Treble distortion

[Paul P]

I've been playing with my oscilloscope looking at the signal through my amp
(brownface Bandmaster preamp, SE 2x6v6 in parallel output). Traces below
were taken across the speaker output, with speaker connected.

The clean channel is pretty clean with the treble turned down to zero. A
sine wave still looks like a sine wave even at max output. But turn the
treble up and half the wave distorts. It not so bad at 500hz but gets
worse as frequency goes up.

500 hz, Volume max, Treble at zero : Treble max :

Volume 50%, Treble at max :

1 Khz 2Khz 4Khz

Any ideas what's causing the spike at the beginning of the wave ?

Paul P

[cbarrow7625]

I would guess that it is the bright cap on the volume. That is probably why you see the spike more at 4kHz than at 1kHz, the bright cap is losing it's effectiveness as the frequency shifts down.

In your first two traces you can see the same treble bump start to appear in the "treble max" trace. The bright cap is just going to throw off your frequency balance a whole lot more than the treble control itself.

Remember, the volume control at 50% is still aver large voltage divider (like 900K series, 100k to ground) because it is an audio taper. A bright cap around 900K will have a pretty dramatic effect on the "flat" frequency response you otherwise have.

[Paul P]

Thanks cbarrow7625. If you mean a bright cap as in a 47pf, 120pf or 150pf
(and not the 250pf treble cap) it was not switched in.

Is this acceptable distortion ?

The vibrato channel is different. For some reason it's a lot less loud than
the clean channel (a sine wave that measures 7 squares p-p on the scope
for the clean channel is only 2 squares p-p for the vibrato channel).

With the volume at 6 (on a scale of 1-10), treble at max I get the following
(again no bright cap) :

500hz 1Khz 2 Khz 4 Khz

Funny how the little hump moves from left to right as the frequency is increased.

Paul P

[Bruce / Mission Amps]

Have you tried this same test with it plugged into a dummy load?

[Paul P]

Have you tried this same test with it plugged into a dummy load?

No, but I'll try that tomorrow and post the results. If it still looks the same
I'll try to find out just where it happens inside the amp.

Paul P

[Enzo]

REmember too that a guitar amp is NOT a hifi amp. It is not designed to be flat and distortion-free. It is not made to REproduce sound, it is a primary sound producer, it is part of the instrument really.

[chipaudette]

I have questions about your input signal: What is generating it? A function generator? What is the signal's voltage?

Depending on the strength of the input signal, your distortion could be in the preamp or in the power amp. Knowing the voltage of your input signal will help me understand the levels in your amp.

Even knowing the signal levels, though, the best way to track down that spike is use your scope to measure the signal at various points within your amp -- not just at the output. I'd setup your 4 kHz case with the big spike and then start moving backwards through your amp...look for the spike at the speaker output, look for the spike at the input to the power tube, look for the spike before the last pre-amp stage, look for the spike before the next-to-last pre-amp stage, etc. That's the methodical way and far more direct than any theorizing that I'd do.

Finally, you apparently seem bothered by the spike in your traces. My question is...does the amp actually sound bad because of it? Given how sharp the spike is in comaparison to the wavelength of even the 4kHz signals, that spike has got to be at a frequency of 16 kHz or higher. That is really high frequency. In fact, that kind of high-frequency signal is NOT going to come through your guitar speaker at all. In my opinion, you are not going to hear it directly.

The only way you might hear it is if the spike appears early enough in the signal path that it gets part of the mix of "good" distortion that may be occurring in your amp. In that case, there might be less-pleasant inter-modulation distortion generated between the "normal" frequencies in your signal and this super high-frequency 16kHz guy.

To see if getting rid of this spike helps your sound, you could try adding a capacitor in your amp to filter it out. Since it is so high frequency, this should be pretty easy. Since you're seeing the spike at the output of the amp, I suppose that you could put the cap across your speaker terminals. What do the rest of you think about this???? is this safe??? Given an 8 ohm speaker, I estimate that a 2 uF cap will have a cutoff at ~10kHz, which should filter out the spike. Alternatively, you could also try a cap going from the grid of the power tube(s) to ground. That might be safer...but I don't know what value to use there. Just an idea.

So, yeah, (1) let me the know the input voltage level, (2) try the oscilloscope at various places earlier in your circuit, (3) contemplate whether the spike is actually audible, (4) have fun.

Chip

[Garydean]

When I scope an amp I never inject a signal greater than 1000hz. Messing with the bass and treble I'm able to sculpt the wave. I expect when I turn up the treble the form will change. Injecting a 4K signal I would expect to see a big change in the signal. Has the circuitry been changed? Do any of the ceramic disc's, in the preamp not match the original schematic? Has the amp been modded for higher gain? If the spike is not audible and you're happy with the sound I'd say move on.

[Paul P]

You've all been most helpful, thanks. I've had to learn how to take meaningful
measurements.

Going into a dummy load got rid of most of the mess right away, except for a
smallish peak remaining. I then measured the input signal, coming from an
HP 204D oscillator. It was up around 450mv. I measure the output of my
guitar which was more like 150mv. So I lowered the input and the signal
cleaned up the rest of the way.

Here are some more traces. The input was kept at 150mv and the scope
display was adjusted for each trace to keep the same sort of image. Volume
is at max, treble is at max and bass is at min which produced the strongest
signal.









Looks like nice soft clipping for the most part. The vibrato channel still has
that funny little hump in it but I'll need to do some more tests to see where
that's coming from.

Paul P


,