So, I have guitar direct in to a single-ended class A (12AX7 and 6V6) Champ-style amp. The amp is from the 60s and unbranded, it's likely a Pepco (Pine Electronics) Silvertone style circuit. It's been re-capped and upgraded with a safety ground.

When I shut it off while playing and keep strumming, I get a really nice compressed distortion that slowly fades off as the unit powers down. It also removes the hum (which makes sense, since AC power is disconnected).

I've tried using a VARIAC with the amp to recreate this effect but it actually cleans up the tone (and lowers volume).

My non-EE's guess at what's happening-- A voltage spike from AC shut down is causing a higher-than-normal B+ (DC) voltage (causing distortion) which is stored in the power tube (?) at a lower-than-normal amperage (causing compression and a lower output volume) which then rapidly depletes (using up stored B+ from power tube) as it shuts down.

If I wanted to replicate this tone, should I use a multimeter to find the B+ line and what DC voltage/amperage it's at, then supply the same voltage/amps myself to that line (with the AC power switched off)? That is, if there isn't some kind of inductance happening or stored AC/DC somewhere besides the B+ causing my "sound".

Of course, my own guess could be totally off. I realize you all might be limited without posting the schematic (which I don't have). Any ideas?

  • 3
    \$\begingroup\$ When you shut it off you instantly disconnect the heater power, while the DC rails may discharge a bit more slowly. Turning down the voltage with a Variac won't reproduce that. You might experiment with reducing the heater voltage (use a 0 to 4.7 ohm 5W wirewound pot or even run the heaters off a DC lab power supply) independently of the HT (B) voltage. But if you don't know what you're doing with the wiring, DON'T! \$\endgroup\$ – Brian Drummond Jan 3 '16 at 0:01

Measuring B+ voltage is a good idea. A class A amplifier draws full plate current continuously, so it will drain the power supply capacitors quite quickly. The heater cools down slowly, so I think it is the B+ voltage drop that is causing most of the compression effect.

One characteristic of Pentodes is that the Plate draws relatively constant current at different voltages (unlike a Triode, which acts like a resistor). As a result the operating point will shift off center when the voltage is lowered, causing compression of the negative half of the output waveform.

In the curve trace below I have added load lines for normal B+ voltage (red) and reduced B+ voltage (blue) to show what I think might be happening. For lowest distortion the distances between Grid voltage steps along the load line should be equal. In the blue line you can see that the steps are compressed at the lower voltage end, and because the operating point (vertical blue line) is closer to that end the distortion is higher.

enter image description here

Once you have determined what voltage produces the desired sound the next problem is how to get it there in normal operation. Tube amps often use a two stage power supply filter with a resistor between the first and second capacitors. You might be able to lower the B+ voltage by increasing the value of this resistor.

enter image description here

Running the heater at lower voltage is not a good idea because the Cathode coating could get 'poisoned' (and it probably won't produce the sound you want anyway, since the lower emission will probably just reduce the gain and make it quieter).

  • \$\begingroup\$ This is a great answer. I plan on measuring B+ voltage and maybe trying a pot in place of the resistor you pointed out. I am still curious if anything else might be going on causing the distortion, though. It's more distorted (but quieter, which checks out with low-voltage pentode behavior) than the tone I get with everything cranked, hitting the strings as hard as I can-- It's very bright with a fuzz-like intermodulation distortion. \$\endgroup\$ – jamcube Jan 3 '16 at 5:23
  • \$\begingroup\$ You probably can't produce enough signal voltage to overload the amp as much normally, and the distortion will be different anyway due to the different load point. Don't use a pot - the resistor may dissipate several Watts so a pot would probably burn out. Get a few wirewound resistors of different values (eg. 1k, 2k2, 4k7) which can be wired in series/parallel to make different resistances. \$\endgroup\$ – Bruce Abbott Jan 3 '16 at 5:39
  • \$\begingroup\$ It looks like the methods suggested can't possibly simulate the conditions of AC shut down (exactly). Not advising anyone do this, but if I wanted to brute force it (and probably destroy the tubes, or worse) a sufficiently slow PWM on the 120V AC going in would simulate my amp turning on and off and get the exact sound I want. Some filtering after the PWM might soften the blow and provide the line with some kind of safety capacitor. Sounds unsafe to me, but maybe better (safer?) than PERMANENTLY wiring the amp to run at improper B+ or heater voltage. \$\endgroup\$ – jamcube Jan 3 '16 at 5:53
  • 1
    \$\begingroup\$ 'improper' B+ voltage won't hurt it, and you could install a switch to select different resistances. However it might be better to identify the distortion and reproduce it in an effects unit (that way you can get the desired effect on demand with any amp). \$\endgroup\$ – Bruce Abbott Jan 3 '16 at 6:27
  • \$\begingroup\$ Good point on cathode poisoning. Brief experiments are OK but it's not good for long term life! Playing with the heater voltage was commonplace in the 1920s and early 30s, but that was in the "bright emitter" days before cathode chemistry changed, saving power, but introducing the "poisoning" problem. Rather than decreasing the anode voltage (increasing that resistor) you can also change the valve characteristics by reducing the screen grid voltage (increase the 25k resistor above it). \$\endgroup\$ – Brian Drummond Jan 3 '16 at 12:07

There is a sonic improvement to be had. During power-down the Tube Amp passes through a better operating point. You must repeat your experiment by removing the AC hum by whatever means. Then you'd be comparing apples with apples. Now if you still notice the momentary better sound then check all the valves. Sometimes valve characteristics move with time. Maybe you have a valve that has gone out of spec. If all your valves are good then it's time to look at biasing and filament voltages. Try to emulate the power-down condition.


You probably have the power tube cathode, connected (grounded) to the same path of the first capacitor ground. Separate them. Connect the power tube cathode to the input jack, and the sound will disappear.

  • 1
    \$\begingroup\$ You seem to have read the question wrong. This isn't about getting rid of something, but rather how to replicate a particular effect that happens while shutting off the amp. \$\endgroup\$ – JRE Apr 5 '16 at 12:32
  • \$\begingroup\$ Agreed. The OP wants to create the distorted sound at will, not make it disappear. \$\endgroup\$ – Transistor Apr 5 '16 at 17:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.