Potentiometer values don't add up

Question

I'm working on a project that involves swapping out a couple of pots of a circuit board for a couple of digital potentiomters. I'm trying to make sure that I'm using the right resistance values, so I'm measuring the resistance of the old pots. These are standard 3-terminal pots. I'm calling the terminals A, WIPER, and B.For the first one, these are the values I'm getting:

A to B: 4.84k

CENTER:
A to WIPER: 1.66k
B to WIPER: 3.55k
            -----
            5.21k

FORWARD DIRECTION:
A to WIPER: 2.333k
B to WIPER: 2.785k
            -----
            5.118k

BACKWARDS DIRECTION:
A to WIPER: 1.00k
B to WIPER: 4.16k
            -----
            5.16k

I had always understood that the resistance from A to WIPER + the resistance from WIPER to B should add up to equal the resistance from A to B, but no matter how many times I measure the A to B resistance is not 5.2k.

Obviously there is something that I'm not understanding correctly. What is it?

Thanks!

Answer 1

This is expected behavior.

So I rummaged around my old parts bin and found similar behavior on most old potentiometers.

There wasn't have a big enough sample of newer ones to see a difference.

Possible reasons for this behavior:

• Oxidation on the resistor track and / or wiper. There was one potentiometer where the resistance sum changed wildly over one sweep, up to twice the nominal value, but became better with exercise.

• Construction: The resistor itself is a flat band that the sweeping contact travels across. If the contact does not reach across the whole width of the band, or is not correctly aligned, the resistances will not add up.

One of the more consistent potentiometers gave its life for geekness: There were two sweeping contacts, each about one quarter width of the resistor, one traveling along the inner edge, one traveling along the middle. This means, when you measure across the resistor, you measure all along the whole width of the resistor. On the dot of the sweeper, you measure only a tapering part of the resistor, which does not have as much resistor in parallel.

Trying to visualize:

• A-B resistance is like this: o========o;
• To the wiper it is like this: o==><=====o.
• It is the taper of the current distribution that leads to the increase in resistance.

Do I care about this? No. First, we only care about the relation of the two resistances. Second, we are dealing with 10% accuracy devices. Third, if precise control over the parameter was required, one would have chosen a different tool.

Ms. Ohm and Ms. Kirchhoff can safely put away their husband's heart medicine.

Answer 2

In general, the resistance of a pot will have three components: the resistance from the contact point to the clockwise end, the resistance from the contact point to the counter-clockwise end, and the resistance from the contact point to the wiper terminal. The first two resistances almost always add up to the resistance between the two ends of the pot, and if one doesn't draw any current through the wiper, the voltage between the wiper terminal and one of the end terminals will be a suitable ratio of the voltage between the two ends. That is by far the preferred way to use the pot. If there is no current flowing through the wiper terminal, its resistance won't matter.

If the circuit requires drawing a non-trivial amount of current through the wiper terminal, things will generally get quite a bit nastier. Unlike the first two resistances I mentioned, which nicely vary as the pot is turned, the resistance between the contact point and the wiper is arbitrarily variable. Sometimes it may be 5 ohms, sometimes 500. On older pots, it may even get to be 50,000. Different spots on a potentiometer may have widely varying amounts of contact resistance, depending upon how the pot has been used and low long it has sat in different positions. If one is going to use a pot to generate a resistance rather than a ratiometric voltage, one should design one's circuit to use the largest value pot one can practically manage (the Atari 2600 Video Computer System's "paddle" controllers, for example, use 1-meg pots wired as rheostats). Unless the pots are clean, the "jitter" caused by varying wiper resistance will cause the game to respond somewhat erratically to small amounts of rotation, but unless the controls haven't been cleaned in ages it usually won't be too bad. Had the console's circuitry been designed for lower-resistance pots, though, even small amounts of wiper resistance variation would have been intolerable.