Custom Taper Potentiometers

[vintageampfixer]

What sort of custom taper are you looking for? You can modify "standard" pots just by loading them with resistors ... here's an example of something I was playing with during the week; vertical scale is the output in mV and the horizontal roughly equates to "percentage of rotational travel" ... the chart was generated in Excel just to give me a quick/dirty view and I wasn't planning to share it which is why the axes aren't fully labelled!

 

[JohnRoberts]

"Slugging" pots to bend the tapers has limited utility. The bulk resistance tolerance of common pots is like 20% so for production slugging might have to be done differently for each batch of pots.

Back in the early 80s when I was using a quad Alps pot to control a 4 pole SVF crossover filter, I had my technician slug the pot sections individually to all match each other at 50% rotation.

JR

 

[vintageampfixer]

"Slugging" pots to bend the tapers has limited utility. The bulk resistance tolerance of common pots is like 20% so for production slugging might have to be done differently for each batch of pots.

Back in the early 80s when I was using a quad Alps pot to control a 4 pole SVF crossover filter, I had my technician slug the pot sections individually to all match each other at 50% rotation.

JR

John, yes repeat no!

1. The initial question related to availability / viability of custom tapers and my point stands because source and load impedances presented will inevitably vary the pot's taper ... and by slugging, this can be further modified .... always assuming one of the 10+ "standard" tapers offered (Alps' snapshot below) isn't quite what's needed. Happy to discuss further!

2. In a similar lab to yours, albeit in cold and rainy England towards the end of the 80s, I was a similar soldier looking at much the same issue. I'll skip forwards; what we pragmatically concluded was that for the most part, most people can't actually hear the differences. In contrast, where accuracy was critical as it was in some of the instrumentation applications which we sold under an altogether different brand name, then we used tapped pots due to the commercial (im)practicality of select-on-test component insertion.

3. Pots are indeed wide tolerance which is a compound of the errors in absolute track resistance plus thermal drift, rotational accuracy, compliance to published taper and mechanical lash which in turn is further compounded by gang error, etc., etc.

4. However, back to your point about 20% - whether the pot is slugged or operating in free air, its tolerance doesn't change, thus if slugging proves an acceptable means to obtain a given taper, then there's no automatic requirement to re-adjust by pot batches because if n% tolerance is OK for the pot itself, it's a stretch to suggest we'd suddenly require (or hear!?) "greater" accuracy just because the pot's been slugged. And re-reading my point above, even a precision pot is unlikely to offer tolerance better than 10% ..

5. But why did I even begin to look at this? Here's the irony: after 30 odd years, a much loved Arcam amplifier had stopped working. The obvious fault was the volume control being physically worn out which had then caused it to fail electrically. The original Alps dual-gang, 22k Log with an 18mm long, flat shaft design isn't now available as a COTS item and the repair of a single unit didn't warrant anything non-standard. The closest I could find was a 10k pot but given the (slightly odd) Arcam circuit, this gave a very different "feel" to the volume control's operation. As I needed wanted to try and keep the feel of the control similar to that of the original, I crashed some numbers into Excel and plotted a couple of curves ... and the rest is history!

Ciao!

 

[john12ax7]

Slugging works really well to make a log pot more linear, not so much in the other direction.

 

[vintageampfixer]

Slugging works really well to make a log pot more linear, not so much in the other direction.

That's funny, I'd have said the opposite! When we consider the circuit in context, the preceding stage's Zout is rarely zero and the Zin of the following stage rarely approaches infinity ... and with that in mind, wouldn't any "load" presented by Zin drive the pot's curve towards being non-linear?

 

[JohnRoberts]

Pot tapers matter... I have long studied the subtle factors that influence how customers perceive product performance. While working at Peavey we didn't make odd tapers casually but some circuits really benefit from unconventional tapers.

The one that comes to mind is the gain resistor in the popular mic preamp topology with gain pot between two transistor emitters. To meet the dual criteria of high gain at max, and only modest gain at min is not trivial. To provide decent headroom the total resistance of our custom gain pot was something like 25k, while simultaneously providing good single digit ohms adjustability at the high gain hop off region. I don't recall all the details but the pot elements involved something like 3 or 4 resistive ink screening steps.

I am not aware of any way to deliver that slugging a standard pot taper.

JR

 

[vintageampfixer]

Pot tapers matter... I have long studied the subtle factors that influence how customers perceive product performance. While working at Peavey we didn't make odd tapers casually but some circuits really benefit from unconventional tapers.

The one that comes to mind is the gain resistor in the popular mic preamp topology with gain pot between two transistor emitters. To meet the dual criteria of high gain at max, and only modest gain at min is not trivial. To provide decent headroom the total resistance of our custom gain pot was something like 25k, while simultaneously providing good single digit ohms adjustability at the high gain hop off region. I don't recall all the details but the pot elements involved something like 3 or 4 resistive ink screening steps.

I am not aware of any way to deliver that slugging a standard pot taper.

