NOT biasing JFETs

[marcelomd]

Hi,

I found a bunch of J201s in my parts box and I was wondering what to do with them.

90% of the circuits using JFETs for distortion suggest using a trimpot in the drain to bias them to whatever value. But some, like Wampler's Plexi drive, just stick a resistor there.

What's the tradeoff? Is Wampler hand picking JFETs? Do they accept a large variation in tones from unit to unit?

Thanks!

[FSFX]

Hi,

I found a bunch of J201s in my parts box and I was wondering what to do with them.

90% of the circuits using JFETs for distortion suggest using a trimpot in the drain to bias them to whatever value. But some, like Wampler's Plexi drive, just stick a resistor there.

What's the tradeoff? Is Wampler hand picking JFETs? Do they accept a large variation in tones from unit to unit?

Thanks!

I don't like the idea of a trimpot in the drain of JFETs. I would rather use one in the source to bias them.
Anyway, if the JFET is used as a simple source follower buffer then with devices like J201s you just need to bias the gate to a few volts or even up to half the supply to ensure that it accepts a large signal swing. With buffers using JFETs with a larger pinch off voltage of around -2 to -4 volts then a simple resistor from gate to ground is all that is required for a source follower buffer.

If you are using it in a common source type configuration then it is a good idea to select JFETs with the appropriate pinch off voltage, Vgs(off) and saturation current (Idss) for the resistor values or measure Vgs(off) and Idss and calculate the appropriate source and drain resistors to give the type of operation you require. The biasing for the 'Fetzer' type distortion is different to that for a more linear operation. There are JFET biasing calculators available on the Internet to help. Also there are simple circuits that allow you to measure the pinch off voltage and saturation current.

This is one of the best and most versatile JFET bias calculators that I have come across.
This link is to a copy of it on my website but I claim no credit for it at all.
It was created and posted somewhere by Frank_NH in 2015.
I found it somewhere and did some minor changes to it, but that is all. All credits go to Frank_NH for a great calculator tool.

http://www.fleetingspider.com/files/JFET_Calc.xlsx

[idy]

Pretty sure they are using sorted FETs in commercial pedals. The drain trimmer is "problematic" for reasons above me, and gives you "acceptable" results with "unsorted" FETs, but the results do not sound like the pedal you are cloning. You may like it, you may like it better, but the RunOffGroove/Fetzer trimmer is not the way "serious" engineers or business people do it.

[marcelomd]

Examples of what I'm refering to. Should've posted in the first message.





So what do you do with all the JFETs that aren't "good enough"?

[FSFX]

Examples of what I'm refering to.

A drain resistor of 15k and source resistor of 1k in Wampler's Plexi Drive should work pretty well with a J201 with typical pinch off and Idss values.
The following screenshot shows how these values bias a typical J201 using the calculator spreadsheet I posted a link to earlier.

[Rob Strand]

The drain resistor works fine but if the circuit requires JFETs with VP = 0.5V and you plug in JFETs with VP=2.0V then you have to expect the circuit won't work exactly like it did before (as intended).    If you stick to the same part often the drain resistor is good enough.

While there's plenty of schematics around for various JFET pedals there's a lack of info on the bias voltages and the JFET parameters of the *actual* pedal.    With a wide range of pot adjustment it not even possible to guess.   So even if you know how to throw electronics ninja maneuvers at the circuit you don't know what the target is!

This old post surveys making different circuit adjustments to set the bias point, then analyses the effect on the circuit.
https://www.diystompboxes.com/smfforum/index.php?topic=119707.0

It will take a bit to get your head around the results.  It's hard for people to process multiple device parameters varying never mind compensating for them.

Circuits with bypassed and unbypassed source resistors are affected differently.   The gain of the bypassed source case will generally be more affected by different JFETs.

[marcelomd]

... So even if you know how to throw electronics ninja maneuvers at the circuit you don't know what the target is! ...

That's more or less what I want to understand. I do get that you can change your bias point by using different components around the transistor.

