GGG 808 mod question: Trying to use a pot w/ two clipping options

[dap9]

I have the GGG 808 PCB and rather than using the on-off-on toggle for the clipping, I'd like to use a pot (one side LEDs, the other side MOSFETs).  I was hoping it'd be as easy as running the Clipping pads (C1 and C3 to the outer lugs of a 1k pot and then have bot lugs jumpered to the wiper).  Of course it wasn't that easy.

EDIT:  I have the LEDs in the spot as pictured inthe diagram, and the MOSFETs in where the diodes (D1 D2) are labeled.

I've been staring at the PCB diagram and schematic and can't figure out what's going on.  I've tried a variety of things including running C2 to the wiper of the pot, and that did nothing.  Here's the PCB layout:  http://www.generalguitargadgets.com/pdf/ggg_its8_lo.pdf

I put the on-off-on toggle in there and it worked fine.  So everything else is up and running properly.  I've done a few of these builds, but not w/ this mod.  Any ideas/insights would be greatly appreciated.  

[aziltz]

Here's what I would do:


I'm not positive that's the way mosfets are hooked up, but you get the picture.  Send the Free ends to C1 and C2.

Use a linear pot.  I'm not sure about value, perhaps 10k.  Maybe someone else can suggest better, but this will give you all mosfet at one end, and all LED in the other, and blend between them.

[dap9]

Thanks, man.  I was hoping not to do any off-board wiring, but maybe I'll have to go that route.  I've done it before.  Thanks for that pic!  It's a good one.  I'll have to double check the MOS orientation too, but I get what yer saying.

[aziltz]

rat shack has these round PCBs, just the right size for fixing to the bottom of a pot with some double sized foam tape.  I use them when I'm replacing an on board component with some kind of blend control, like what you are doing, or a bass cut (cap blend) or range control.

if you can see under the wiring on the POT that is on the top right, that's exactly what I did there.

[dap9]

Very cool.  I'll have to check out the Rat Shack for that.  Looks like I will have to go off board after all.  Oh well.  Thanks again

[dap9]

Well, I sorta got it working, but not quite right.  First off I've tried 1k, 10k, 100k, 250k, 500k, and 1M pots.

The higher the pot value, the brighter the LEDs get.  Anything below 100k didn't really do much - I did not get a clear distinction between the different settings.  I figured maybe there wasn't enough resistance and there was too much bleeding.  So my first question is am I right?  And why is that?

My main problem, though is at either end of the taper, I get the appropriate sound.  But as I travel towards the middle, I do not get a blend.  Instead, I get a volume boost similar to having the on-off-on switch where the off position is a diode lift.  The way I wired up is right (I believe) and uses a little technique I came up with using an op amp socket:





You'll see that I split up the 'diodes':  The Green wire on the left connects the MOSFETs to one of the outer lugs on the pot, the White Wire on the right connects the LEDs.  The Blue wire connects to the side where all four are joined and goes to the PCB.  Then I have another wire that goes from the wiper to the PCB.

Per the GGG instructions Here:  http://www.generalguitargadgets.com/pdf/ggg_its8_lo_no_clipper_sw.pdf  I've jumpered C1 and C2 therefore disabling the switch that's supposed to get wired in.  So I have my wires that go to the PCB going to D1.  I'm pretty sure this is done properly as either end of the taper sounds good, but that middle settings aren't doing the 'blend thing' at all.

Thanks in advance!

[BAARON]

You're darn right, the middle settings don't blend/clip/do what you want them to do!  Normally, when the voltage across R1 in a non-inverting op-amp feedback loop (i.e. the gain pot) passes a diode's forward voltage, the diode basically turns into an open wire with only a few ohms of resistance, which results in a unity gain copy of the signal floating on top of the diode's forward voltage (because the gain is basically R1/R2 + 1 (where R2 is the 4.7k and 0.047µF resistor and cap), and if the diode (in parallel with the gain pot) is acting as R1 and is only a few ohms, you'll basically end up with gain of 1 whenver you're over the forward voltage level of the diodes).
However, when that happens Now, you still have all the resistance from your pot being added in series with the diode, so rather than change to a gain of 1 and giving you that nice soft "clipping" we expect when you pass the forward voltage, it's calculating the gain by putting the diode+blend pot in parallel with the gain pot, and treating That as R1 in the R1/R2 + 1 formula.

By adding the resistance of the pot in series with the diodes, you're basically stopping them from doing their job properly, and your description of it as a cleanish boost matches pretty much what I suspected was going to happen with this experiment when I started stalking this thread. 

When the pot is at one of its far ends, the set of diodes that has the big resistor attached to it basically runs in parallel to your gain pot to form R1 (whenever the signal is over its forward voltage level), and the set of diodes with no resistance before it acts as clipping diodes.

I think the clipping diodes + blend pot technique is probably best saved for hard-clipping a signal to ground, because the diodes are actually clipping there, whereas diodes in a non-inverting op-amp feedback loop are wacky and don't behave like we expect "clipping" diodes to behave: they don't so much clip as much as they delicately mangle the way the op-amp amplifies the signal.  I find that many diodes sound so alike in TS circuits at high volume levels that the fine distinctions between them are lost anyway... oodles of clipping options = too many useless switches and knobs most of the time.  (I used to have a TS with 6 clipping options and a bass knob.  Now I have a TS with 2 clipping options and a simple "bass boost" switch that's set where I like it, because that's all I ended up using in the real world.)  My advice is to use the socket to find a few clipping options that you really like and stick with those ones on a toggle switch, rather than trying to find some way to dial in a million different kinds of clipping.  (For me, those options are LED//LED, and LED//2xSilicon diodes with an small extra cap to smooth out the highs more.)  Your audience won't notice the subtle distinctions anyway - only the big changes.

[dap9]

Well that's a kick in the dick!  I drilled the hole for the pot already.  Guess I'll have to figure out a way to put the switch in so it's not all loosey goosey.

Thanks a lot for that explanation!  It's somewhat over my head, but I'll print it out and go over it a few hundred times until it makes sense, lol.  I have a rough idea what you said which is good.  Guess I'm moving out of the noob phase (ever so slowly, lol).

Thanks again.

[aziltz]

that's weird.

i think the KOT circuit uses resistance in series with the clipping diodes, and I was basing things off of that.  I also thought lower would be better.  Did you get anything different out of the 1k?

[dap9]

that's weird.

i think the KOT circuit uses resistance in series with the clipping diodes, and I was basing things off of that.  I also thought lower would be better.  Did you get anything different out of the 1k?

No, the 1k, 5k, and 10k all had pretty much no effect whatsoever.  So I skipped to the 100k.  From there on up, I got decent effects on either extreme, but also the aforementioned problem.

I did this mod w/ a Klone clone and it works well.  I have a 1k in there and need to try some different values b/c there's a very limited sweet spot and I'm wondering if a 5k or 10k may be better.