How much clipping are cascading JFETs supposed to generate?

Started by ninjaaron, May 15, 2009, 07:17:33 PM

Previous topic - Next topic

ninjaaron

I'm in the process of trying to prototype a JFET overdrive pedal for myself. So far, It's pretty much just two J201s with a gain pot and a decoupling cap between them, very similar to the begining of the Runoffgroove Odie (http://www.runoffgroove.com/odie.html), except I'm using significantly larger caps from source to ground. There are also a few other different values, but nothing much. I'm basing it on the Odie as well as another similar circut

Anyway, I was assuming that this would be a light overdrive, but when I turn up the gain, I'm well into heavy metal territory. When the gain is lower, I can still get overdrive sounds, so it's ok, just totally not what I was expecting.

Just checking to make sure this is 'normal'...

anchovie

Have a look at the recent thread about the JFET adaptation of the Mesa Lonestar. It explains how JFETs set up like triodes will clip much quicker because of the ratio of supply voltage to signal voltage being so much smaller.
Bringing you yesterday's technology tomorrow.

ninjaaron

Ok, I think that makes sense... and you are saying this is normal?

tackleberry

Yea theres only so much before they clip then cascading 3 of em will definately clip that last 1 pretty hard.

ninjaaron

Well, I've only got two of them at the moment (though It looks to me like the last one on the Odie is wired about like a buffer, as far as I can tell).

MohiZ

I've read cascaded JFETs clip hard, and it looks that way based on my simulations as well. If you want to, try putting several cascaded stages in series, each set with lower gain.

EDIT: The Odie is not cascaded. Cascaded is the design where the FETs are connected source to drain, as in BSIAB 2 (http://gaussmarkov.net/layouts/bsiab2/bsiab2-schem.png).
I think your situation is still normal, though. The source bypass cap increase the gain tenfold. Try using smaller drain resistors.

ninjaaron

Quote from: MohiZ on May 16, 2009, 01:50:23 AM
I've read cascaded JFETs clip hard, and it looks that way based on my simulations as well. If you want to, try putting several cascaded stages in series, each set with lower gain.

EDIT: The Odie is not cascaded. Cascaded is the design where the FETs are connected source to drain, as in BSIAB 2 (http://gaussmarkov.net/layouts/bsiab2/bsiab2-schem.png).
I think your situation is still normal, though. The source bypass cap increase the gain tenfold. Try using smaller drain resistors.

Hmm... I thought cascading was just running several in a row... but I have been known to be wrong... especially when I'm a major n00b. I did mess around with source bypass caps a little...

However, once I finally a volume pot on my breadboard, I figured out that I was getting quite a bit of overdrive from my amp because I was hitting the front end so hard. After I set the volume to unity, the gain was somewhat less. There was still a lot, but not quite enough for heavy metal anymore. The range is pretty much where I want it now, from clean to hard rock, with some nice overdrive sounds in the middle.

Thanks for the advice though. Interesting stuff.

slacker

Quote from: ninjaaron on May 16, 2009, 02:56:46 AM
Hmm... I thought cascading was just running several in a row... but I have been known to be wrong... especially when I'm a major n00b.

That's generally what people mean by cascading, running a number of stages in series.

I think MohiZ is confusing this with cascOde circuits http://en.wikipedia.org/wiki/Cascode which is something completely different.

valdiorn

if it's clipping too hard for your taste put a voltage divider in front of the gate. This is actually one thing that is often forgotten when people do the Triode-to-FET conversion. Also take into account the threshold voltage of the given FET, which varies somewhat with unit to unit.

So, if you're going to use a cascade of FETs (which is quite normal, actually), just make sure to divide the signal down between stages to get multi-staged soft clipping, instead of one massive hard clip at the last stage.

ninjaaron

Quote from: valdiorn on May 16, 2009, 09:10:08 AM
if it's clipping too hard for your taste put a voltage divider in front of the gate. This is actually one thing that is often forgotten when people do the Triode-to-FET conversion. Also take into account the threshold voltage of the given FET, which varies somewhat with unit to unit.

So, if you're going to use a cascade of FETs (which is quite normal, actually), just make sure to divide the signal down between stages to get multi-staged soft clipping, instead of one massive hard clip at the last stage.
Sounds smart... Gives me a couple of interesting ideas too. To be honest, it actually sounds quite good when it's clipping hard, but it's just not what I'm going for at the moment. Maybe I'll revisit this design later, add an extra FET and a couple diodes to ground and come up with a decent 'metal' distortion.

ninjaaron

As a side note, I think the J201 I had in Q1  seems to have particularly high gain characteristics (and took a litte more voltage on the drain). I've messed around with stuff, and I put in a MPF102, and it has much less gain available (and slightly better dynamics, I think). I might try cascading two of those into a J201, hopefully to get two stages of soft clipping (and if it doesn't sound right, I'll try a J201 with a voltage divider).


If anyone wants to know what I'm doing, It's similar to the Odie, with these differences.

1. a treble bleed across the gain control
2. a pre-clipping bass cut (hopefully to become a bass control eventually)
3. possibly a little built-in roll off of very high frequencies on the ouput
4. No JFET buffer on the end, but a BJT instead, after the volume control (I want LoZ out).

It looks like the gain stage is slowly evolving into something else as well.

This is ultimately going to be one of the halves in a dual drive pedal, where the other half is a more conventional opamp+diode based overdrive/distortion circuit (I'm thinking tube reamer with optional diodes to ground). The order will be switchable, and I'm considering adding switch that selects either true bypass or a TL071 input buffer for the whole box.

MohiZ

QuoteI think MohiZ is confusing this with cascOde circuits http://en.wikipedia.org/wiki/Cascode which is something completely different.

Yes, I confused those. Sorry about that, English is not my native tongue  :icon_redface:

Caferacernoc

I've been messing with cascaded jfets and mosfets for several months now. The difference between j201 to MPF102 can be substantial. But all things being equal. 2 jfets equals overdrive, 3 is hard rock, and 4 is metal. And one is boost, of course!
Also, there is a difference in sound between having one high gain stage blast the heck out of a second one and having 3 or 4 in a row but with each attenuated back down so it's just stacked soft clipping. Two stages full blast is obviously more like a fuzz pedal and the multiple stacked soft clipping stages aremore like an classic tube amp sound.