Multi-Flanger from Japanese stompbox book

[StephenGiles]

Dare I ask if the trigger levels follows input dynamics?

[armdnrdy]

Dare I ask if the trigger levels follows input dynamics?

You wish!

Hey Stephen,
I never got around to bread boarding your circuit. That weekend turned upside down and I didn't get much of anything that I wanted to accomplish done! Never enough hours in the day...or days in the week! 

[anotherjim]

Output of IC1a and IC3b are effectively driving into each other via the regen path. You might get away with that if you put a resistor immediately after IC1a, otherwise I'd thought the low output impedance of IC1a is going to pull the regen signal  - effectively gating it.

BBD paths usually have just under unity gain, this looks no different. So if you want Regen to hit feedback, you'll need some gain too.

So a simple 2 input inverting opamp mixer before the BBD input filter is what I'm suggesting - like your IC6b but with 47k as the Regen input resistor so it's got gain enough for feedback.

[armdnrdy]

Okay,

I see and understand what you are getting at. I missed the no gain issue.

The original board layout (which leaves a bit to be desired) has an actual jumper (for no apparent reason) between IC1A and R5 (33K)

If the regen circuit were to connect to that point, it would just need a single pad. So...There must have been more going on than a standard regen circuit consisting of a pot, trimmer, and a couple of caps.

For the likes of me I do not understand why the designer would put every bell and whistle on this thing and leave the regen circuit as an "option". 

[Fender3D]

...You might get away with that if you put a resistor immediately after IC1a...

You may just connect the FB path @ R5/R6/C4 node...
But I think you'll need smaller value resistors for R74 and TR8...

I added a footswitch (SW2) to choose between straight flanger and Envelope/Triggered flanger....

Then shouldn't SW2 better select between ENV/TRIG mod and LFO (ala Mutron), (not just deactivate ENV/TRIG)?..

...It seems that when the Envelope/Trigger switch (SW1) is in the Envelope position, the trigger, which includes the Attack and Release controls, is still connected to the mixing node (R47-R50) through the trigger sens. control. (P6)
My question is: are the Attack and Release controls still active when SW1 is in the Envelope position?

They are...
Just notice that P6 and P7 voltages are INVERTED each other, I guess it "might" work as ENV/TRIG's upper limit-lower limit, otherwise it "might" correct what Mark said about the envelope control:

...when it comes to downward sweep, it is not simply the mathematical "opposite" of upward...

BTW
is R39 strictly necessary?

[anotherjim]

Think kindly of the designer. They may have wanted to keep options open with the Regen until a later stage of development.
The easiest and most common way is to blend regen with the incoming clean signal with some kind of mixer. That usually needs an inverting opamp for a decent virtual earth mixing circuit. That will also give an opportunity to add in some more low pass filtering which could be handy if there's still some clock noise making it's way to the delay output - and you wouldn't want to regenerate that.
It could also provide an opportunity to insert companding noise reduction, if it was found necessary, or maybe just a soft clip limiter to tame the howl at feedback.

[armdnrdy]

Hey Federico,
Thanks for the input!

You may just connect the FB path @ R5/R6/C4 node...
But I think you'll need smaller value resistors for R74 and TR8...

I agree with Jim. I think that it would be wise to add an op amp with gain and dial it back with the regen trimmer. In every circuit that I've looked at for reference, the feedback section is taken from a point with gain.

I added a footswitch (SW2) to choose between straight flanger and Envelope/Triggered flanger....

Then shouldn't SW2 better select between ENV/TRIG mod and LFO (ala Mutron), (not just deactivate ENV/TRIG)?..

I think that the envelope CV reacts with the LFO to create a certain effect...
but...You got me thinking to switch between straight flanger and Envelope/Triggered flanger with a footswitch, and add a toggle switch to cut or add the LFO when the footswitch is in the Envelope/Triggered flanger position. That configuration will give three options, with two available on a footswitch.

Just notice that P6 and P7 voltages are INVERTED each other, I guess it "might" work as ENV/TRIG's upper limit-lower limit, otherwise it "might" correct what Mark said about the envelope control:

I looked up "envelope invert" on the net and read that some ADSRs come with the ability to invert the wave shape. I think that when the trigger or envelope is passed through IC7B the wave shape is inverted for a different effect. When P7 is CCW, the wave passes through the 100K mixing node. (R47-R49) When P7 is CW, the wave is inverted. The original name for P7 on the original drawing is Inv. Sens (Invert sensitivity) Does this sound correct?

BTW
is R39 strictly necessary?

R39 in parallel with P3 looks kind of redundant when P3 is CCW but, it might serve some purpose when P3 is CW.
P3 is connected to the input signal via R35 & R17, to IC 5 non inverting input, and to the mixing node. (R47-R49)
It's hard for me to tell if it serves a purpose.
I double checked the original board layout....and that's how it is all connected.  

[armdnrdy]

Think kindly of the designer. They may have wanted to keep options open with the Regen until a later stage of development.

This doesn't seem like a work in progress....it was published in a Japanese Effector book back in the eighties. No disrespect to the designer.....I think that this is an intriguing design. Did you design this Jim? If so...I apologize! 

