Question about the Hollis Omnidrive and Digital Potentiometers

[wayland]

Hi all! 

I'm looking at building a Hollis-style Omnidrive pedal (schematic links below), but with digital potentiometers.  My main problem is that it contains a 470kΩ potentiometer, and unfortunately I can't find any digital pots between 250kΩ and 1MΩ.  Does anyone have any thoughts about whether I could use one of those (or 100kΩ would be even better) if I altered some of the other components around the pot? 

http://www.hollis.co.uk/john/circuits.html
http://www.geofex.com/PCB_layouts/Layouts/omnidrv.pdf

[antonis]

You probably refer to Gain pot (470k)..

I should suggest to double its value (1M, if availiabe) rather than lowering it down to 100k..
Make C6 value 22-27pF, R16 2k2 and C7 4.7μF..

If you insist on 100K Gain pot, make C6 220pF, R16 220R and C7 47μF..

[Ripthorn]

Since it appears to be just a variable resistor, you could use 1M and put a 1M resistor across the terminals. This would give 500k at max, with less influence as the pot resistance goes down.

[wayland]

Yes, I'm referring to the Gain pot, but forgot to mention that; thanks! 

Well, you guys seem to be unanimous at least in referring me to the larger pot.  Slightly annoying, but good advice, I'm sure.  Thanks!  I'm still in the parts-buying stage, so definitely worth knowing at this point.  Unfortunately, the 1MΩ digital pot is also SMD, which I haven't done before.  But can't be avoided, I guess, if I want this to be useful. 

Wouldn't putting a 1MΩ resistor in parallel change the taper of the pot (is that bad though?)?  And in this scenario, wouldn't I be better off with an 82kΩ resistor, giving me 450kΩ  at max?  Or is it bad to be less than the 470? 

So, I like the simplicity of Ripthorn's solution, but Antonis, your solution was very educational; from studying your solution, I can see what needs to change in which directions to account for the change I'm looking at. 

My choice regarding which of your solutions to go with will depend on feedback I get regarding the taper of the pot (ie. whether that's good or bad).  I got the impression from online reading that, in order to make the pot act like a log, the thing to do was add a 1/5 resistor in parallel.  (ie. if I wanted a 470k pot to be log, I'd aim for a 96 ohm resistor in parallel). 

So, thanks (ευχαριστω!) to both of you! 

[antonis]

Wouldn't putting a 1MΩ resistor in parallel change the taper of the pot (is that bad though?)?  And in this scenario, wouldn't I be better off with an 82kΩ resistor, giving me 450kΩ  at max?  Or is it bad to be less than the 470?

You're overthinking (not bad but overkill for such a project..)
But, yes to taper alteration.. (although you may not even notice it..)

How comes 82K//470k -> 450K..?? 

470k set max gain to 471 (1 + 470k/1k).. Ripthorm's suggestion raise it up to 501..
(if only pot set to min could be indeed 0 Ohms.. - rare situation for commercialy available pots..)

So, thanks (ευχαριστω!) to both of you!

You're welcome (παρακαλούμε!)

[wayland]

Apologies: 820kΩ, not 82kΩ. 

So does that mean that your suggestion would give me a gain of ~1000?

ευχαριτω παλιν (warning: my Greek is from doing a year of Biblical Greek 10 years ago, so could be inaccurate and/or outdated). 

[antonis]

So does that mean that your suggestion would give me a gain of ~1000?

Can't figure out the way you've resulted into it.. 
Both my suggestions (either for 1M or for 100k pot) are followed by respective R16 value altering just to maintain original circuit gain..
(well, the closer enough for resistors E12 series values)

P.S.
Your Greek is by far more modern than mine..

[wayland]

That was just me getting over-excited.  Half of my original electronics knowledge comes from physics, and the other have from "Dick Smith Fun Way into Electronics" kits, so while I have some idea what individual parts do, I'm not very good with how they affect each other.  I'd forgotten your other changes. 

Another question on this circuit; is there something I should alter that controls the smoothness of the knee on the clipping? 

