Switching between series/parallel effects using a rotary switch

[liddokun]

Hello everyone, and happy new year. 

I was wondering if some of you switching savantes could assist me, as I have always been lousy at drawing up switching schemes. 

I am looking to build a dual fuzz pedal, but give it some extra routing options.  I will call the two fuzzes A and B.  I'm looking to achieve something similar to what the Boss/JHS Angry Drive does

Using a single rotary switch, can switch between:

A>B in series
B>A in series
A only
B only
A and B in parallel
Can this be achieved with a rotary switch?  How many poles/throws would I be needing, and how would I wire this? 

I plan to use a ROG Splitter-Blend circuit to make the parallel thing work. 

Suggestions appreciated!

[rankot]

I did it this way - 4P3T switch, which facilitate A>B in series, B>A in series and A and B in parallel, so if you want A or B only, simply turn the volume on unwanted one. I've built this, BUT had a huge problem with a feedback, so be careful. I didn't fix it yet, because I was a little bit disappointed, so I left it on the shelf "for another day".

There's also my PCB, so maybe that's just the layout problem, but maybe it should also have some kind of active splitter/mixer. Not sure.

[Mark Hammer]

Though having all desired options on one rotary switch seems ideal, one of the things to keep in mind is that rotary switches can be bulky, and conceivably take up more room than their toggle counterparts.

I've built a number of dual-circuit pedals with individual on/off switches for A and B, plus an order-flipper switch.  My gut sense is that your plans are complicated by each of two elements: the desire to have order-flipping, and the desire to have series or parallel.  Eliminate either one of those and the project gets much easier.

Ranko may have the most feasible arrangement and switching scheme, though.

[liddokun]

Though having all desired options on one rotary switch seems ideal, one of the things to keep in mind is that rotary switches can be bulky, and conceivably take up more room than their toggle counterparts.

I've built a number of dual-circuit pedals with individual on/off switches for A and B, plus an order-flipper switch.  My gut sense is that your plans are complicated by each of two elements: the desire to have order-flipping, and the desire to have series or parallel.  Eliminate either one of those and the project gets much easier.

Ranko may have the most feasible arrangement and switching scheme, though.

Mark,

I've spent the better half of my morning trying to do some reading and figure out what I'll need, and I agree with this.  While what I described above is ideal, I would be fine with having a rotary to do A+B series/A+B series/ A / B as well.  The order switching can be its own toggle. 

[rankot]

I actually did this as an experiment - I hated the idea to breadboard all this, so I made a PCB. Maybe stupid decision. However, I will definitely come back to this one, SOME DAY.

[liddokun]

@rankot

When you said you had a huge problem with feedback, what exactly did you mean by this? 

[rankot]

I almost forgot what happened - but as far as I can remember, both segments had feedback almost constantly. I will have to wire it up again to see what happens, and I have also recently built a little blend circuit with op amp mixer, so I will try that one too.

[antonis]

Maybe due to Q1 & Q4 not biased Gates..??

[rankot]

Maybe due to Q1 & Q4 not biased Gates..??

Indeed, Antonis, silly me - I moved bias resistor in front of the switch, so it's "out of equation" in some positions! I will try to fix this right away! But I thought that they will be biased via VOL pots of preceding section.

[rankot]

Nope. I tried that, but it's not the cause. I have even tried to install two op amp based buffers after both channels volume pots, trying to avoid interference, but it doesn't help, and it seem sto behave still very strange. Maybe the problem with my switch?

So, this is what happens:

1. Switch to the left - input goes to PT (Professor Tweed), then from it's volume pot wiper to the buffer, then to the input of SD (Supreaux Deux), then from it's output to another buffer and finally to output jack. It shall be expected that both channels work in series, and they do.

2. Switch in the middle - input goes to both channels in parallel, then from respective outputs to their buffers, and then both buffers join at output jack. It is expected to have both channels in parallel. Sounds as expected.

