cutoff period as hoped. Still reaches cutoff but I can now reach "8" on the volume (drive) control with no fart which allows for nice additonal crunch. I'm not seeing the transformer saturation on the wave, or at least the type of examples I found. The idle is now 8.8ma total so I think the tube will be ok.

Here's a pic with the amplitude of the primary at the same peak to peak. voltage. Much cleaner. Note that the grid trace is a little less strong while it achieves that same output/primary amplitude. The increased idle has increased the gain too I guess.

Thanx for all the help and comments!
 

Attachments

  • Reverb-SFDR-2.JPG
    Reverb-SFDR-2.JPG
    24.8 KB · Views: 78
I tried a 500pf cap across the primary like on the Twin. I caused ringing at the edge of the horizontal area of the wave so I removed it. I then replaced the cathode resistor with a pot and dialed in from 500-2200R. Definite improvement in reducing the cutoff period as hoped. Still reaches cutoff but I can now reach "8" on the volume (drive) control with no fart which allows for nice additonal crunch. I'm not seeing the transformer saturation on the wave, or at least the type of examples I found. The idle is now 8.8ma total so I think the tube will be ok.

Here's a pic with the amplitude of the primary at the same peak to peak. voltage. Much cleaner. Note that the grid trace is a little less strong while it achieves that same output/primary amplitude. The increased idle has increased the gain too I guess.

Thanx for all the help and comments!


what was the value you ended up dailing in on the pot?

Can you draw the stage again, and add a few voltages (plate, bias, etc.)?
 
Pot was around 750R where it looked improved enough for me. I settled on an 820R that measures 836R. Cathode voltage is 7.4 so idle is 8.9ma. Some googlin implied that the stock tranny may be good to 15ma dc but it's all speculation apparently. Plate was measuring 435 originally with the PS node at 443. Plate is now 423 with PS at 440.

I got concerned over the additional gain and also remembered that the Twin scheme does not use a cathode bypass cap. So I tested with the 836R cathode and no bypass cap and the p-p amplitude was about half. So I rigged a pot in line with only the bypass cap (with the 836R in place of course) and adjusted for about the same gain as before the experimentation began. The pot read about 200R so I added a 220R to the bypass line.

Sounds good so far but the little woman works tomorrow so I'll be able to test at all volumes.:cool:
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.