The Problem with Parallel Pots

[MoltenVoltage]

I need to run 4 pots in parallel as voltage dividers, but since each pot will act as a parallel resistor for the other 3, the taper of each pot is significantly altered as discussed here:
http://www.diystompboxes.com/smfforum/index.php?topic=70732.0

Does anyone know a way to connect 4 parallel pots as voltage dividers (to power, ground, and output the voltage at the wiper) but avoid the taper alteration?

Thanks!

[Earthscum]

I've been battling at this for some time now and haven't come up with anything better than large extremes. If you take a 1M pot, for example, and strap a 10k resistor across it, the major amount of taper doesn't happen until the end of the rotation. If you are only using a small center portion of the pots' rotations, then this may work. Basically go look up a parallel resistor calc and just start punching numbers and it quickly becomes apparent what's going on.

ATM, I'm actually working on, possibly, a similar problem with LDR's. I have pics up in the Lounge of a Woody Wah I'm working on. I used over/under dual LDR's, so the trap exposes the bottom LDR first, then the top. Resistances go from about 500k at heel down to about 20k at a quarter open, down to about 10k. The top LDR goes from >20M, 16M at quarter open, and down to about 500k at half c o c k e d. From there the top LDR jumps down to the 20k range real fast. I am working on translating that into something a bit smoother using parallel resistors. But, no matter what I do I will always have that taper in there (I could invert the taper with a transistor, I guess.)

And that is where I am on that problem... no solution, just move the tapers to the extremes and work in between, unfortunately.

[Suicufnoc]

What are they being used for?  In some situations I you could use buffers to isolate the pots, or maybe even just caps.

[Earthscum]

you know, I was just thinking... depending on the setup, there may not be a taper problem with it. For instance, if you have 100k/100k, no matter where the wiper is, it's still a set percentage of the overall resistance from pos to gnd. If you are using the pot as a variable resistor (center and lug tied together), then there is a taper as the resistances shift. At half rotation, in that case, you would have 100k/50k. That's where the taper problem comes in. If you aren't varying the resistances, there should be no taper, just a percentage of total resistances from that lug, for the most part (assuming no 'loading' on the lugs, i.e. buffered, or otherwise high impedance).

[PRR]

> 4 parallel pots as voltage dividers (to power, ground, and output the voltage at the wiper) but avoid the taper alteration?

From Power to Ground? Like this?



There is NO interaction. In theory the power rail is SOLID 9V to ground.

[MoltenVoltage]

This is the circuit:



There is some interaction because when I measured the resistance across the outer lugs of a pot alone I got 5K, and when I put two in parallel I got 2.5K.  Then when I measured the resistance between each side lug and the wiper, each value was lower with a parallel pot connected (but oddly the sum the lug to wiper resistances was above 2.5K but lower than 5K with a parallel pot connected).

I also assume that since the resistance is lower, more current will flow through each wiper.  Is that not the case?

Thanks again!

[earthtonesaudio]

If you take the battery away of course there will be interaction.  But that's not the circuit.  Side note: don't measure resistances where current is flowing, it's bad for your meter.

With one 5k pot across the battery, 1.8mA flows through both the battery and the 5k pot. 
With FOUR 5k pots, 1.8mA flows through EACH pot so the battery must provide 7.2mA.

Notice how the current through any pot is not affected by the addition of more pots, hence the voltage divider taper at each wiper is unaffected by the addition of more pots.

[PRR]

> a pot alone I got 5K, and when I put two in parallel I got 2.5K.

The battery can pull a 5K, 2.5K, even 1.25K load.

Hook it up as shown, with a battery.

Measure the Voltage from Out1 to ground, while you turn Pot2, Pot3, Pot 4.

As shown, turning "other pots" has no effect on the pot you are measuring.

If you use taper-resistors (as shown on R.G.'s page) AND use a very sick battery, there will be some slight interaction. Fix: use healthy battery, appropriate pot values (100 ohm would be too-small for 9V block battery).

> more current will flow through each wiper

As shown, current in the wiper is zero.

In testing, wiper current is the loading current of your voltmeter. For a DVM measuring say 3V, about 0.3uA or 0.000,3mA.

In real life, you need to specify what is connected to the wiper. If a naked opamp buffer, maybe 30pA. If a 12V 50mA light-bulb, could jump to 30mA near the top of the rotation (and may burn-up the pot track).

[R.G.]

In real life, you need to specify what is connected ...

Precisely, Paul.

[rustypinto]

Don't forget a capacitor on each wiper. I suspect that will cut down on the "interaction" problem, but then again, we don't know what they are connected to.

[MoltenVoltage]

The issue came up because I am working on putting switchable expression pedal jacks on 4 different pots that feed a microcontroller.

I was thinking about how to wire the switching jacks and remembered the old thread about pot tapering and assumed that I couldn't just jumper the Sleeve and Ring (ground and power) from the switching side to the non-switching side because that would alter the taper of the pot since the outer lugs of the permanent pot would remain connected in parallel.

What happened is I remembered the old thread wrong, which is not about Potentiometer taper, but about Rheostat taper alteration.

A Potentiometer is a voltage divider that changes the Potential (voltage) coming out.

A Rheostat is a variable resistor that change the resistance (and therefore current) coming out, and whose taper is affected by a parallel resistor.

The only problem I see with jumpering the Sleeve and Ring on the switching jacks is that there will be excess current flowing which will tax and heat up the voltage regulator (and waste electricity), so it makes more sense to switch either the Sleeve or Ring to eliminate that "vampire" current.

In any event, thanks again for pulling me out of the rabbit hole!