My first transistor overdrive / distortion

[Banjan73]

Well Antonis. Not exactly Schematics, but the mods you suggested ;-)
No, the Marshall isn`t original a 4-channel. It is a `71 1959T (super tremolo). It is been modded to have 4 channels.
Anyway, here is the updated schematics. I would love to have some suggestions on how to make a better volume control.

[willienillie]

No, the Marshall isn`t original a 4-channel. It is a `71 1959T (super tremolo). It is been modded to have 4 channels.

I see.  I think I can imagine about how it was modded too.  Is every channel voiced differently, or 2 each of the regular normal and bright?

Sorry to drag you off topic.  I like old amps.

[Banjan73]

willienillie:
Its totally refurnished. The whole preampsection is new.
Its this guy who have done it:
https://tommy-folkesson.nu/modifieringar/modifieringsprocessen/
Sry. Its in swedish. My amp is the mkVII mod that he mentiones (in swedish. Use google translate)

[antonis]

Got you now..

Although your oscilloscope waveform is almost "perfect", let's see how it could be "absolutely perfect"..
(in the mean of pure squared waveform (pulse)..)

For zero vertical slope, you need infinite gain, so let's leave it as it is (for the time being)..
For zero horizontal slope (clipped top & bottom waveform lines absolutely parallel to x axis abscissa), you need time constant(*) value veeeeery lower than signal half period(**)..

(*) time constant is refered on each coupling capacitor value multiplied by the sum of impedances before and after it..
(e.g. for Q2 stage, t = C3 x { R4 + [R6//R7//(h_FE+1) X R9)]} resulting into about 6.5ms (for h_FE=150 & negleting r_e)

(**) period is the inverse of frequency (wavelength)

You can see now that, for 10kHz signal (200μs half period), 6.5ms time constant has little effect on horizontal waveform slope (about 3%) but ten times more for 1kHz (30%) and 100 times more for 100Hz (300%)..

If you wish "perfect" horizontal waveform for all of your frequencies of interest you have to enlarge enough all coupling capacitors..


Same (but in reverse period/time constant relationship) stands for vertical waveform "linearity" (not slope), taking into account shunt capacitors (like C2 & Load capacitance..)



P.S.
I might puzzled you a bit but I presume that it should be good to understand the reason(s) for obtaining a particular waveform before you proceed on any altering/improvement..

[Banjan73]

Hmm.
You scare me off the forum, mate
Been a long time since i've done calculations like that, but as I understand this, you're talking about the that the energy of the capacitors need to be released slow enough to make the square wave? Almost like filtering a rectified AC voltage?
But I don't really get why the waves need to be a perfect square wave (or do they? Have I misunderstood?). I get why we want a symmetric wave, but a perfect square?
So, wheres the formulas for calculating coupling caps?
I thought that 100nF was pretty large, but apperently not, hehe..

[Banjan73]

Hmm. Trying to read a little about this now. Wasnt just straight forward. Well, the formula for finding the cap is ok (straightforward filter formula), but the impedance was more difficult.
https://www.electronics-tutorials.ws/amplifier/input-impedance-of-an-amplifier.html

[Renegadrian]

Nice work but nothing new here, just look at this schem there are a lot of others similar just googling, this one has a baxandall tc but otherwise is almost identical. Built this one, it's very good.

[Banjan73]

There you have that 1M/100k biasing.
Might just well be this I stole the idea from. I frankly don't remember😉

[Banjan73]

If I understand it right.
On that scheme, the 10k pot is gain and the 250k pot before the tone stack is volume?

[Renegadrian]

If I understand it right.
On that scheme, the 10k pot is gain and the 250k pot before the tone stack is volume?

Yes

[Renegadrian]

There you have that 1M/100k biasing.

just google bjt gain stage, you'll find it's pretty common arrangement. just look at the lpb1

[antonis]

But I don't really get why the waves need to be a perfect square wave (or do they? Have I misunderstood?). I get why we want a symmetric wave, but a perfect square?

You don't need a perfect squared waveform..!!
(or else, we would deal with Smitt Trigger oscillator perfection)

You stated on previous post (the one with oscilloscope pic) that your waveform symmetry isn't satisfactory so I just tried to point to you some of the particular wave formation causes..

But, let it be and keep your worthwhile endeavor..!!

[Banjan73]

Antonis: that was just an observation (not perfect symmetric). I'll leave it as it is, then.
I remember now, that I stole the 1M/100k from LPB-1, off course.
Now, for the volume control. I'll experiment with putting a pot between q1 and q2. Maybr I need some adjustements for the Rc and Re then, but we'll. See.
I think that I anyway need to read a lot more theory regarding transistors. When I went to engineering school, we learned a lot about OP-amps, but nothing about transistors. That has always irritated me😉.
I am very grateful for all past and future inputs here. Thx guys. Fantastic forum!

[antonis]

I'll experiment with putting a pot between q1 and q2.

Pot between Q1 & Q2 will lower Q2 input signal voltage, hence OUT distortion waveform..
(on pot low end you might only have amplification of Q1 out  - no extra second stage distortion..)

But, yes.. keep trying..!!
(experimental results are much more mind-established than theoretical assumptions..)

[Banjan73]

Yeah.
I actually knew that. (that the q2 wont work as a dist stage in the same way, due to lower input signal).
But, how shall I do it then? My wish is to maintain the same distortion with varying volume. Seems like this circuit lacks a third stage for doing just that, or what?

[PRR]

> wish is to maintain the same distortion with varying volume.

Distortion naturally rises with signal level.

Any other result is un-natural. Requires much glitchy trickery to approach that. And may not be musical.

[Banjan73]

Well.
I know that. That not my point here. I think i expressed myself unclear.
I need a volume control on the pedal, to control the volume of the effect, not distortion. Like a boss OD-1.
Overdrive (here distortion) and volume.
Is it possible to put a volumecontrol to this circuit without messing with the amount of distorion?

[Banjan73]

Actually, when i look at that italian schematic thing, I think I eill try to increase the gain of the first gainstage. It may self oscillate at a point, or what?
But if I maximize the gain in the first stage, the second hasnt that much to say in creating distortion. The, perhaps, the volume control at the end will work more properly (if I lower the gain in the second stage).
Just what I am thinking right right now..

[PRR]

> perhaps, the volume control at the end will work more properly

This is a useful thing, and very popular.

[Banjan73]

I've been experimenting a little bit, trying to get more gain out of stage 1 and not very much out of stage 2.
The Rc and Re are as follows now:
R4=18k, R5=40, gain pot=40 (dont know if it exist, but 50ohms is ok), R8=3,9k, R9=1k.
Heres the pictures of the scope.
The first is measured after q1. The second is only q2 isolated. The third is the total circuit with max volume.
I think this might work, and if its too little distortion, i think I might have a little more to go on.