Low Watt Tube Amp Clipping Diodes Circuit

[K Teacher]

Has anyone tried this circuit? Does it work (well)?
 

Questions about components:

- Marshall uses 1N4001/ 7 for D1 & D2. Are these OK, or 1N4148 would be better there?
- Estimating the signal on V1a grid at 30mV-peak and V1a gain of approx. 50, the signal on V1a plate would be 1,500mV-peak. The silicon diodes will clamp this signal at 600 to 650mV (little less than half). Isn't that too much?
- The signal after C1 will swing positive and negative above ground. In this case, shouldn't C3 be 2x 20µF back-to-back (negative leads connected) to form a non-polar 10µF electrolytic to avoid component failure overtime?
- What voltage for C3?

Thanks in advance.

[Merlin]

- Marshall uses 1N4001/ 7 for D1 & D2. Are these OK, or 1N4148 would be better there?

1N4148 have less capacitance, but that's not necessarily "better" in a deliberate distortion circuit. Whatever sounds good.

- Estimating the signal on V1a grid at 30mV-peak and V1a gain of approx. 50, the signal on V1a plate would be 1,500mV-peak. The silicon diodes will clamp this signal at 600 to 650mV (little less than half). Isn't that too much?

Too much for what? Do you want distortion or not? Besides, there will usually be more clipping later on in the preamp, which will mask some of what these diode do.

- The signal after C1 will swing positive and negative above ground. In this case, shouldn't C3 be 2x 20µF back-to-back (negative leads connected) to form a non-polar 10µF electrolytic to avoid component failure overtime?

If the cap is big enough there will never be more than one diode drop across it, which is safe for a non-polar. 10uF is probably big enough, since it's reactance is only a few hundred ohms even at 20Hz, which is a hundred times less than the output impedance of the tube. To be honest I don't know why C3 has been included, it looks unecessary to me.

- What voltage for C3?

Since the voltage at the anode is clipped to about on diode drop, a low voltage cap can be used.

[K Teacher]

Hello Merlin,
Thanks for your quick response and invaluable insights.

C3 doesn’t seem necessary to me as well. One reason I can think of is it could round the edges of the clipped signal, giving it a more tube-like sound. Does this make sense to you?

Thanks again.

[Merlin]

C3 doesn’t seem necessary to me as well. One reason I can think of is it could round the edges of the clipped signal, giving it a more tube-like sound. Does this make sense to you?

It could do that at low frequencies if it was small enough, like <1uF, but at 10uF I don't think it will be having any audible effect.

[K Teacher]

OK.
Thanks again.

[stratomaster]

C3 doesn’t seem necessary to me as well. One reason I can think of is it could round the edges of the clipped signal, giving it a more tube-like sound.

In pedals this is accomplished by putting the cap across the diodes.  The cap is usually very small in value (<220pF) unless a drastic softening of the high end harmonics is desired. 

I think Friedman and others use a pot between signal and diodes or between diodes and ground to adjust the effect of the diode clipping without switching diode types for clipping threshold.  This is also done in pedals.

[K Teacher]

Stratomaster, thanks for your input.

I've seen what you're describing here:
   http://www.geofex.com/article_folders/tstech/tsxtech.htm

However, the diodes (and the paralleled cap.) are inserted in the opamp feedback loop, instead being from the output to ground, as in the schematic on the initial post. I'm not sure if the opamp circuit will work well on a preamp tube circuit. Additionally, this parallel capacitor will have an increased effect on the next tube stage due to the Miller effect.

What I'm trying to accomplish is this:
   

 
This video shows the end result of the mod, but not how it was done.

Does anybody here have an idea how to do it?

[Zenabi]

Stratomaster, thanks for your input.

I've seen what you're describing here:
   http://www.geofex.com/article_folders/tstech/tsxtech.htm

However, the diodes (and the paralleled cap.) are inserted in the opamp feedback loop, instead being from the output to ground, as in the schematic on the initial post. I'm not sure if the opamp circuit will work well on a preamp tube circuit. Additionally, this parallel capacitor will have an increased effect on the next tube stage due to the Miller effect.

What I'm trying to accomplish is this:
   

 
This video shows the end result of the mod, but not how it was done.

Does anybody here have an idea how to do it?

The video provided a link in the description. Here it is.
https://www.instructables.com/Add-Diode-Clipping-Distortion-to-your-Guitar-Amp/

Friedman does similar Jose Mods in his amps. He uses diodes and transistors. You can Google it. These mods are quite well documented.

You could also YouTube "headfirst Jose mod" to watch videos where he explains everything.

[stratomaster]

However, the diodes (and the paralleled cap.) are inserted in the opamp feedback loop, instead being from the output to ground, as in the schematic on the initial post. I'm not sure if the opamp circuit will work well on a preamp tube circuit. Additionally, this parallel capacitor will have an increased effect on the next tube stage due to the Miller effect.

In the pedal world there is "soft" and "hard" clipping.  The former typically refers to antiparallel diodes in the feedback loop of an opamp. The latter typically refers to antiparallel diodes to ground (or some other reference voltage functioning as AC ground). 

You found RG's technical dissection of the TS circuit (probably the most famous soft clipper). 

For an example of a hard clipper take the DOD Overdrive 250/Dist+ (or even a Klon for a more advanced application of this principle).  I've included a screenshot of a schematic with the relevant portion highlighted.

The cap can be employed in either config, and since it's on a switch in your implementation the Miller capacitance effect is only happening when the diodes are engaged--which you might want in this case.

The Jose mod uses the diode function of transistors (whose internal capacitance effects might nullify the need for an additional parallel cap) or even Zeners for higher clipping thresholds.  The principle is the same, clamping signal to ground for hard clipping.  You can have fun with different thresholds for the positive and negative swing as well for asymmetrical clipping. 

I'm not personally a fan of seeing this in an amp as I think it can be done better in the context of a pedal (with additional filtering, EQ, etc) but there is a whole world of possibilities with clipping thresholds, asymmetry, parallel caps, and series pots available for experimentation in an amp.

[K Teacher]

Zenabi & Stratomaster,

Thanks for the info.