Analog ctaver how it works

[POTL]

Hello.
People on this forum have repeatedly helped me in finding answers to questions, I continue my way in learning the principles of the work of various effects and began to study octaves.
Unfortunately, I only partially understand the principle of this effect and I want to understand it completely.

I was interested in the project on the site madbean, it is a combination of two famous pedals BOSS OC-2 (Pearl Octaver) and Foxrox Octron.
Help to understand how this scheme works.
I will break the scheme into parts and ask questions and write how in my understanding they work, please correct if I misunderstand.

1) IC 5 A + B work with something like peak detector
Each half of the scheme limits one half period, but for what?
2) IC 5 C + D summarizes the signal before the limitation and after, C is out of phase with D
3) The summable signal falls on IC6A, what does this stage do? Does it work like a mixer?
4) IC6B is a flip flop divider - it lowers the signal by an octave, IC7A does the same thing, dropping the signal by one octave, am I right?
5) The processed signals after IC6B and IC7A fall on the differential amplifiers IC3D and IC3B, respectively.
6) What functions in differential amplifiers perform field-effect transistors and germanium diodes?
7) Why used differential amplifiers, but not summing amplifiers? Is this made to reduce noise?
I understand correctly that IC3A, IC3C, IC4B filters reduce distortion after a square wave from flip flop divider?

[Rob Strand]

There's a detail write-up here,

http://moosapotamus.net/stompboxology.html
http://moosapotamus.net/files/stompboxology-freq-div-sub-synth.pdf

[POTL]

There's a detail write-up here,

http://moosapotamus.net/stompboxology.html
http://moosapotamus.net/files/stompboxology-freq-div-sub-synth.pdf

Thank you very much 

[POTL]

Hi again
I read the article, but I'm not sure that I understood the article correctly.

Let's check it out.
IC4B forms a pre-filter.
IC5A / B / C / D turns the sine into a square.
IC6A summarizes 2 signals in phase and antiphase to produce a full square wave.
IC6B and IC7A divide the signal into 2 and 4, respectively, thus obtaining sounds 1 and 2 octaves lower than the original.
IC3B and CI3D summarize the signals and at the output we get the difference of the processed signals, which are then processed by low-pass filters.
I understand how it works?

[Rob Strand]

I read the article, but I'm not sure that I understood the article correctly.

I understand how it works?

You did well.  That's pretty much the it.

There's a only a few finer points.  Which you probably should ignore, except the last one.

IC5A / B / C / D turns the sine into a square.
IC6A summarizes 2 signals in phase and antiphase to produce a full square wave.

For a sinewave input it's pretty much square.  For a guitar signal roughly square.
The "square" wave edges aligned to the peaks of the sine-wave.

C6B and IC7A divide the signal into 2 and 4,

Once you go through the divider it is square.

IC3B and CI3D summarize the signals and at the output we get the difference of the processed signals

It "flips" the input signal to mold it into a lower octave.  It does it in a tricky way which is explained in fig 9. in the article.  A less-tricky way would be the method in fig 8 of the article.   The fig 9 method is nice and smooth.

Notice also that by using either of the methods in  fig 8 or 9 the level of the input signal is retained (and also some of the character of the input signal).   If we took the lower octave from the output of IC6B the level of the input signal would be lost and it would just a buzzy square-wave.

[POTL]

Thank you now all understandable.
I found another version of the analog octave that is different from other analog pedals, it's the Mu Tron Octave Divider, I've never played it, its circuitry looks much more complicated and I'm wondering - does its octave sound better than the octave of the BOSS OC-2 or is it inferior in sound quality and tracking?

[Rob Strand]

I found another version of the analog octave that is different from other analog pedals, it's the Mu Tron Octave Divider, I've never played it, its circuitry looks much more complicated and I'm wondering - does its octave sound better than the octave of the BOSS OC-2 or is it inferior in sound quality and tracking?

I haven't played through one either.  The sound of the basic octave down is similar but you would have to put them side-by-side to get a final call.

If you chuck out a lot of the circuit like the octave up and the power supply the circuits do have a general similarity and complexity.   The OC-2 obviously doesn't have those items.

As for tracking I don't know.  Most analog octaves mistrack to some degree.  I use an OC-2 for bass and some basses tend to mistrack more than others.  The part of the circuit that handles tracking is different but I wouldn't like to speculate which is better.

[merlinb]

I understand how it works?

Yes. The square wave is used as a control signal to alternately invert or not invert the audio (e.g. sine) wave.

[POTL]

I understand how it works?

Yes. The square wave is used as a control signal to alternately invert or not invert the audio (e.g. sine) wave.

