Jfet Biasing - 9v vs 18v

[MAO]

I've read a few threads regarding biasing jfets and I'm a little confused about jfets self-biasing proportionally to their supply voltage

I'm working on a Plexi Drive variant. Switchable 9v/18v

All 3 drain trimmers are connected together at the power source, no intermediate resistor between any of the trimmers.  Jfets are not matched.

Powered with 9.3v with all 3 jfets biased at 4.65v, it sounds great.

Powered with 17.5v, the bias on all 3 jumps up to 12.75v. Well above mid point (8.75v)

Question: considering the bias for all 3 jfets increased to the same 12.75v, am I to assume that's the proper bias for the Plexi circuit with a 17.5v supply?

Makes me wonder about pedals like the JHS Superbolt which state to bias at 9vs with an 18v supply. Based on what I'm seeing, I would assume the bias would be well under 4.5v if switched to a 9v supply.  











 



   

[Frank_NH]

I used my handy, dandy JFET calculator spreadsheet for a single common source stage of the plexidrive (1K source resistor) and got the following:

* Assuming a J201 JFET (Idss = 0.64 mA, Vp = 0.83V) - For 9.3V source voltage, and a drain voltage of 4.65V (source resistor = 1K), the drain resistance is calculated to be 16577 ohms.  Then for Rs = 1K, Rd = 16577, and source voltage = 17.5V, I get a drain voltage of 12.85V, which is close to your observation (science!).  The slight discrepancy is due to the fact that your JFET's Idss and Vp are likely different than my assumptions.

Note that the calculated gain (with bypass cap) for the 12.85V bias point is about 17, whereas if you rebias at 8.75, you increase the gain to 32!  But your original 9.3 source voltage case with bias at 4.65V has a gain of 17 as well.  So, switching to 17.5V without rebiasing should sound similar.  But rebiasing should allow you to increase the gain by a factor of 2, which you may like.  Try it and let us know.  

[MAO]

Very cool Frank_NH!

Seems confirmed,  I wasn't sure as the pedal's output is louder and sharper at 18v.  

I'd like to get my hands on that spreadsheet

I did try a few lower bias settings, 10v, 11v and 11.5v. Couldn't get much lower than 10v with the 50k trimmers.  

Volume increased as the bias voltage was lowered, but the distortion seemed to loosen up too, or maybe muddy up is a bit more accurate.

I'm using humbuckers though, all that additional gain might still be good with single coils or vintage output.

11.5v seemed to be the lowest I could go without compromising the distortion.  Note: this lower bias at 17.5v supply did not sound good at 9.3v supply.

It would seem if switching between 9v and 18v, biasing at 4.5v with a 9v supply and leaving it as is when switching to 18v would be best. If just using 18v, maybe biasing between 11v and 12v would be best.    

I'll try some more tomorrow as my ears are probably burned out with all the tweaking today.  

Thanks again Frank_NH!

[Frank_NH]

"I'd like to get my hands on that spreadsheet"

Here it is:

http://guitar-fx-layouts.42897.x6.nabble.com/JFET-common-source-gain-stage-calculator-ver-2-1-tp25021.html

One of the interesting things about JFETs is that the input voltage range for the common source stage is set by Vp and not the source voltage.  So if you use a J201, you can't have a signal amplitude much above Vp ~0.8 to 1V before it starts clipping.  That may not matter at the input stage, but later stages may clip sooner with higher source voltages.  You may like this or not, which is why people advocate setting the bias by ear.

I probably would design a Plexidrive to be either 9V or 18V and set up the bias to taste.  For op amp based distortions, you don't have the same biasing issues and thus switching the source voltage from 9V to 18V can more useful (see for example the EHX Glove).