Reverb Noise

[Daver]

Enzo,
I have 10.5 VDC on the cathode of the driver and about 6 mVDC on the grid. If I raise the cathode resistor from 2.2K to 3.3K it helps some. I get 11.5 VDC on the cathode then. I still don't get why a resitive load or speaker gives me a clean sine wave, but with any pan connected I get this distortion. Thanks.
Dave

[jmaf]

Enzo,
I have 10.5 VDC on the cathode of the driver and about 6 mVDC on the grid. If I raise the cathode resistor from 2.2K to 3.3K it helps some. I get 11.5 VDC on the cathode then. I still don't get why a resitive load or speaker gives me a clean sine wave, but with any pan connected I get this distortion. Thanks.
Dave

Dave, what is the standby DC voltage between pins 2 and 3 of the 12AT7 at the moment when distortion starts? Going by Enzo's idea, I think that grid is going positive against the cathode right when the distortion begins.

[Daver]

jmaf,
When the distortion starts I get 14 VDC between pins 2 and 3. Thanks.
Dave

[jmaf]

jmaf,
When the distortion starts I get 14 VDC between pins 2 and 3. Thanks.
Dave

Dave, I think that's your problem right there. I assume you had the positive lead on pin 3, so you have 14 V negative, it's too much bias. If it's positive then worse yet.

Could be that BOTH your 2n2 and 500 pF may be leaky, but it's unlikely as they're in series.

Most likely you have too much current on that tube for some reason. A coworker here at our workshop just peaked over my shoulder and I told him your case, he said he fixed a similar issue by inserting a 47k one watt resistor between the transformer and the 12AT7 anode to reduce tube current. If you have the time you might want to give it a try.

[Tom Phillips]

… I still don't get why a resistive load or speaker gives me a clean sine wave, but with any pan connected I get this distortion. Thanks.
Dave

Dave,
The reverb can input transducer is not exactly equivalent to a speaker or a resistive load. Therefore, if some parameter in your system is just crossing over the edge of acceptable limits, then the difference between the speaker and the reverb transducer could give different results.
I’d suggest that the problem may be the 500pF cap. Is it really 500pF?. It’s in the signal path and it seems that the same part has been there during the whole troubleshooting process. May be it’s mismarked or defective and the capacitance is much larger than 500pF. Anyway, it would be an easy thing to change.

Tom

[Daver]

Thanks, guys, for your help.

jmaf,
That voltage measurement is indeed positive with the positive meter lead on pin 3 (cathode) and the negative on pin 2 (grid).

Tom,
I removed the 500pf cap and it measures 494pf. I then replaced it with a 470pf (all I have) and it changed nothing. I realize the pan transducer is not properly replicated using a speaker, transformer or resistor, but the difference is so clear cut. The pan produces distortion and all the others are clean.

Dave

[Tom Phillips]

Oh well. So much for that idea.
The circuit drawing you showed should work of course. It this case one clue is that your amp has always had this problem since you built it. Right? Therefore, either there could be a mis-wiring / layout problem or there could be a combination of two different small problems that would not normally be noticed but would together cause the symptom you have described.

As Trem suggested, maybe let it rest a few days and start with freash eyes.
As already discussed the voltages you measusred don't seem quite right.
It might be helpful to post more of your schematic with the measured voltages written in and a phot of the layout.
These things can be frustrating. Hang in there.

Tom

[tubeswell]

Hi Daver

Out of curiosity, what does the signal look like on the grid of the other stage after the EQ recovery 1/2 12AX7? (seeing as how you have a 220k plate resistor there) - Its probably clean (by the way you have described getting a clean signal with a resistive load) but it might be worth considering albeit briefly.

[Daver]

tubeswell,
Do you mean the mix stage or the reverb recovery stage?

Tom,
If I use the resistive load, I get the same grid to cathode voltage measurements as I do with the pan attached. So, it seems that doesn't change how the tube is operating. But I still get a clean signal with the resistor.

Maybe I'll just ignore it for the weekend. That's hard for me to do.

Thanks for the brainstorming, gentlemen!
Dave

[tubeswell]

The EQ recovery stage (from whence you are tapping the reverb driver input)

[JoeM]

Dave,

Can you possibly post scope screen shots with the Pan connected and one with a resistive load?

[Daver]

tubeswell,
Got ya. The signal at that point is clean.

JoeM,
Here are the scope pics. The first three are with the reverb pan/tank and the second three are with the resistor. The volume was on full. The signal is traced at the reverb input jack. The first pic in each set are with the dwell at about 1/3, the second with the dwell at 1/2 and the third with the dwell at full. Enjoy!
Dave

Reverb Tank






Resistor

[Tom Phillips]

I think the reason for the difference between the resistor and the tank is that you are seeing the reverb kickback from the springs in the tank when the tank is the only load.
Hold down the springs with some foam so they can't vibrate and note how the waveform changes.
Tom

[Daver]

Tom,
Excellent idea! Unfortunately it didn't change anything. I used the foam the tank shipped with and jammed it tight in there so nothing moved. I even tried banging around the tank and it didn't change the input waveform. I appreciate the help. I was hopeful that would change something.
Dave

[Enzo]

Connect the drive to a small speaker instead of the reverb pan. The reverb pan transducer has some inductive kick, and so would a speaker voice coil. What the hell, it's science.

By the way, chasing scope traces is fun, but at reasonable amp settings how does the reverb sound?