Breadboarding Davisson's JFET Vulcan

[Ben Lyman]

Sounds great at high gain but is there a better way to adjust the gain? Turning the 1M gain pot down doesn't sound right, strange undertones. I jumpered the pot to make it always full gain, still sounds great so is there another way to set up a gain pot somewhere else in a circuit like this one?
I used Gauss Markov's schematic here:
http://gaussmarkov.net/layouts/jvulcan/jvulcan-schem.png

[robthequiet]

I've been wanting to try this one too -- shot in the dark here, but maybe try a tiny cap, say 100pf, across wiper of gain pot, like a treble bleed? Might change the tone a bit. HTH

[Ben Lyman]

Thanks Rob, that's a good idea and I will try it. However, the sound I am getting isn't so much a "too much bass" sound,
it's more of an out-of-phase-octave-down-plus-dissonant-fuzziness-in-the-background  or something...
maybe I should make a sound clip. It's a great sounding circuit, right off the bat I noticed how powerful and "present" my high E string became...
it's not like it's too bright or anything, it's just really loud and clear as compared to some other distortions i've used in the past.

I think there must be another way to set up a gain pot on this, maybe something a little more traditional? Like on a FF or TS or something?

[robthequiet]

Hm, so the second stage really likes the first staged dimed, so my next wild stab would be maybe tweak the values of R6 and C5. Does it clean up better if you roll down the volume on the guitar? The bjt version has different values for everything which makes sense but instead of 100pF and 2.2M it's .0022uF and 470K which to me throttles the tone a lot less, but the guy that designed this knows way more than I.

Might take a look at ROG's Odie, they use a 250K drive pot with a 22K to ground under it, and also a 100K trim between the FET and +V rail. Something to try. Might make things even worse, but maybe worth a shot. 

http://www.home-wrecker.com/odie.html

Edit: Or note Joe Davisson's note on this page:

http://www.diystompboxes.com/analogalchemy/sch/jfetvulcan.html

[Ben Lyman]

Thanks Rob, I did mess around with C5 and R6 and it definitely helped but I didn't stop there, I think it's coming along pretty good. I still wish there was a quick fix instead of all the changes I made but here's where it's at now.
C5 1nF
R6 1M 

C7 100nF
C9 47uF

Added 100k to ground from Gain pot to prevent total shut down at lowest gain setting

Added two 1N4148 clipping diodes and 100pF at output to ground

If anyone can think of a better system for a gain pot, I will gladly return everything to normal and test it out. As it was, the gain pot wasn't really doing it for me, I wouldn't want that sound it was making at low gain, and backing the guitar volume down was even worse.

[robthequiet]

Cool. I will definitely refer back to this when I get around to my attempt 

[duck_arse]

why not - replace R10 with a 20k or 25k pot (plus some resistance to total 33k, if you must), and connect C7 to the wiper instead of the source.

[anotherjim]

When you turn the gain pot down, everything up to Q2 is floating on very high impedances as the top portion of the pot track adds in series to Q1's output impedance. This might just be making it highly receptive to noise on the breadboard & could be fine in the box.
You might make it bearable on the BB by using 470 or 500k pot for gain without losing to much signal out of Q1 or try a 1M resistor from the 1M wiper to ground.

[Ben Lyman]

Edit: Or note Joe Davisson's note on this page:

http://www.diystompboxes.com/analogalchemy/sch/jfetvulcan.html

Cool, thanks for that, I think my brain is starting to understand a little bit of this thing now. And as for Mr. Davisson's advice, "for higher gain..."  Holy cow! Not sure if it needs it but I'll try anyway! 

why not - replace R10 with a 20k or 25k pot (plus some resistance to total 33k, if you must), and connect C7 to the wiper instead of the source.

Yes! Thanks, and thanks to kipper for suggesting the same, it seems like an obvious experiment now that you point it out to me. I don't have time right now but I can't wait to try this, I hope it works. I am wondering what about this schematic makes you guys zero in on Q1 or Q2 sources for attaching a gain pot, while FF and MKII, etc would have a fuzz pot on the last stage, Q3 in this case? Anyway, thanks again, I am going to try that ASAP.

[Ben Lyman]

When you turn the gain pot down, everything up to Q2 is floating on very high impedances as the top portion of the pot track adds in series to Q1's output impedance. This might just be making it highly receptive to noise on the breadboard & could be fine in the box.
You might make it bearable on the BB by using 470 or 500k pot for gain without losing to much signal out of Q1 or try a 1M resistor from the 1M wiper to ground.

