DIY Plexi Drive - can a guitar drive a JFET into cut off?

[Rockhorst]

I've addressed a similar issue on DIYSB months ago, but it has come up again and maybe (hopefully) someone here has some interesting views to contribute.

I've build a modded Wampler Plexi Drive for a friend a while ago. The circuit is basically three cascaded J201 stages with filtering between stages. The mods are some clipping diodes between 2nd and 3rd stage and a 'boost', which switches a bypass capacitor in/out of the first stage and a pot in series with the level pot to get a rhythm and solo setting. Here is the first stage of the effect (with the switchable cap):



Now there's two weird issues that keep coming back. Any and all ideas much appreciated.

-Circuit runs fine on 9V and 18V of a Voodoo Labs PP2+ or a OneSpot. My friend likes to run it at 12V with the PP2+. This works fine in normal mode, but when he switches to the boost (increase level and bypass cap) this induces a sort of low pitched hum, that slowly oscillates at about 1 Hz (I would describe it as 'hovering'). Sometimes it goes away for a short while, then it comes back. Like something charging up? I hoped to solve this by using a Road Rage with a 12V regulator, but no luck. No problems at 9V or 18V, just 12V (and possibly 15V).

-He likes to stack a booster in front of the Plexi. Now this is sort of a quest for more gain where no more can be found, AFAIK, but he likes it. He uses quite hot pickups in a Les Paul and it seems the high output levels can completely swipe the JFET out of operation until the signal dies of to a lower level. I have remedied this somewhat by lowering the input cap to filter out more low frequency content, but on certain setups the signal is still cut off. Can somebody explain the mechanism behind this? It's like a sudden dive bomb going to complete silence then slowly coming back up again. Why doesn't the JFET just clip the waveform off? Can a really large signal at the gate (say a few volts) upset the bias operating point of the device or something?

Thanks mucho in advance!

[Rockhorst]

Finally, a clue! After searching on the net for aeons, I stumbled across this page at Duncan's Amp page (used the wayback machine to dig up a missing image). The article discusses how a large signal can pull a current through the grid of a preamp tube, charging up the input cap and cutting off the signal. My guess is that this also happens with a JFET, does that make sense? Is it forward biased? Does it damage the JFET?

The original Plexi Drive doesn't have an input cap, so would not have this problem.

[madbean]

Have you tried a pullup resistor connected to the gate? Try a 1M @ C1/R2 connected to 9v (keep R2). This will set a constant DC on C1 which might solve the problem.

[Rockhorst]

Haven't tried that. I was more thinking of dumping the input cap and pulldown as in the original, I'm just trying to understand what's happening. It's exactly as described on Dundans page, but with a JFET. I just seems odd that nobody else ever mentions the problem

Thanks mucho for the suggestion Brian!

[stecykmi]

It's unusual that the circuit operates at 18v but not 12v. I would definitely try brian's suggestion for biasing the gate. the switching action may change the biasing of the circuit. it seems unlikely that a guitar is driving the jfet into cut off, but a high gain stage that is not attenuated (enough) can drive a 2nd stage into cut off. this is called farting out in tube amps.

Also, 1uF (C4) seems somewhat small for a DC bypass cap, have you tried a larger value? these are typically much larger, 50uf is not uncommon.

[Rockhorst]

@Stecykmi: thanks for the reply. The 1 uF is a typo, sorry for that. It's higher in real life. When I get home later today I'll do so bias checks on the jfets. The farting out happens either in the first or second stage (since after that, there's diodes to ground). It's not the amp, but the pedal (checked the pedal output with a scope). So reducing gain is the obvious solution. Let's say the second stage gets slammed too hard, does it happen analogously to the farting out in a tube amp? Or is the mechanism different?

[stecykmi]

ya, it's a very similar phenomenon, i probably did not make that clear. it's because tubes and jfets have similar biasing characteristics, with cut-off voltages that are negative.

[Rockhorst]

Thanks. For completeness, this is the Duncan link I was referring to (wasn't working).