My parallel distortion box

[ianmgull]

So...

I've been reading up on parallel mixing and love the idea of being able to run multiple distortions simultaneously. I want to build a box with 3 or 4 distortion circuits in it as well as a boost. I plan on using on of these:

http://www.diystompboxes.com/smfforum/index.php?topic=67323.0

I've never done anything like this before so I was hoping to get some idea's. Right now I know I'm going to put my GGG Tubescreamer as well as my UglyFace in there. I know I want some sort of high gain circuit (BSIAB maybe?). Something that still has a decent amount of picking dynamics. Any thoughts on this?

I'm also wondering what sort of clean boost to throw in. I'd like to be able to switch between having the boost either pre-splitter/post mixer. I want to be able to blend the clean in, as well as a master bypass switch for the whole box.

This is definitely my most ambitious project to date but I think I'm about ready for it. I'm wondering if I would need a polarity switch. Does the tubescreamer invert it's output? Any advice anyone could throw at me would be GREATLY appreciated.

[GibsonGM]

Pretty ambitious box you have going, let us know how you're making out!   
Looks to me like the input buffer and output stage of the TS is in a Common Collector setup, non-inverting...each opamp is in the noninverting config., so signal stays noninverted...signal in and signal out should be the same phase (or close to it)....

[petemoore]

GGG Tubescreamer as well as my UglyFace in there.
  UF has a 555timer in it, that could make ugly spikes appear in the TS signal path, perhaps there's a way to turn it off or shunt the input to ground when bypassed.

[ianmgull]

Hey Pete, I was thinking about something similar actually. I'm looking at JD Sleeps parallelyzer:

http://www.generalguitargadgets.com/index.php?option=com_content&task=view&id=67&Itemid=26

And I was curious as to why the switching would be at the input and not closer to the end. Wouldn't it yield cleaner results to switch at the output? That way when I'm not using a particular effect, NOTHING from it's signal path meets the mixer at the end of the circuit?

What exactly do you mean about spikes in the TS signal path? If everything is running in parallel, how would the Ugly Face's signal end up in the TS path?

[ianmgull]

Ok so I'm etching a big muff pcb as we speak. I'm leaning towards the Rams head...sounds nice and chunky. I hate to keep bringing light to my noob status but does anyone know if the big muff inverts it's output?

Thanks,

ian

[ianmgull]

Tubescreamer and UglyFace complete! Big Muff and MicroAmp half way done!!!! Enclosure is on the way!  Can you tell how excited I am?

[tehfunk]

sounds good man, can't wait to see the finished product. 

[petemoore]

I hate to keep bringing light to my noob status but does anyone know if the big muff inverts it's output?
  Each transistor gain stage inverts the signal once.
  that's means it's inverted 4 times, and comes out...
  Non Inverted.

[ianmgull]

Ok I'm somewhat confused now... Assuming I throw in a TubeScreamer (non-inverting), Big Muff (inverted 4 times, non-inverted in the end), Uglyface (shouldn't matter), and clean signal do I need some sort of phase inversion circuit? The reason I'm asking is because even though the big muff sends out a non inverted signal, won't it still be four cycles behind the clean signal? I would imagine this would cause some sort of comb filtering. Or am I completely missing the point here.

As always, thanks for the advice everyone.

ian

[ianmgull]

Ok. So here is the update. I drew up a schematic based on the GGG IC buffer and Simple mixer. I've never done this before so If someone more knowledgeable would give me some feedback it would be greatly appreciated. I didn't include any phase switching (still wondering if it's necessary). I also left out the switches. I am going to put a Master bypass for the whole box in. Also bypass for the boost (which will stay at the end) and each individual channel (still wondering if I should do that at the beginning or right before the mix pots.

I would appriciate any advice available!!!!!! Thanks!!!!!!   

ian

[ianmgull]

So I'm starting to put everything in the box and I'm wondering where to put the switches for each of the five channels. I'm just using a spdt so the switch will just ground the signal when it is bypassed. I'm wondering if the switches for each channel should go after the buffer and before the effect OR after each effect. I was thinking after because I plan on putting a fuzz factory in there and thought that even if the input is grounded (switch before the effect), it might still oscillate and be audible. Any suggestions for switch placement?

[Processaurus]

After the effect, definitely, otherwise you get all the self noise from 4 or 5 distortions.  Also you may want to decouple & star ground the power supply for the uglyface as it uses a logic chip and may put noise on the power rails.  A 100 ohm resistor and a big cap to ground should do it.

If it were my project, I might think about putting 2 sets of volume controls after each effect, in parallel, each set summed together with their own summing resistors, so you could have 2 different mixes available on a spdt footswitch, if that makes sense.  An individual output for each effect too would be awesome.

[ianmgull]

So I have another question about switching. Assume I'm using a SPDT for switching each channel. I put the switch after each effect board (before mixer obviously). I run the signal from the effect to the middle (common) lug of the switch, one lug to the mixer, and the third lug to ground. Will this work???

The reason I wonder is I figure as soon as I send one channel to ground, wouldn't all of the channels then be grounded? Or do the buffers act as some sort of "force field" which protects the other channels from grounding when one is grounded.

Does this make sense? It just never occurred to me until now that this might be a problem. 

[ianmgull]

Oh, and for those interested, here is a crappy camera phone pic of my progress. (I'll get nice pics when it's done!)

[Processaurus]

So I have another question about switching. Assume I'm using a SPDT for switching each channel. I put the switch after each effect board (before mixer obviously). I run the signal from the effect to the middle (common) lug of the switch, one lug to the mixer, and the third lug to ground. Will this work???

