Asymmetrical clipping at output

[nsubulysses]

Hi I am curious what would cause an output that at the point of clipping goes asymmetrical and eventually ends up looking like a picture I just came across here -- A Musical Distortion Primer


I have a build that does this and another identical build that does not (well to me it seems identical). Having trouble figuring out why. The good amp has a normal looking output and sounds great. The weird amp has a very asymmetrical clipping the farther you go in volume, and it lacks bass response and volume.

The good amp clips at 105W and the weird one only does 84W. It is not a complicated amp. V1 triodes are split for input 1 and input 2, a normal and bright input, V2 is a cathode follower that drives the tone stack and V3 is an LTP phase inverter. Output section is ultralinear with 4xKT77.

I will try to scope it again but any suggestions in the mean time are appreciated.

[loudthud]

A symmetrically clipped wave sounds hollow. An asymmetrically clipped wave sounds more interesting and contains even harmonics.

When a sinewave is fed to a vacuum tube through a capacitor, as the amplitude of the sine is increased past the clipping point, grid current tends to flow on the top of the wave and this pushes the center of the sinewave in the negative direction. At the output of the tube (plate) the negative portion of the wave gets narrower and the positive portion gets wider. If the tube was biased perfectly in the center of its range initially, the sinewave will clip symmetrically up to a point but the grid current will shift the bias and cause asymmetrical clipping at the output.

Most iconic guitar amps clip asymmetrically, on the order of 40%/60%. The wave in your post is more like 25%/75%. Probably too much of a shift to sound good although it might be useful in some situations.

One of your amps probably has a bad tube, a failed component or build mistake. Follow the signal through the amps to find where there is a deviation.

[nsubulysses]

Thanks LT

I misplaced a solder connection in the tone stack and of course had strange results. Build mistake oh yeah

[MarkusBass]

Most iconic guitar amps clip asymmetrically, on the order of 40%/60%. The wave in your post is more like 25%/75%. Probably too much of a shift to sound good although it might be useful in some situations.

This is very interesting information. I'm fixing Soundcity 50 amp that has 21 and 60 Ohms resistors in cathodes of the power amp (EL34). I thought that this is a mistake but then found out that several Soundcity amps were build with two different cathode resistors. Does it mean that they did it on purpose to get asymmetric clipping - to make the amp iconic?

Mark

[loudthud]

Just when you think there is nothing new, a new wrinkle appears. It's hard to say exactly what the thinking was to use different resistors in the cathodes

Before clipping the offset cathode resistors would have a subtle effect, but you could see it with a distortion analyzer. I would expect even order distortion to increase with power level similar to a preamp tube. The imbalance would disturb the even order distortion cancellation effect of the push pull stage but global feedback will try to correct the imbalance. As the stage starts to clip I would expect one side to clip softer. At full squarewave overdrive one side may have slightly less amplitude, but the OT hides this because it can't pass DC to the output.

I wonder if the side to side difference in OT resistance is compensated for by the cathode resistors, or made greater.

I'm just shooting from the hip here. I could be totally wrong. My experiments suggest that the asymmetry is created at the PI input. The glaring contradiction to that theory is the Trainwreck Express where a cold clipper stage prevents hard overdrive on positive peaks.

[Enzo]

My first thought was to check the OT primary for uneven resistance in the winding, but as I read LT's post, I see he just said exactly that. I wouldn't be surprised. If I arbitrarily assume 40ma through the tubes, the 68 ohm gets me about 2.8v dropped, while the 22 ohm makes about 0.88v. So that would unbalance the bias by 2v out of 36v.

[TimmyP1955]

If you open the NFB loop, you can swap the OPT's primary leads. This should tell you if it's a bad OPT or something ahead of it.

[nsubulysses]

You guys are wonderful and I appreciate all the suggestions but I wired my slope resistor in the tone stack wrong, resulting in a weird output.

Of course I stared at the amp for hours and hours, gave up, posted a thread, then noticed the wrongly placed resistor about 20 minutes later while having a conversation with a friend, not even working on the amp. yep sometimes all you have to do is start a thread and you'll figure it out. But not before.