Making linear pot logharitmic

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

rotation

Well-known member
Joined
Jan 24, 2006
Messages
402
Location
slovenia
Hi,

i'm adding 1k pot as attenuator at the output of transformer balanced preamp (600ohm). I only have linear pot at hand and no time to wait for logarithmic pot.
I found this article about making log pot out of lin:
http://sound.westhost.com/project01.htm
It says that ratio between pot's and resistor resistance must be 6:1 to 10:1.
Is this correct, can i use something like 100R resistor with 1k pot?
I don't care if i can't get nice log curve, i would be happy if it's at least a but similar.

Miha
 
Do a google search on this.  I had to do it once, and found tons of info.
The Neve 8068 AUX module has 10K log pots made by soldering 8K2 between CCW and wiper of a 10K lin.
Experiment in and out of circuit.
Mike
 
The only problem with this is that, when the pot is raised, the load it presents to the source becomes lower and lower. At max output, the load would be 90 ohms, which the output stage may not agree with.
 
Unfortunately, the very last part of it is utterly wrong.
There's no way you can turn a lin pot in a Log or Revlog variable resistor, whatever you try with "padding" and "trim" resistors.
 
This quickly turns into a (you can't get) something for nothing exercise. Typical pots are 20% bulk resistance. These various schemes look good on paper and are an interesting intellectual exercise, but in practice are not very accurate or reliable.

JR
 
Sorry to disagree, but all my parametric EQ's use lin pots with law-bending resistors for frequency control. The resulting law is remarkably linear (well, logarithmic!) over a 1:10 range. Very nice 20, 25, 31, 40, 50, 63, 80, 100, 125, 160 and 200 ticks at 30° increments.
 
Opinions vary.

I recall back in the '80s using a good quality (alps) 4 gang pot in a 24 dB/oct LR active crossover, and despite using the pots as intended without any tricks, I still had to have a technician measure and tweak the pots (by adding a shunt resistor, at 50% rotation, to get acceptable (to me) overall tracking.

In my sundry parametric designs IIRC I used dual linear or variants. One I recall was 50% at the middle and had some curvature at both extremes (S taper?).  I always had trouble holding decent frequency accuracy, especially if I tried to give the pot sweep too much frequency range. x10 range switches helped a lot in that regard, but that was a luxury I didn't often use. While customers rarely notice if the pot sections don't track, Q can vary some with frequency settings.

I found my best performance for parametric frequency control in a state variable (virtual earth integrator stage) was using the pot so that it was completely out of the circuit at full up and full down. Using a real resistor in parallel with the pot meant that I could be 5% accurate (or whatever R I used in parallel) at the LF extreme, and likewise got the same tolerance as the wiper resistor at the HF extreme. As long as the wiper resistor was high relative to the pot's nominal resistance, it would track the taper, and not be shifted by the resistor loading and interacting with the 20% tolerance pot. 

Maybe I was just too picky, or had bad luck with pots. I thought the alps were pretty good at the time.

JR

PS I could tell stories about some cheap pots I ran into while working at Peavey, and quite humorously (to me) one well know pot brand, decided to no longer bid for our business on one pot part number because "we were too picky".  ;D  FWIW that pickiness was not related to their pot's tolerance.
 
 
Back
Top