Hmm.  Now I'm wondering whether I should be including ϝ if I say οιδα :p

[antonis]

Nothing to alter..

Place a cap of 1nF, say, across diodes pair..

P.S.
Would you prefer a modified schematic full of ancient greek symbols and numberings..??

[ElectricDruid]

Personally, if it was me, I'd reduce the gain pot to an easily-available 100K, and then scale the other components accordingly to keep everything else the same. C6 would have to go up by x4.7, but 47pF is pretty tiny already anyway, so 220pF makes more sense.
R16/1K could go down to 220R if you really need that much gain. C7 is 10u and gives a bottom end -3dB of 15Hz with the 1K, so you might not even notice if that moved up to 80Hz. If you do, change it for 22u, or decide that the gain's a bit much and use 2K2 for R16 instead.

[wayland]

Regarding the capacitor, great!  Thanks!  I had a feeling a capacitor would be invoved, but wasn't quite sure how . 

Regarding the modified diagram, I'm actually considering drawing one of those myself, but since the Greek symbol would be Ω, that's nothing much new there :p . 

I want to be able to play my bass through this (I do bass as well as regular guitar), so I'm hesitatant about the bass rolloff; thanks very much for the advice on alternative components. 

I'm also interested to hear from Antonis why he thought that the 100k pot an inferior option. 

[antonis]

I'm also interested to hear from Antonis why he thought that the 100k pot an inferior option.

Just because, by lowering pot value you'd have to also lower R16 (scaling for maintain equal gain) resulting into a C7 big electro..(for maintain same HPF corner frequency)

The opposite, of course, deals with elevated noise level & unwanted distortion but we aren't dealing with Hi-End builds, aree..??

[wayland]

Oh, OK.  Not 100% sure I followed the last sentence; by going with the 100K pot, I'd keep down noise level and unwanted distortion, but would have to use a larger capacitor? 

Regarding high-end builds, I'm not sure what I want to do about it at this point; I'm actually considering (probably heresy on this forum) doing the design (up to and including the PCB design) and then having someone else build it for me.  Where by "design", I'm thinking also of the controller for the digital pots.  Then I'd have to program it afterwards. 

[antonis]

Oh, OK.  Not 100% sure I followed the last sentence; by going with the 100K pot, I'd keep down noise level and unwanted distortion, but would have to use a larger capacitor?

Yeapp..

[wayland]

OK.  That actually sounds like a pretty good option, especially when pricing is factored in.  Well, I think that's all my questions about the Gain part of the circuit.  Thanks very much!  I think I need to go back and like all of the posts in this thread; they've been great!

[wayland]

No, I've discovered another substitution question: Can I replace the TL072s with an MCP6004? 

[11-90-an]

The supply voltage of mcp6004 is 1.8-5.5V and the tl072 is 7-36V... would add a lot of complications to your circuit... would make more sense to use 1 TL074 instead

[antonis]

No..!!
(for several resons..)

MCP6004 is a quad op-amp.. It might substitute TL074, but see below..

Gain bandwidth product is only 1MHz
Slew rate only 0.6 V/μS
Input bias current up to 170μA
(all the above back to 741 days..)

Supply rail voltage up only to 6V..

Secondary drawbacks are: PSRR 86dB, CMRR 76dB..

[wayland]

Hmm.  My pedal already has a voltage regulator with a 3.3V output on it anyway (for the digital pots), so that's not necessarily a drawback.  But that 1MHz is a good point.  OK.  I like to use MCP chips where I can because they're readable, but you've convinced me here; thanks! 

I don't even know what some of those other things are (looks like I've got some reading to do, but not now; it's 1am ). 

[ElectricDruid]

If the pots are going to be digital pots running at 3.3V, then using 3.3V op-amps might make sense, because the digipots can't cope with signals outside their supply range. And if everything else is on 9V in a gain pedal, that's going to be pretty likely.

But there might be a bit of redesign involved. At this point, it becomes truly yours, since it'll have little left except "inspired by John Hollis' Omnidrive".