Modifying Pots

[Ginsly]

Typically, the answer is "just get the right pot, dingus". And I agree!

But - modifying pots has come up here and there, between reading The Secret Life Of Pots to a couple members suggesting I add parallel resistance in a couple instances.

I've successfully added series resistance to pots which limits the minimum resistance and bumps up the maximum resistance. Lowering overall maximum and/or altering the taper is more complicated, and seems to depend on how the pot is set up (rheostat vs voltage divider).

Here's a specific scenario, and I wanted to see what the solution would be, if any. Let's say I'm building a Fuzz Foundry, which has its Volume control at the end of the circuit:

The Vol control calls for a B5K pot. Let's say I only have B10K pots, but also want a more gradual increase in the first half of rotation (log taper).

My first instinct would be to add a 10K resistor to lugs 1 and 3 on the Vol pot, but I'm really not sure. Seems like sometimes resistors are added across 2 and an outer lug instead, and I've also seen both used at once - it's possible that's what I need to do, as I'm trying to both lower the overall value of the pot as well as change the taper.

In addition to Secret Life Of Pots there's a good Amplified Parts article that also talks about the effects of doing this on a rheostat vs a voltage divider, which I'm just starting to understand a little better. They're helping, but I'm just trying to make sure I'm understanding this correctly!

 

[pbrommer]

Analog Alchemy - EMH

 

[GAStan]

If you do the math at, say, every 10% of the 10k pot measuring lugs 2 & 3 with a parallel 10k resistor connected to these same lugs you get:
0k || 10k = 0k Loudest, wiper at top
1k || 10k = .91k
2k || 10k = 1.7k
3k || 10k = 2.3k
4k || 10k = 2.9k
5k || 10k = 3.3k
6k || 10k = 3.8k
7k || 10k = 4.1
8k || 10k = 4.4k
9k || 10k = 4.7k
10k || 10k = 5k Quietest, wiper at bottom

Plotting this on a graph will make it more obvious, but the first 50% from Quietest gives only 1.7k of change while the other half gives nearly twice that.

 

[Ginsly]

If you do the math at, say, every 10% of the 10k pot measuring lugs 2 & 3 with a parallel 10k resistor connected to these same lugs you get:
0k || 10k = 0k Loudest, wiper at top
1k || 10k = .91k
2k || 10k = 1.7k
3k || 10k = 2.3k
4k || 10k = 2.9k
5k || 10k = 3.3k
6k || 10k = 3.8k
7k || 10k = 4.1
8k || 10k = 4.4k
9k || 10k = 4.7k
10k || 10k = 5k Quietest, wiper at bottom

Plotting this on a graph will make it more obvious, but the first 50% from Quietest gives only 1.7k of change while the other half gives nearly twice that.

Ok, so in this case one would NOT want to strap a resistor across 1 and 3, but across 2 and 3 (Vol lug 1 goes to ground in my Fuzz Foundry example and is left alone - correct?). And if this is only to produce a log-like response, not reduce the overall max value of the pot- what if I also wanted to do that?

 

[GAStan]

Check out the calculator on the link @pbrommer provided. It uses two resistors, one to lugs 1 & 2 the other to 2 & 3. The two in series will add up to 10k in parallel with lugs 1 & 3.

 

[GAStan]

Scratch that last post

 

[GAStan]

if this is only to produce a log-like response, not reduce the overall max value of the pot- what if I also wanted to do that?

I don't believe you'll be able to keep the overall value of the pot at a fixed value using this method. The overall value is going to change as the wiper is moved any time you tie an end lug to the wiper via resistor or wire.

To the best of my knowledge about all you can do to keep the overall value at a constant is tie the appropriate resistor to the two end lugs, in this case 1 & 3, and live with the response this gives you.

Hopefully someone will be able to contradict me and I'll learn something new

 

[Ginsly]

I hate to admit it, but I'm not even sure if this Volume pot would be considered a voltage divider, rheostat or "other". Only a year into all of this, and I'm learning a bunch of things I kinda glossed over along the way.

I realize rheostats only use two lugs and typically tie the unused lug to the middle, so I assume this is considered a voltage divider..? Neither?

 

[GAStan]

In this circuit the overall pot value needs to stay the same as varying it will impact how Q3 operates.