The reason I wonder is I figure as soon as I send one channel to ground, wouldn't all of the channels then be grounded? Or do the buffers act as some sort of "force field" which protects the other channels from grounding when one is grounded.

Does this make sense? It just never occurred to me until now that this might be a problem. 

Hi, you don't want to send the effects outputs to ground generally, as it makes the output have to source as much current as it can, as it is shorted out.  Basically makes it work hard for no good reason.  Much better to ground the input of the mixer, duplicating what the circuit looks like if the volume control is all the way down (to ground).  Flip the switch around, put the cap and summing resistor to your opamp summing stage on the common of the switch, the effect output (wiper of each's volume pot) on one throw, and ground on the other.  Grounding here will not affect the other channels, because the inverting opamp makes a virtual ground on its inverting input.  Weird concept, but it makes sense, as an opamp works by using the output to try to keep the 2 inputs at the same voltage.  So here, it has the 4.5v on the non-inverting input, OK, and when audio comes in on the inverting input, the output works to put out what will be the exact opposite signal,  enough to cancel it out, and get that pin back to the steady 4.5v, once it makes its way back through the feedback resistor to the inverting input.  Say you make the feedback resistor bigger, the ouput has to swing bigger (ie louder) in order to get the same nulling to 4.5v on that inverting input.  Kind of a funny trick to pull on that opamp, like the mule and the carrot on a stick.

[Gus]

Two things that might be issues at the input section.

What is the input resistance as drawn?

Do you want to change the 1/2 power supply splitting to another design?

Things to think about

Why so many 2.2megs?

Make a 1/2 supply divider like in a TS etc two 22Ks and a cap from center node to ground

A 1 meg from the divider node to + input of an opamp buffer and your input cap connected to the + input,  NOW this will have the output of the new opamp at about 1/2 supply, use this output to drive the opamps setup as buffers before the effects no need for caps or 1/2 supply resistors between (Ask yourself do you still need caps after if the effect after the buffer has a cap coupled input).
  You might want add series resistors of 100 ohm to 1K before each of the parallel opamps + inputs.  One more opamp but a reduction of resistors and caps.

If you don't understand this post ask

[Processaurus]

I wasn't going to say anything because he had it in a box and it will work fine, but from an economical  design perspective, yeah, you could do the input with only one set of biasing resistors, like the JFET splitter at AMZ (made me realize the high high impedance of JFETs, including JFET input opamps, means you can fan a properly biased input out to a ton of them, without loading).  But if it works, why fix it?

http://www.muzique.com/lab/splitter.htm

Do you want to change the 1/2 power supply splitting to another design?

Did you mean the R22, R24 voltage divider?  That should be fine, right, as the only loads are the two noninverting inputs of the opamps?

You might want add series resistors of 100 ohm to 1K before each of the parallel opamps + inputs. 

Do you mean the input section, where the input fans out to the 5 opamp buffers? Not to make a habit of contradicting you    but while that wouldn't hurt anything, it shouldn't be necessary.  Nothing happens on the non-inverting input pin to the signal, so there would be no interaction between the 5 opamps' inputs.

[MicFarlow77]

Hey Ian,

That enclosure is really looking good! Great Job! I cannot wait to see it painted up!

Thanks,

Mick

[ianmgull]

Thanks for the tips everyone. As far as the buffers are concerned I thought about trying to do the whole buffer section on one board. Problem is I SUCK at pcb layouts and couldn't find a 5 way buffer layout. I ended up using the single op-amp buffer from GGG. This is why I have the redundancy of the bias resistors in the schematic. Space is getting tight though so I'm wondering if it might be better to try to whip up a layout using dual op-amps. I imagine it wouldn't be to much bigger than what I'm using now with the 5 independent ones.

Ben,

I see what you are saying about putting the switching later on in the circuit. I already have the mixer section all wired up in the box so I'm wondering; Would putting the switch after the mixers cap and summing resistor effect anything aside from the current draw of each individual channel? I'm all for being green, I just know I would have to heavily modify the mixer pcb which is already in place.

Gus,

Does what you were saying relate to input impedance? If so I have no idea what the answer to your question is!    Really I'm just pulling whatever circuit fragments I can find and trying my hardest to fit it all in one box.

Mick,

Thank YOU!  I may be in the market for a couple more of these boxes in the near future if you have the time to whip them up ( I know you're probably busy with the wah enclosures).


Thanks all!

ian

[Processaurus]

Ben,

I see what you are saying about putting the switching later on in the circuit. I already have the mixer section all wired up in the box so I'm wondering; Would putting the switch after the mixers cap and summing resistor effect anything aside from the current draw of each individual channel? I'm all for being green, I just know I would have to heavily modify the mixer pcb which is already in place.

Hi, are you saying that your switches are already wired, and all of them are hooked up after the summing resistors, on the inverting input of IC1B?  If so, that won't work, as all the effects are connected directly to that point. Turn one off and they would all be off.  The best place for the switch is between the pot and the cap leading to the summing resistor and mixer.  Example, for the clean signal, between pot1 and C11.

Why so many 2.2megs?

Oh, I just noticed what you meant, there are 5  2.2M pulldown resistors connected to the input jack.  There should only be one, to drain any minute DC leaking back through the input caps.  Actually, Ian, if you haven't made the buffers yet, you can get rid of all of the input components up to the opamps, except for one set (one pulldown, one input cap, two biasing resistors).  Then connect all 5 of the buffer opamps' + inputs together, like orman does with the JFETs in his 4 way  JFET splitter in that link.