Omnidrive switch mods

[jishnudg]

So I've been checking out the omni drive circuit...a thing of beauty,really, very elegant design, and I feel rather bad for even suggesting any modifications but here goes...what if all the switches were replaced by potentiometers..? i.e I could fade in/fade out each effect (octave, filter etc) ....
How would I go about doing this? what pot values could I use and how would I wire up the circuit?
Many thanks,
J.

[Mark Hammer]

I have to say that, as elegant and seemingly flexible as it is (and why wouldn't one want all of "that" in one package?), the sound has never impressed me all that much.  I have one, boxed up, and it sits gathering dust.

Having said that, there are really only two switched functions that lend themselves to continuous control.  Using the GEOFEX drawing as a reference ( http://www.geofex.com/PCB_layouts/Layouts/omnidrv.pdf ), you can do the following:

1) Make the gain/drive variable.  Omit R1 / C1 / R3, and the switch.  Replace R5 with a 470k resistor and a 1M pot (wired as variable resistor; i.e., two adjacent lugs only) in series.  That will get you a range of input stage gains from around 5x to around 15x.  The original goes from around 3x to 10x.

2) The scoop control can be made variable (though varying it in tandem with the tone control itself would drive me nuts).  Omit the switch and R20, and replace R19 with a 10k pot (wired as variable resistor) and a 470R resistor in series.

None of the other switched functions can be replaced with pots.  The octave is either on or off, and same for the filter.  Well, let me revise that.  You COULD connect the filter permanently, omitting the switch, and make the corner frequency variable with a dual-ganged pot.  But quite honestly, and without wanting to be dismissive, I would not recommend it to someone who asked the question you did.  Too complex at this point.

Note that, no matter what the particulars of the design, matching signals is important in getting octaving.  So, if you can get 1% resistors for the octave part (R10-14), or can match up some 10k units from your parts bin with your meter, do so.  It will reduce the likelihood of you muttering "Is this thing even on?" when switching back and forth between octave and no-octave.

Personally, I don't find much use for the switch that connects the diodes in two different ways.  It is NOT going to do TS-type clipping very well.  Best to treat the unit as a DOD250 and just connect the diode to ground directly, omitting C12.

Again, it does a LOT of things, and as such provides a good learning platform.  It just doesn't do any of them especially well.  God design for learning.  Bad design for gigging.  At least that's my take on it.

[jishnudg]

Mark,
Thank you for the very informative reply, it has actually helped me understand this circuit better.
It has always surprised me that although a lot of people on this forum have shared your view about the sound of the pedal (i.e that it's somewhat average, despite the sheer amount of variable parameters) there hasn't been any group effort on doing a massive overhaul of this pedal (or has there ? ) do you think this may be possible - - can the unit be modified enough (both in terms of parts values and circuit topologies) to make the tone more appealing, whilst still maintaining its hallmarks of relatively low parts count and high versatility?
In the meantime,I'll try out the mods you have suggested.
Thanks again,
Regards,
J.

[Mark Hammer]

I think the general reaction has been that it is okay for learning, but not as good as the individual effects, because the step from the Omnidrive to having individual circuits that are optimized for the various functions is a very small and inexpensive one.  If it was a bigger leap, there might be more enthusiasm, but if it only takes $2 more worth of parts to get a sub-component of the Omnidrive to do a particular job well, why not.

I give massive props to John Hollis for producing these minimalist designs for teaching/learning purposes, but they are minimalist - the absolute least amount you need to come within spitting distance of a particular sound.  The same thing has been true for the Zombie Chorus.  The circuit will work, as is, but once people start aiming for the things they usually expect of a chorus, the mods start to become numerous enough that it is easier (and no more costly) just to attempt a chorus design that has already attended to all those things.

All of that said, as learning exercises go, yu won't get much cheaper than the Omnidrive, so give it a whirl.  The worst thing that will happen is that you're out a couple of bucks, and a couple of hours, and you've learned a bit about what probably isn't "enough".  And if a couple of simple component swaps here and there give you something you like, it's gravy.