The question is how much of a difference does it make on the final sound. Is it a meaningful difference or something that only tone sommeliers can appreciate? I want to avoid having to tune each device if I can. Also, trimpots are too big.

Now this is my nerd side: If the biasing is important to the final sound, how do large (-ish) scale manufacturers do it? "Is this pedal crap or did I get one with bad bias?"

[antonis]

Now this is my nerd side: If the biasing is important to the final sound, how do large (-ish) scale manufacturers do it?

By (trying to) design circuits almost "immune" from bias variations..

[marcelomd]

Now this is my nerd side: If the biasing is important to the final sound, how do large (-ish) scale manufacturers do it?

By (trying to) design circuits almost "immune" from bias variations..

Any pointers? Wampler's doesn't look any different from Runoffgroove apart from the trimpot.

[antonis]

Any pointers?

https://www.vishay.com/docs/70595/70595.pdf

[marcelomd]

Any pointers?

https://www.vishay.com/docs/70595/70595.pdf

Wow! That was quick. Thanks!

[FSFX]

Any pointers?

To design for JFETs you need to know the two basic parameters that effect biasing, that is the drain saturation current, Idss, and the pinch off voltage, that is Vg(off).

They can easily be measured using a few resistors, a 9 battery and a DMM. There are many circuits on the web showing how.

Below is a simple one I use.

Then if you use the spreadsheet that I posted a link to above, you can work out the best resistor values to use for drain and source.

This explanation of JFET biasing may help you understand.



JFET TESTER

[FSFX]

Wow! That was quick. Thanks!

This is also a good write up by Rod Elliott about JFET biasing.

https://sound-au.com/articles/jfet-design.htm

[antonis]

I think OP's query is the way for massive JFETs implementation into a specific circuit..
(no shorting, no individual measurements, e.t.c.)

[marcelomd]

I think OP's query is the way for massive JFETs implementation into a specific circuit..
(no shorting, no individual measurements, e.t.c.)

That's the idea. Use mass production techniques so I don't have to go through the hassle.

[FSFX]

Well not everyone just takes what they buy or have without verifying them. There are just too many fakes around.
JFETs are quick and easy to test with simple testers or things like the DCA75.
I have hundreds of J201s and other JFETs and I know all their characteristics.
Batch testing of SMD parts is quick and easy so no problems using them in bulk for production.

[Frank_NH]

Glad my old spreadsheet is getting some use. 

There are some other ways of biasing jfets without trimmers.  I came across the following circuit from 1975 while studying this topic a while back (the NPN transistor is incorrectly specified as 2N3906):



Reference: https://worldradiohistory.com/BOOKSHELF-ARH/Technology/Encyclopedia-of-Electronic-Circuits-Vol.5.pdf#search=%22jfet%20bias%22

I simulated this current source biasing arrangement in LTSpice and built it on the breadboard and it works (I used +/- 9V from an LT1054 charge pump).  You can get more gain my adjusting the drain resistor.  It seems to be tolerant of different jfets (i.e. you can use an MPF102 or a J201 and it will bias OK).  But the downside is that you need a split supply and some extra parts (the NPN transistor, zener diode).  Anyhow, something to think about...

[Clint Eastwood]

That's the idea. Use mass production techniques so I don't have to go through the hassle.

What you want is like going to the Mcdrive and ask for haute quisine. You are not going to get it.

It really isn't such a hassle, just the next step in your diy experience. I find cooking with jfets very gratifying.

[FSFX]

Glad my old spreadsheet is getting some use. 

Frank_NH, thanks for doing that spreadsheet. I have shared it with a lot of pedal builders on the FB groups.

It is a really nice JFET calculator and it looks and works far better than a lot of the online Javascript ones.

I use it a lot for checking designs and comparing results against LTspice simulations with all the JFET models I use.

I just added a couple more example JFETs and a bit of text to try to stop people entering data in cells that are calculated ones.

[antonis]

Although schematic is good, description is rather confusing (especially for noobs..)..
(current source consists of n-p-n BJT and relevant circuitry with its Norton equivalent output resistance bypassed via 10μF cap for MPF102 CS amp to have some gain..)