Overdriving a JFET.

[nisios]

I have been messing with a jfet stage and when i overdrive it too hard it sort of stops amplifying and then recovers from this state. It behaves like some sort of compression with a long release. What's hapening? Is is the same thing that happens with tubes and for wich people stuck a resistor in front of the plate?

[aron]

>Is is the same thing that happens with tubes and for wich people stuck a resistor in front of the plate?

Yes, similar.

[stm]

Placing a 100k or larger resistor in series with the gate usually helps to tame this situation.  Experiment with larger/smaller values until you get the desired effect.

[brett]

Hi
when the input capacitor is large (e.g. 0.1uF or greater) and the "pull down" resistor is large (e.g. 1M or greater), the time constant for any charge held on the capacitor and the JFET's gate becomes large.  (1M and 0.1uF = 0.6 seconds, 1M and 10uF = 62 seconds !) 

I almost always use a 0.01uF cap and a 470k resistor.  The 470k resistor is a good value for keeping the input impedance high enough for the pickups to be well "loaded", but not so high that extraneous noise is picked up.  In my opinion, the commonly used value of 1M is a bit too high.
cheers

[DougH]

http://www.aikenamps.com/BlockingDistortion.html

[Caferacernoc]

Once in a while I like to overboost the heck out of an input to get some of that on purpose. Sounds like Neil Young during an amp melting freakout!

[juse]

Hi
when the input capacitor is large (e.g. 0.1uF or greater) and the "pull down" resistor is large (e.g. 1M or greater), the time constant for any charge held on the capacitor and the JFET's gate becomes large.  (1M and 0.1uF = 0.6 seconds, 1M and 10uF = 62 seconds !) 

I almost always use a 0.01uF cap and a 470k resistor.  The 470k resistor is a good value for keeping the input impedance high enough for the pickups to be well "loaded", but not so high that extraneous noise is picked up.  In my opinion, the commonly used value of 1M is a bit too high.
cheers

Hi Brett,
Is this your technique for jfets only, or for all of your builds? I've often wondered about the 1m pulldown & it's affect on the signal, if it is a slight tone-sucker.

[nisios]

Thanks for all your answers.

How does charge accumulate on the coupling capacitor wich makes the time constant of importance in this effect? Im not getting that.


I decided to try a low voltage pentode (6814) instead of the jfet and it worked very well.....i got a very pleasing distortion with the tube.
The tube is a pentode wired in triode, only because i was getting more gain this way but a problem arrises: when i touch the tube i get a wine. The tube reacts alot, probably because its such a tiny tube and has the electrodes very close. Is there anything that i can do to reduce this effect? I tried it in pentode mode and i wasnt getting better results and the gain seemed a bit lower. Is there anything else i can do to help in this matter?

[brett]

Hi
the charge thing is fairly simple to describe.  Capacitors have a "resistance" (impedance) that depends on the frequency of the signal.  When the capacitor is large (ie more uF), it allows slow charging and discharging of the capacitor's two plates. When there's also a high resistance on the "far side" of a capacitor in the signal path, then the resistance of the capacitor is very low relative to the resistor, and the charge is held for longer than if the resistance is small.  The "time constant" is the time it takes for half of the charge to pass through the resistor to ground.  The capacitor and resistor form a "1-pole" filter.

About the 1M pull-down resistor.  The 1M value is high and more likely to cause bleeding ears from excessive highs than to lose tone (if that's what you mean by "tone sucking").  Most good quality classic amps had considerably lower input resistance than 1M ohms.
cheers
Brett

[bioroids]

I think the standard 1M pull down is not meant for having a 1M input impedance, because it usually goes in parallel with the impedance of the input stage.

I once used 2M2 resistors for pulldown and Vref (opamp input), so the input impedance was in fact 1M1, and yes it sounded much more trebly that the direct connection to the amp (with same length cable). So you're right, according to my calculations around 500K total input impedance is the best choice: it preserves tone, but it doesn't preserve it too much!

About that "fet compression", I experienced it too. It sounds like the pedal chokes on the signal.
I think it happens when you drive a FET hard from a low impedance source (like an opamp output). If you have several Fet stages, the output impedance of the first seems to be high enough to avoid choking the next. Beware, because although you may not be hearing an obvious compression, it may still be affecting your tone (can be good or bad, that will depend on taste).

Regards!

Miguel

[grolschie]

I noticed that fet compression with my ROG "Eighteen" when hitting it hard with humbuckers. When strumming hard the signal completely overloads and then fades/blooms in with the decay of the strings. So do you reckon inserting a resistor before the gate of the first J201 would do the trick? Thanks in advance.