Hi Merlin.
I understand correctly that your method is close to what Mutron octave divider does?
(I'm talking about the waveform)
Already studying the article on your site 

[POTL]

Bump

Hi, I read again all the articles and understood the principle of work. I want to thank you!
I have a few more questions.
1) Pretty simple - I studied several octave schemes - DOD / EBS / MXR / BOSS / Pearl / Ampeg - they are close, but what I don't quite understand is the position of the diodes in the circuits. In the scheme in the first message, the diodes in the IC5A and IC5B harness are located in different directions, with the DOD and EBS diodes being identical. Does it have any meaning go or not?
2) A rather complicated and extensive question - the work of the lower octave in a mu tron ​​octave divider.
The principle of operation is similar to other octavera, but the harness itself is very different.
Who can tell about how the cascades are called using IC4A / B, IC5A / B, IC6A / B, IC9, Q3 / 4/7
and how they work.
I wonder, is this approach more modern / perfect or, on the contrary, more vintage and inferior to simple octavera?
I don't understand the need for a stabilization switch - if it improves the operation of an octavera, then why there is a choice of modes, you can make it turned on by default.
http://byocelectronics.com/dividedoctaveschematic.pdf

[Rob Strand]

In the scheme in the first message, the diodes in the IC5A and IC5B harness are located in different directions, with the DOD and EBS diodes being identical. Does it have any meaning go or not?

I don't have a link to the EBS.

It doesn't make sense to me.  I suspect a bug on the schematic.   I found the DOD FX35 schematic and some PCB pics and the both diodes on IC3A are reversed on the schematic.

A rather complicated and extensive question - the work of the lower octave in a mu tron ​​octave divider.
The principle of operation is similar to other octavera, but the harness itself is very different.
Who can tell about how the cascades are called using IC4A / B, IC5A / B, IC6A / B, IC9, Q3 / 4/7
and how they work.

It's a complicated circuit and I haven't got the time to today to go through it in detail.  I can give you some tips.

1)  The first point to notice is 4013b requires positive clock edges to change state.  That means it will change state when 4013a gets a positive pulse on the *reset* line.  For that to happen the input to IC4b must go negative.  So to summarise the circuit synchronizes the to the negative zero crossings of the waveform.

2) The whole point of the rest of the circuit is to re-arm 4013a with the set signal (S) at the right point so it doesn't mistrack.   You should realize that when you get a negative zero crossing the signal might jitter  +/- right after and you don't want to re-arm the detector so soon. 

3) IC9 has two trigger paths for the set signal. 
   The output of IC5b selects the trigger path:
   - When the output of IC5b is low the trigger comes from IC4a. 
   - When the output of IC5b is high the trigger comes from IC6a.
   - When the stabilize switch is on the output of IC5b is held low so it selects IC4a as the trigger.
   The 4013a is rearmed when the trigger signal goes high; from either IC4a or IC6a.

4) The IC4a trigger:   When the circuit starts-up with no input IC4a pin 2 will be at zero voltages.
    The output will generate a trigger when the input in pin 3 goes from positive to negative.
    Once the signal has been present for a while, the DC level on C24 will build up to near the signal peak
    then IC4a won't generate triggers until you are near the positive peak.
    So this delays the re-arm "set" signal to quite late in the cycle push it close to the next "reset" and
    so preventing false re-triggers.

5) IC6a retrigger.
    This is tricky and I haven't got time to go through it.

    The IC6a trigger comes from IC5a.
    To get a positive output at IC6a, Q6 must turn on, and that means we need a positive edge at
    the output of IC5a.
    R40 and C26 try to filter the ripple off the rectifier.  C25 and R41 let the ripple pass through.
    When the ripple has a negative direction the output of IC5a will go positive and generate a trigger.
    *** More work required here:
    IC6b is some sort of lock-out circuit for IC6a.  I have a feeling it tries to tune the circuit
    so it doesn't generates the re-arm pulse late in the cycle.   I have not studied the details
    of how it does it.

   The other thing is with the stabilize switch off IC5b will now be toggling on and off depending
   on the signal.  It's kind of automatically selecting between the two triggers, presumably trying to pick
   the best one.  That needs to be analysed too.

[Please post any fixes.  I'm sure there's typos here and there.]

FYI,  I think C11 connects to the wrong spot.  Maybe it should go to the opamp output?
http://byocelectronics.com/dividedoctaveschematic.pdf

[POTL]

I found the original circuit on the FSB.
Indeed, in the scheme C11 is installed incorrectly.
I thought that the tone cancels the work of the entire filter, but it turns out it moves us between the 1st and 2nd stages of filtering, it looks more reasonable.


Thanks for the extensive answer, I am not a native speaker of English, so it will take me some time to comprehend all the information, thanks again)

[Rob Strand]

I am not a native speaker of English, so it will take me some time to comprehend all the information, thanks again)

If what I said doesn't make sense just ask again.  I tried to get through as much of that circuit as possible but I could have made an error or typo.   Also I have used the terms 'trigger' and re-arm interchangeably.

[POTL]

I am not a native speaker of English, so it will take me some time to comprehend all the information, thanks again)

If what I said doesn't make sense just ask again.  I tried to get through as much of that circuit as possible but I could have made an error or typo.   Also I have used the terms 'trigger' and re-arm interchangeably.

Thank you, I figured out everything, you really helped me