3. Switch to the right - input goes to SD (Supreaux Deux), then from it's volume pot wiper to the buffer, then to the input of PT (Professor Tweed), then from it's output to another buffer and finally to output jack. Works as expected.

What I experience, is that there's really lots of feedback when running channels in series and in that case any volume at more than 50% makes feedback nosie - but that is maybe normal, due to the fact that both circuits are high gain? However, switching schematic is fine.

[antonis]

I should suggest to place a Miller cap both on Q3 & Q6 toghether with global feedback (i.e. from second series effect out to first series effect in) according to effects order switch setting..

[rankot]

There's one more thing - I have motorboating on Professor Tweed if Gain is on maximum, even if I completely disconnect another channel. I tried to figure out what's going on, even tried to add few more 100n coupling caps, but no success. Pot is A1M, but total resistance is 1M1. I found out that if resistance between wiper and CW is less than 843k, motorboating starts. It was even worse before I doubled 100uF filtering cap after voltage doubler (C23) to 220uF - the limit was 950k. I'm puzzled.

[antonis]

I'm puzzled.

Me too 'cause I don't see any schematic of power supply..

[rankot]

There it is:


Also, it seems that adding 33p Miller capacitance in front of Q6 reduce motorboating, but it's still there. Gain limit is at 384k right now. So I might add more capacitance?

[liddokun]

This is helpful to me in case I run into the same issues as you!

I'm planning to do a Harmonic Perculator and a Tonebender in one box! 

[rankot]

Then take care of not having low and high level signals (input and output) running in parallel, and use buffers (or even better) simple mixer when running channels in parallel.

Regarding my motorboating problem, there's definitely something connected to Gain of of Professor Tweed. I tried to add Miller caps almost everywhere, but it didn't help much. I could move this part of a discussion to another thread, if you guys think it's trolling.

[antonis]

When you cascode two high gain circuits, motorboating should be considered natural consequence..

After a very brute estimation, 6 stages of X10 gain (at least) result into 120dB total gain..
(more than open-loop gain of most of general purpose op-amps..)

PSRR of ultra-high gain discrete configuration always is a problem..

[rankot]

When you cascode two high gain circuits, motorboating should be considered natural consequence..

After a very brute estimation, 6 stages of X10 gain (at least) result into 120dB total gain..
(more than open-loop gain of most of general purpose op-amps..)

PSRR of ultra-high gain discrete configuration always is a problem..

Good to know! However, I managed to find the usable range for gain pot of the first channel (the one which cause MB), and after adding 360k resistor between Gain pot CW and Tone pot W, it's not existent anymore. However, its VERY feedback prone pedal, so I decided to put it into enclosure and add big red letters on it, to remind me before I start switching and turning pots, like this:

[BJM]

Don't know if it helps but some time age I built the ROG splitter/blend (well, actually a TH customs Madbox, basically the same thing) with a separate box with three rotary switches. The idea was to have two parallel effect chains, three stompboxes and the possibillty to place the effects either in chain A or chain B, so a lot of different combinations. I'm no good at this but surprisingly it worked.

Totally not what you're after but from your question, with one rotary and five possible combinations you'll need at least a 2/6 rotary switch. With four combinations you'll need a 3/4 at least. I think I used a 3/4, so I'll ry to dig up the schematic to see if make some changes. if you don't want Led switching also it may be doable.

Btw, I think the series switch is the most difficult one, but with using the splitter/blender you can have A, B, or A+B also switched on or off by switching the inputs or outputs to ground. That would be simple with one rotary, or two DPDT footswitches. I think.... someone please correct me if I'm wrong.  Don't know anything about motorboats either.....

[BJM]

Hi,

This is what I came up with,  I hope it's clear (and works....)


Btw, as mr. Hammer wrote, these switches are quite big. The plastic sealed ones without the "double decks" are the smallest, I have also seen some miniature versions of those (16 mm versus 27 mm) but not yet a 3P4T.


Good luck,