I was busy with business related matters for most of the day. I'll add an inverting mixer for the regen circuit and repost the schematic.
I appreciate your ideas and feedback. You've been very helpful.

[anotherjim]

Hah! No, wasn't me. That modulation section alone would have put me in the funny farm 

[armdnrdy]

Here is a quick preliminary revision with the regen input mixer and a few op amps moved around. The op amps will find a final location while routing the board.

With this mixing configuration it seems as if the regen gain will change with the adjustment of the regen trim (TR8)

Is this correct?

If that is the case, I'm wondering if it wouldn't be better to connect the regen gain op amp (IC4B) at the output of IC3B.

[armdnrdy]

I just had another thought...

What if the regen circuit were connected as I originally had it (drawing on first page of this thread) and added gain to IC3B then lowered the gain resistors in the two output op amps? (IC6A & IC6B)

If this will work with no unintended consequences...this would be the easiest solution.

[Fender3D]

C'mon Larry...
R74 AND R78?

The only potential issue (but it might be as well a feature) is the feedback Phase inversion...

I just had another thought...

What if the regen circuit were connected as I originally had it (drawing on first page of this thread) and added gain to IC3B then lowered the gain resistors in the two output op amps? (IC6A & IC6B)

If this will work with no unintended consequences...this would be the easiest solution.

+1
I'd try that way, , first...
It will save 1 op-amp (and 2 resistors if you listen to me  )

[armdnrdy]

C'mon Larry...
R74 AND R78?

I know, I know!    That's why I'm here....to learn! I just cut around what I originally had and was waiting for you to reprimand me!  

I just had another thought...

What if the regen circuit were connected as I originally had it (drawing on first page of this thread) and added gain to IC3B then lowered the gain resistors in the two output op amps? (IC6A & IC6B)

+1
I'd try that way, , first...
It will save 1 op-amp (and 2 resistors if you listen to me   )


This configuration is similar to standard regen circuits that I've referenced. The regen circuit is connected to an output with gain and then dropped in after the first op amp.

[anotherjim]

I'm going to say MK2 is just fine. But you don't need R74 (100k). I think the trim will be handy.
Input impedance - Do you want 47k? It could be 1M for direct guitar.

In mk1, you had regen feeding into the output of IC1a which isn't going to be nice -  the incoming clean signal will gate the regen. You have to have a mixer.
Polarity inversion of the regen and incoming clean to the delay isn't a problem -  the delay line is removing any fixed phase relationship. You only need to maintain true polarity of the dry signal path.

Regen source could be either of the post filter amps. It will probably "dull" out the regen faster from the last filter. Maybe a jumper option here so you can choose? There are designs that take regen before the post filter (from the BBD null), presumably depending on the pre-filter to clean it up. That could be another jumper option?

The Blend control isn't really a blend is it? It looks to me that it adjusts the dry level to a fake stereo anti-phase output, and you can switch phases turning the blend and dry off is at centre. There seems to be nothing to adjust the delay output level itself.

If you don't want the hassle of that 2 gang centre tap pot or can do without the fake stereo, you can save one opamp there. Ic4a as a unity gain inverter (needed to maintain dry path polarity) feeding a Dry output level control to IC6a and output of IC3b feeding a Delay level control to the other input of IC6a.

If it the intention to replicate the original design, then please ignore my ramblings   You could always leave room to attach a daughter board, fit jumpers for now as per original, and maybe the original "regen circuit" may come to light?

[armdnrdy]

Hey Jim,

Thanks for the help.

I haven't got around to input impedance or pull downs yet. I usually redraw the schematic, make modifications, and then put on the finishing touches when I have something that is a bit more concrete.

I'm going back and forth with this regen circuit! As far as the original regen circuit coming to light....I have searched all over the net to find either the missing pages or a copy of the original Japanese book....no luck! What was scanned and posted on the net is all I can find.

I actually sourced the dual 50KB center tapped pot so....I was going to build that section per the original drawing.

I'll make some revisions and repost the "work in progress" drawing.

[anotherjim]

It seems like the original was meant for use with synth. I can imagine some fun could be had twiddling the blend control while holding a string pad chord.

[armdnrdy]

It seems like the original was meant for use with synth. I can imagine some fun could be had twiddling the blend control while holding a string pad chord.

Yes sir.
You can tell by the Gain control that I'm going to delete, the Peak level circuit that's already gone, and by the oversized (for guitar) caps in the signal path.

Here's the drawing getting a bit closer to being finalized.

[StephenGiles]

Impossible to read drawing on my Blackberry! Won't focus for some reason.

[armdnrdy]

Impossible to read drawing on my Blackberry! Won't focus for some reason.

Resized JPEG with reduced quality.

Try this PDF:
https://dl.dropboxusercontent.com/u/53299166/DIYstompboxes/Multi-Flanger.pdf

[StephenGiles]

Impossible to read drawing on my Blackberry! Won't focus for some reason.

Resized JPEG with reduced quality.

Try this PDF:
https://dl.dropboxusercontent.com/u/53299166/DIYstompboxes/Multi-Flanger.pdf

Thanks Larry, much better on my PC - that is a superb drawing. I don't think mobile phones were made to read schematics!!