Real McTube II Waveform Study (pics)

[Paul Marossy]

I had a spare moment tonight, so I thought I would use my somewhat dusty oscilloscope to look at the waveforms produced by my latest build, the Real McTube II. Kind of interesting. All pics are with volume at 3:00, input signal was 600Hz, tube is a Sylvania 12AU7 (the waveforms didn't really change at all with frequency in terms of the clipping that occured)
Volts/Div control changed as necessary to get waveform to fit on the CRT screen for each picture. Slope +.

Here is the waveform with the gain control at 9:00. Pretty much a sine wave as it comes from the signal generator.



Next, with the gain control set at 11:00. The bottoms get a little more rounded.



Gain control set at 1:00. Now the bottoms are getting more squashed.



Gain control set at 3:00. Pretty much your classic tube clipping.



Gain control at max (5:00). Look ma, asymmetrical clipping!



Just thought someone might find this as interesting as I did...  

[Sic]

woo, that is awesome, i love seeing that kind of stuff

[Ge_Whiz]

Okay Paul, it looks like it sounds good. But does it sound like it sounds good?  

[RobB]

That’s really interesting Paul.  Thanks for posting those pic’s.  
I notice that in pic’s 2,3 and 4 there was more clipping at the bottom than at the top.  In the last pic (maybe I’m not interpreting this correctly) most of the clipping seems to be at the top.  Did the Real McTube II do this by itself or did you by chance reverse the wires when this shot was taken?  

Also, shots 2 and 3 look exactly like what I saw when I scoped  Joe Davisson’s diode compression module a few weeks ago.

[Paul Marossy]

Ge_Whiz-

Well, how it sounds it subjective, of course.  :wink:
I think it sounds pretty good with a 12AU7. A 12AX7 sounds a little raspy to me. I have a few 12AT7s on order, I expect those to sound great with this circuit the way I have it built.

RobB-

Yeah, the way it clips the signal is curious, isn't it? The way those pictures were taken was with alligator clips connected to the scope probe, so there is no chance that I accidentally reversed polarity. Regarding the last picture, it appears that it is now clipping the top more a lot more. That is what the McTube was doing all by itself.

What kind of diodes does Davisson's diode clipping module use? It looks kind of tame for diode clipping...

[Doug_H]

[;lkj

[Paul Marossy]

Yeah, I think the next thing I am going to do is try a 12AT7 and then a 12AX7, too. It would be interesting to see what I think sounds "raspy".  :wink:

[Transmogrifox]

Yeah, the way it clips the signal is curious, isn't it? The way those pictures were taken was with alligator clips connected to the scope probe, so there is no chance that I accidentally reversed polarity. Regarding the last picture, it appears that it is now clipping the top more a lot more. That is what the McTube was doing all by itself.

I think you could shed a little more light on this phenomenon by probing both stages (this is using 2 stages, correct?).  When I see something like that, I suspect the first stage keeps clean at lower volume settings and therefore has more headroom on the positive cycle to cause the second stage to clip, thus the initial assymetry on the bottom.  As you crank up the volume pot, the signal amplitude increases enough to where the first tube begins to break up.  When this happens, the first tube likely limits the signal more on the saturation side, which is inverted on the output of the second stage.  The effect is that the second stage is then amplifying the assymmetry of the first stage after the first stage begins to break up.

This isn't completely well thought through, but it would be interesting if you dug up the scope again and tried to track down the source of this phenomenon.

Thanks, by the way, that was interesting.

[Paul Marossy]

Transmogrifox-

You're welcome.  :wink:

That sounds like a reasonable explanation to me. I guess I should check the first stage next, huh? Incidentally, I fiddled with the volume control while looking at these waveforms, and I couldn't see any noticeable difference in the shapes other than the amplitude changing. So I left the volume at 3:00 and took the pictures. The gain control is what causes all the action to happen. It seems to me that it is the first stage pushing the second stage into distortion, which is controlled the by gain pot.

[Transmogrifox]

The gain control is what causes all the action to happen. It seems to me that it is the first stage pushing the second stage into distortion, which is controlled the by gain pot.

Yeah...gain pot...duh :wink: , I'm using the wrong word for the picture in my mind.  The gain pot is like a "pre-volume" pot driving the tube gain stages.  That must be why I was thinking "volume". Yes, the *gain control*

[Jason Stout]

Have you tried sweeping through all audible frequencies yet :wink:

[Paul Marossy]

Yeah, I did some spot checks. The frequency didn't appear to really affect the wave shape at all. I remember trying 80Hz, 100Hz, 600Hz, 1kHz, 2.5kHz, they all looked pretty much the same. I didn't try all the audio frequencies since the frequencies produced by an electric guitar generally aren't above 3kHz.

[puretube]

the 1st 3 pics are asym clipping already
(you`d scope this also after 1st stage);
next stage swaps polarity, and clips the
other side of the wave in the same manner.

afterwards, you`re entering "varying pulswidth area",
mentioned around here somewhere, lately...


PS.: it could wll be vice versa - 2nd stage clips firstly,
1st stage clips later...

[Paul Marossy]

Yeah, I know. They really are all asymmetrical clipping. It's just really obvious in the last picture.

Varying pulse width. I do remember that the waveforms in the last two pictures didn't change in amplitude, but the pulse width did...

[Paul Marossy]

OK, next question: How in the world do you get pulse width modulation with a tube circuit? I thought that could only be done by digital means...

[WGTP]

So with the control at 4 o'clock does the pulse width become the same on both sides?  Isn't it between 3 and 5 that it flops?  

[Paul Marossy]

IIRC, at 4:00, the pulse width is halfway between the way it looks in pictures 4 & 5.

[puretube]

that kind of PWM has been done with tubes in some delta/sigma A/D converterters (tube) since the early 50`s...
(the comparator/integrator thingy)

[RobB]

What kind of diodes does Davisson's diode clipping module use? It looks kind of tame for diode clipping...

Diode compression.  Joe's reverse biased diode technique is what I was refering to.  
http://analogalchemy.com/diodecomp.html

[Paul Marossy]

puretube-

Hmmm... very interesting.  

RobB-

Thanks for the link. Now I get the idea, pretty clever use of diodes. I used to wonder what would happen to a signal if it passed through a diode instead of a diode connected to ground for the usual method of clipping the signal. Now I wonder if I could use one of my duo-diode tubes in my next tube "stompbox" to create a similar effect with a 12AX7 preamp tube...