G PULTEC resistors in parallel with pots to get correct values?

[daveee]

Hi,

I'm modifying my G-Pultec to make it stereo... and I'm planning on using dual-gang stepped pots, as I figure it's the best way to get symmetry between the channels.

However I can only get (affordable!) 10k and 100k linear and log stepped attenuators.

Do you think it would be OK to put a resistor in parallel with a 10k pot to make a 1k pot and also the same for the 2.2k pot? I know this is possible as it's just parallel resistors; for example, using Rparallel = 1/(1/R - 1/Rpot) ... I would need a 2.7k resistor in parallel with a 10k pot to make it 2.2k.

Or is it not as simple as this? I understand this can change the taper, which I'm not as worried about if it's not too different.

many thanks in advance

dave

 

[abbey road d enfer]

I understand this can change the taper, which I'm not as worried about if it's not too different.

Putting 2.8k in parallels with a 10k pot would make the taper hugely different.
Take the advantage of rotary switches to make them exactly the right value.

 

[daveee]

ah I thought as much.

so you mean make my own stepped pot using a rotary switch and a lot of resistors? yes I had considered that, guess it's reasonably straightforward.

thanks for your response

 

[daveee]

Just in case anyone else is wondering how to do this, I've now made my custom stepped pots using 24 position dual-gang rotary switches, by soldering resistors between the pins.

For the 2.2k pot I used 46x 1% 95.3ohm resistors (23 on each wafer), and for the 1k pot I used 43.2ohm. It works great, if I measure the resistance of both wafers at different positions, they're usually bang on, or within a few ohms.

 

[abbey road d enfer]

Just in case anyone else is wondering how to do this, I've now made my custom stepped pots using 24 position dual-gang rotary switches, by soldering resistors between the pins.

For the 2.2k pot I used 46x 1% 95.3ohm resistors (23 on each wafer), and for the 1k pot I used 43.2ohm. It works great, if I measure the resistance of both wafers at different positions, they're usually bang on, or within a few ohms.

Indeed, rotary switches and precision resistors have very good accuracy.
However, the most linear pot is probably the most linear-sounding.
Typically, in a Baxendall-type EQ using Lin pots, most of the progression will happen at both ends of travel; a somewhat better alternative is W-taper pots. In a Pultec-type EQ, both HF controls should ideally be somewhat RevLog. I guess adequate RevLog pots were not available so they opted for the closest acceptable.
The ultimate process would be to trim each resistor for the most accurate audible logarithmic progression.
This is certainly tedious, but simulation is a great help.
Manufacturers, when using pots, very seldom have the possibility to order custom-taper pots, so they have to use either linear or audio Log, whichever matches best the expectations. This generally results in progression being cramped in either end of travel.
Rotary switches offer the unique capability of providing no-compromise progression.