JR

The unusual arrangement in the Arcam amp seems intended to deliver exactly the "feel" from the volume control that you talk about. It's amazing how much it is possible to influence a pot's inherent taper using external loads however these will always - can only ever - give broad changes rather than "at nine o'clock position, I want 18.3 dB of gain and then at ten o'clock position I want 33dB". I suppose to clarify, my advice here is to look at whether one can get a design "close enough" to the desired feel by slugging or whatever because it's usually cheaper than having custom products made. Peavey's actually a great case in point ... I repaired a (rather lovely old) Transtube, probably a Bandit 112 or something, and it was a PIA getting pots for that too! I was very lucky and found some NOS advertised which solved the amp issue, and sold out almost immediately. When I was at Audix, we didn't tend to have many custom items made - there were some but generally our direction was to avoid anything bespoke, esoteric or single source unless there really was no alternative - we had a lot of transformers wound, mind you ... happy days! Cheers!

 

[JohnRoberts]

The unusual arrangement in the Arcam amp seems intended to deliver exactly the "feel" from the volume control that you talk about. It's amazing how much it is possible to influence a pot's inherent taper using external loads however these will always - can only ever - give broad changes rather than "at nine o'clock position, I want 18.3 dB of gain and then at ten o'clock position I want 33dB". I suppose to clarify, my advice here is to look at whether one can get a design "close enough" to the desired feel by slugging or whatever because it's usually cheaper than having custom products made. Peavey's actually a great case in point ... I repaired a (rather lovely old) Transtube, probably a Bandit 112 or something, and it was a PIA getting pots for that too! I was very lucky and found some NOS advertised which solved the amp issue, and sold out almost immediately. When I was at Audix, we didn't tend to have many custom items made - there were some but generally our direction was to avoid anything bespoke, esoteric or single source unless there really was no alternative - we had a lot of transformers wound, mind you ... happy days! Cheers!

funny I don't think of transtubes as old, because that technology was pioneered while I was still working there (my mixer engineers shared a common lab space with the guitar engineers).

When making and selling high volume products customers have strong expectations about what should occur for specific pot rotations.

{TMI warning} another old Peavey anecdote. While I was managing mixer engineering, my senior engineer came up with a circuit improvement for a popular powered mixer. These simple top box mixers used single knob channel controls for mix and level. A common weakness of those circuits is pot kill, or audio leakage still getting into the mix when the pot is turned completely down. His improved circuit using a pot that was already in the system delivered superior kill but had the unintended consequence of being almost 2dB less gain at 12 o'clock, knob straight up. In use this is no big deal, but almost immediately I started getting complaints (from around the world) that these top box powered mixers had less power than the previous version. BUT these top boxes used the exact same power amp section as before.

The customers (and dealers) were just turning all the channels up to 12 o'clock and ASSuming the power amp in the new top box was weaker. Of course the customer is always right and I could not possibly educate every single user around the world. So the solution was to tool up a new custom pot that delivered the improved kill when fully off, and delivered the same gain as before when set at 12 o'clock. I had that luxury to do that because the SKU was selling thousands a month. A new custom pot was cheaper than answering all the customer complaints and perception of reduced power. [/TMI]

JR

 

[john12ax7]

That's funny, I'd have said the opposite! When we consider the circuit in context, the preceding stage's Zout is rarely zero and the Zin of the following stage rarely approaches infinity ... and with that in mind, wouldn't any "load" presented by Zin drive the pot's curve towards being non-linear?

You are correct in a voltage divider type scenario. But often people want a certain taper for a terminal to terminal variable resistor , eqs, gain setting, etc. I suspect this is what OP is intending. In those cases, for example, you can make a 10% log into a 30% but can't make a linear 50% into a 30%, or a 10% log into 5%.

 

[Newmarket]

tbh you have to decide what parameters are important in your application. Matching non-linear tapers is always problematic and basically comes down to someone scratching away at the printed track with a scalpel to get it within limits. Below -40dB it's pretty much anyone's guess and the important thing becomes offness.
For real accuracy you need stepped and switched resistors as in mastering kit.

 

[cyrano]

If you can't get what you need *exactly*, most kludges will do. Even for measurement devices.

No pot means you own a lovely piece of scrap.

I've heard this discussion mainly from guitar players. They expect that level 6 is the same as before, even if they always play at level 11...

 

[vintageampfixer]

I've heard this discussion mainly from guitar players. They expect that level 6 is the same as before, even if they always play at level 11...

that's only funny cos it's true :¬)

 

[abbey road d enfer]

"at nine o'clock position, I want 18.3 dB of gain and then at ten o'clock position I want 33dB".

Seems quite radical to me.
Most ordinary log pots use only two segments, so the variation between 9'o and 10'o is only about 4-5dB. This suggest a specific pot with at least 4 segments, or a negative load impedance.
Do you have the schemo of the amp?

 

[vintageampfixer]

Seems quite radical to me.
Most ordinary log pots use only two segments, so the variation between 9'o and 10'o is only about 4-5dB. This suggest a specific pot with at least 4 segments, or a negative load impedance.
Do you have the schemo of the amp?

it was meant to be ironic

 

[abbey road d enfer]

it was meant to be ironic

Well, it flew 10 miles high for me.

 

[abbey road d enfer]

I've heard this discussion mainly from guitar players. They expect that level 6 is the same as before, even if they always play at level 11...

Guitar players think 10 delivers twice the power than 5, however hard or soft they play.
I once had to play on a borrowed set, where clean sound was notoriously lacking. I didn't want to tamper with his pedal board, so I cranked up the volume knobs on the power amp. The owner immediately turned them down, explaining me that the power amp was too powerful for the speakers, so that way, his 2x90W amp delivered only 2x50.