Fuzz Face Bias Brainteaser (resistors in parallel)

[italianguy63]

A Monday brain-teaser.  Maybe hard, maybe not.

I tend to like to experiment, and ask questions.  I hope you don't mind.  It seems some people like answering said questions.
Anyway, this one is concerned with simple Fuzz Face Q2 biasing and variable parallel resistor(s).  This one may be easy with some kind of simulator (which I don't have or use).  It has a lot of variables, but I think an answer is attainable.  BTW let's try to keep the answers in scope!

Problem:  (part 1)  A basic classic PNP FF circuit uses a 8.2K resistor as a bias resistor on the collector leg of Q2; generally it is varied to seek the magic voltage range of 4.5 to 5.0V.  I have built several FF's, and my experience has found a range for this resistor of around 5.0-8.2K depending on the gain of the transistors in question.
 
(Part 2)  I have seen several schematics substituting the 8.2K resistor either with at trim pot or a user manipulated "bias" pot, or a variable pot and with a fixed resistor in series.  Usually this is a 10K pot, but I have seen 15K, and even 20K.  So, what is the real desired range that is wanted?   (A guess), I think 0 ohms is too little and 15K (maybe even 10K) is too much.  What is a real USEFUL range?  Maybe 4K to 10K???  Can it be reasonably calculated?

Solution:  (and beginning of the discussion) (Part 3) Solutions sometimes cause more problems sometimes.  Anyway, I started drawing a FF board for prototyping.  I thought it would be cool to have the ability to use either a trim pot OR a user manipulated bias control depending on the build.  So I drew a circuit with BOTH components in parallel.  You just solder in which one you want.  Trim-pot or potentiometer?  Cool.  Maybe.  This got me thinking about Part 2 again.  What should the range be?  A user controlled pot should be limited to be useful and not potentially damaging to the transistors.

OK—so it looks like this (I can post a schematic if you need it) but I think you can get it.  2 variable resistors in parallel; with the wiper and one leg combined on each.  What happens if you populate BOTH components?  You get 2 resistors in parallel ranging in values of 0 to 10K each.  If I remember my theory right, the max combined value would be 5K (and not enough).

THE QUESTION:  What should the range be?  Assuming we have R1 as a fixed value, and R2 as a 10K-Alpha variable pot.  What value could R1 be to satisfy Part 2?  If 10K is not a useable value, what is easily attainable as a substitute?  Can someone simulate this or graph it?
Thoughts?!
MC

[digi2t]

Just spit-balling here, but how about an internal DIP switch? One setting to the properly biased resistance, and the other to allow the external pot to be tweaked by the user.

As for range, I would assume whatever range that would produce usable results. I believe that would be dependant on the transistors used, gain, leakage, etc.

[nocentelli]

Pots in parallel give strange tapers: If you've identified 5-10k resistance as being the sweetspot for most fuzzface bias requirements, you would assume two 20k pots (or one pot and one trimmer) would be ideal. Set the unused trimmer/pot to max (20k), and adjust with the other, giving a sweep of 0-10k. However, a pot with a fixed resistance in parallel gives non-linear sweep making setting the exact resistance harder than with just a single linear pot. I'd suggest using a fixed resistor of 2-5k (whatever you think is the absolute minimum resistance you would want) in series with a 10k pot or trimmer: The fixed resistor stops you accidentally giving the transistor the full 9volts.

[italianguy63]

That's kind of where I am-- thinking somewhere in the range of 1K to 4K in series... then the 10K with the other value in parallel.. Just wondering if anybody has tools to graph the tapers or has data to support what the values "should be."   MC

[nocentelli]

If your pot is 10k, your maximum resistance with a resistor in parallel is 9.9999999999..... k. Parallel resistor calculators are handy, I use this one -

http://www.electronics2000.co.uk/calc/series-parallel-resistor-calculator.php

As for the ideal tapering resistor value, i've got a great picture which shows how the "anti-logness" of the combined parallel values varies depending on the difference between the pot and fixed resistor value: I'll post it in a minute when I can get to the pc.

Edit: The below thread has some good information, the image I was looking for is on page 2 -

http://www.diystompboxes.com/smfforum/index.php?topic=70732.0

[italianguy63]

I saw that.  Very cool.  My original idea was to try to solve a problem by using some unused board real-estate to add a component.  Now it looks like 2 are needed, and the solution is not as easy as I hoped.  I would still love to hear if anyone has had experience on what the sweet-spot of the resistor value is.  Most of mine turn out in the 6Kish range.  So, to give it upper and lower range go down for 2K from there, and up 2K from the original spec.  So, 4K to 10K?  Only a guess though.

Bueller?  Bueller?

[italianguy63]

i.e. the best answer may be a 3K9 fixed resistor in series with a 5K-A pot (instead of the widely published 10K value).  MC