Ah, yes, I am learning more and more. Thanks Jim. You may be right about the BB noise because I know that you know a lot more than me. If I can get time I will make a demo of the noise to see if it helps decide what to do about it. I am kind of thinking it's worse than simple BB noise, it's like an octave down note doubling up with each note I play and it sounds terrible and dissonant.
The mods I listed up above seem to have solved it and it sounds great but I will try all these other experiments as soon as I can.
Thanks guys, much appreciated!

[Ben Lyman]

Here's a sample of the weird sounds. I know my breadboard is a mess but I think the undertones are from something else. Last night I had it sounding much better with the mods previously mentioned above but I returned all components to the original schematic values for this vid except for a 22k on the gain pot ground connection to prevent it from going all the way off, I hate it when gain pots do that.

[robthequiet]

I see what you mean. Sounds like you have some kind of oscillation or bleed-in down low, maybe around 80-100Hz?

That's a pretty generous input cap. I wonder if taking C1 down to maybe .047uF or .022uF makes any difference?

[Ben Lyman]

Thanks again guys. I've tried everything and the only thing that seems to have any affect on the bleeding low tone (oscillation?) is lowering the value of Q2 source cap. Too low and the effect becomes something entirely different though, so I decided maybe a 100nF poly cap is a nice compromise. I added a tone pot and some extra clipping on a switch, it's pretty cool with or without.
It's like this on my BB and I made a perf layout for it too so I think I will box it up and find out if it was a waste of time or not:

EDIT: Note that I also removed the 2x 470K divider at the output. Not sure what it was doing there so I hope it wasn't too important. Removing it brought the volume back up a bit after I added the tone control and extra diodes.

[Kipper4]

Good work doing it your way man.
R16=100k?
2 knobs 1 switch should be an easy 1590B with a battery.

[Ben Lyman]

Good work doing it your way man.
R16=100k?
2 knobs 1 switch should be an easy 1590B with a battery.

Yes, oops! and input cap is 100n, oops again!
I was messing around with it today and found a cool mod that could be put on a 2nd stomp switch. Take a jumper wire from just after the input cap and bypass Q1 and Q2, attaching the other end of the wire directly to Q3g and I get a nice low gain crunch OD without losing any volume. Then pull the jumper and it's right back to high gain distortion.
It's a fun idea but I'm not doing it to this pedal because I think it would require another LED and a bunch of layout planning that I don't want to do.

[GibsonGM]

Ben, can you audio probe and find where that 'low freq oscillation' is first audible?  Like, is it there after C4?  C7?   What if you remove the diode on the gate?

Maybe you can trace it back to where it begins. I've had things do that, too, and it is annoying.  Now I am curious why it's doing that. There is probably some rule we can learn if we find out why it's happening...

[robthequiet]

From the inventor himself:

The larger values increase the gain to more usable levels, so three should still be enough. Might need a small .1uF cap across power supply in places to curb the JFETs desire to freak out. Have fun...

Thread http://www.diystompboxes.com/smfforum/index.php?topic=22474.0

Dang those freakin' out JFETs...

[Ben Lyman]

Mike, good idea, I will try that asap before I box it up, shouldn't take long to probe it.

Rob, thanks, very interesting... I don't know what he means by "across power rail" and "in places" 

[GibsonGM]

Mike, good idea, I will try that asap before I box it up, shouldn't take long to probe it.

Rob, thanks, very interesting... I don't know what he means by "across power rail" and "in places" 

Cool, just curious...it sounds like he knew it would/could do that.

"Across the power rails" of course means between plus and minus...."in places" means in areas you might "deem to be sensitive".  Could be at your drain resistor's connection to "+", from there to ground.   Like the way you literally place the filter cap in a tube amp right at the stage it serves.

Maybe play with a .1u while you audio probe, see if you can improve that osc.    No scope, huh?

[anotherjim]

Maybe you can isolate the gate bias supply from the power rail a bit. Set up another power filter block like you have on the power input, but the series resistor can be larger ~ 1k. Then connect the 3 10M's R3,8 &13 to that instead of the main +9v. Looking at the high resistance values throughout that circuit, the FET action shouldn't wobble the +9v rail much at all - but those gates are all so sensitive - plus you have quite a bit more stray capacitive coupling between breadboard socket strips compared to a solder build which could be introducing a bit of unwanted coupling back to the gates.

Those gate diodes, I think, have the potential to be great radio detectors. All FET's could have a small cap between gate & drain - start at 22pF.