Thinking about it, if the overall value of the pot were to change it would have an interesting effect on Q3, whether it's a desirable effect would be up to you. Might be worth experimenting if it's on a breadboard.

Definitely not a rheostat, not controlling current. I'd call it more of a voltage divider.

 

[Ginsly]

In this circuit the overall pot value needs to stay the same as varying it will impact how Q3 operates.

Thinking about it, if the overall value of the pot were to change it would have an interesting effect on Q3, whether it's a desirable effect would be up to you. Might be worth experimenting if it's on a breadboard.

Definitely not a rheostat, not controlling current. I'd call it more of a voltage divider.

Well... I kinda did! Part of the result in using a B10K over B5K is that it gets very loud very fast. Plus, 10K is just way more than it needs to be, thus the interest in reducing it. Of course, the clear solution is to use an A5K pot to solve both problems. I just thought this would be a good opportunity for me to actually learn something!

 

[GAStan]

If you have it on a breadboard you might try the B10k with a 10k resistor tied across lugs 2 & 3, then across 1 & 2 to see if either one works for you. The overal pot resistance will vary from 5k to 10k depending on wiper position. I'd be interested in the results if you do try this.

 

[Ginsly]

10K didn't do a whole lot, but putting a 1k across Volume lugs 1 and 2 definitely brought the volume down in the lower reaches of the pot. On the other hand, it resulted in a lot of hiss and oscillations in the last 5-10% of travel.

Still curious about this! If anyone has experience modding pots with parallel resistors I'd love to hear about it. @AmplifiedPartsTom would love to hear what you think!

 

[darwin999]

You might want to look at these, including their interactive tool:

https://www.tubesandmore.com/tech-articles/potentiometer-taper-modifications

https://www.tubesandmore.com/tech-articles/potentiometer-taper-modifications-interactive

 

[AmplifiedPartsTom]

In this case it would probably be best to track down an actual log 5k pot, but there are some tweaks you can make to get close to what you're looking for without one.

The 5kΩ of the pot in series with 5.1kΩ of R5 are important for biasing of the transistor here, so unfortunately it's one of the cases where modifying the effective taper of the pot by wiring a parallel resistor between lugs 1&2 or 2&3 is going to have some notable other effects (because the effective resistance between lugs 1 and 3 changes as it's rotated in that case). With a parallel resistor between lugs 1 and 2, rotating the pot clockwise will gradually (but not linearly) decrease the resistance seen between lugs 1 and 3.

With a 1k resistor wired between lugs 1 and 2, with the pot fully clockwise (full volume), the effective resistance between lugs 1 and 3 is going to be 1k || 5k which is ~833Ω. Fully counter-clockwise, it will be the full 5k. That will change the bias quite a bit as it's rotated. You can see it by using a DMM to get a voltage reading on the collector and adjusting that pot.

You can make a 10k pot work the same (or at least very nearly the same) as a 5k pot by wiring a 10k resistor parallel to lugs 1 and 3, which would make the effective resistance 5k (10k || 10k). It would not modify the taper though, which would still be linear:



You could also remove the 5.1k R5 and wire lug 3 straight to the collector. This keeps the collector resistance the same (or close enough - 10k vs 10.1k) for biasing, but it will also give you more output volume which I'm guessing isn't something you want if you want a more gradual volume increase.



If you want a log taper for volume but don't have a log 5k pot, I would replace the 5k pot with a 5.1k resistor and replace the 220k resistor with a 250k log pot:



It's not exactly the same as using a log 5k pot and the 220k resistor, but it would be close enough for me that I wouldn't think twice about it. You can also modify the taper of a pot in that location with fewer secondary effects.

 

[Ginsly]

@AmplifiedPartsTom Wow! Thank you SO much for taking the time to explain all of that!

I do actually have 5K pots, but accidentally used 10K for ALL 5 controls . Not really an issue for Stab, but seems like a bad idea to stray from 5K for Vol- although I’d probably use a log taper. I can certainly swap it out, but I thought this would be a opportunity to learn more about modding pots.

This is indeed a particularly weird circuit/control to be doing that on, considering it’s part of the biasing scheme for Q3. How odd!

In a more standard Volume control it seems like it would be more straightforward, as detailed in Secret Life Of Pots and the excellent Amplified Parts articles.

Thanks again!!