# Potentiometers: Mechanical & Electronic, Part 1

I am far from an expert on the use of potentiometers and writing this blog for a site like Planet Analog is a prime example of “leading with my chin.” I expect any number of corrections, but since no one else seems to have decided to tackle this lowly subject and since I have worked quite a bit with pots, I thought I would try to put something on paper.

As I understand it there are two different reasons for using an adjustable resistor. In the first we are trying to control a device through its complete range of operation. Such a device might be a lamp dimmer and typically these devices tend to be high wattage. Modern design tends to frown on this simple solution, preferring additional circuitry to improve efficiency, simplify operation, and add features. In the second approach we (or at least I) design a circuit using standard parts and then improve the circuit's performance by using a potentiometer (pot) to trim the circuit values.

Pedantically, a device that acts as a voltage divider is called a potentiometer and is a three terminal device. The pin that allows for the variable resistance is called the wiper which is named for the way the mechanical connection is made to the fixed resistor. A variable resistor is a rheostat and is a two terminal device. With mechanical pots I am not sure if there ever was a pure rheostat, but with the advent of electronic pots it is possible to realize one. Of course it is very easy to configure a potentiometer into a rheostat as shown in Figure 1. Having said all of that, it is rare to hear the word rheostat used nowadays, being almost always being replaced by pot, potentiometer, or even trimmer.

Figure 1

Two methods of configuring a pot as a rheostat. At first glance they are identical, but the configuration of (b) keeps a current flow even if the wiper of the pot goes open circuit (a not uncommon event).

When trying to create a trimmer circuit, it is not normally a good idea to use the setup of Figure 2 (a) for use in the rheostat mode because the thermal characteristic of the pot is normally several orders of magnitude worse than a fixed resistor even for an expensive multi-turn pot. In Figure 2(b) the value of the pot is normally much smaller than the fixed resistor and so its thermal performance has a much reduced effect on the overall resistance.

Figure 2

How to improve the thermal performance of a rheostat configure potentiometer.

As a potentiometer you might be tempted to use the arrangement in Figure 3a. It may be cheap, but the resolution of 1% is difficult even with multi-turn pots and the temperature drift will probably preclude precision design. Figure 3b is an improvement since the overall percentage error contributed by the pot is much less provided its value is much less than the fixed resistors. However the thermal variation of the end-to-end value of the potentiometer will affect the output.

I understand that this parameter is far worse than the ratiometric drift. In addition, the value of the pot tends to be relatively small and may be difficult to obtain. I have only seen figure 3c in a 1987 Maxim seminar (the source of the whole of this block of information), and I think it is worthy of a repeat performance. The absolute value of the trimmer has no effect on the output, and large pot values can be used. Also the resolution of the adjustment is maximized to the center of the range.

Figure 3

Potentiometer mode. In ©, the initial value of the resistive divider R1/R2 is set to the expected value. To set the value of R3 consider at one end of the range of Rp, the minimum voltage will be the resistive divider created by R1 and R3||R2. Conversely the maximum will be the resistive divider of R1 ||R3 and R2.

Now what happens if you want to adjust the gain of an amplifier by adjusting a pot? There are several application notes that deal with this in far more detail than I can in this blog. I would recommend AN1316 from Microchip, Tutorial 864 from Maxim, and Circuit Note CN-0112 from Analog Devices. All these refer to electronic pots, but the deliberations are the same for mechanical ones working with the specifications for the device you want to use.

Here I am at the end of Part 1 and I am still trying to figure out where to insert the fact that the resistance value adjustment (the taper) can be either linear or logarithmic. Logarithmic pots are often used in audio volume control circuits and are sometimes called audio pots. Generally you will know from your application which taper to use.

In Part 1 of this blog I have tried to treat the potentiometer as a generic device. In Part 2 I will move on to tackling the different types and the advantages of both mechanical and electronic pots.

## 12 comments on “Potentiometers: Mechanical & Electronic, Part 1”

1. eafpres
August 11, 2014

Hi Aubrey–I always love your posts becuase they are straightforward and contain clear examples.  I for one will give you a pat on the back and not a jab to your chin.

Your article reminded me of a long time ago when I was not yet a teenager.  We had electronics around the house, and I did a project once where I somewhere obtained a couple of speaker drivers, and decided I would add them as a remote output from our “audio system” which was all contained in a nice piece of furniture which had fold out doors, lift up lids for the turntable, and built in speakers.  These features give you a hint as to how long ago that was.

Well my dad (who knew plenty about electronics) pointed out to me that the impedance of the speakers I was using was not the same as the output of the tube amp that was the heart of the Hi-Fi.  Somehow I got it into my head that if I could match the impedance I would get better sound in my room.  I found an old pot, and more or less tried to create a purely resistive matching circuit.  I spent hours tweakig that potentiometer trying to get more sound and less hum.  I have to laugh at my lack of fundamentals, but that was my introduction to a tuning pot way back in the '60s.

2. eafpres
August 11, 2014

Hi Aubrey–your mention of volume control pots for audio reminded me of a problem of mechnical potentiometers.  This was especially true of cheaper audio equipment.  In many cases the range of operation of the pot was a relatively smll fraction of the possible range.  The follower would wear down the windings and you could start to get mechnical hysteresis near the edges of the most used range.  In very cheap stereos you could feel this in the friction of the knob.

3. antedeluvian
August 12, 2014

Blaine

I for one will give you a pat on the back and not a jab to your chin.

Thanks for you kind words.

Your article reminded me of a long time ago when I was not yet a teenager.  We had electronics around the house, and I did a project once where I somewhere obtained a couple of speaker drivers, and decided I would add them as a remote output from our “audio system” which was all contained in a nice piece of furniture which had fold out doors, lift up lids for the turntable, and built in speakers.  These features give you a hint as to how long ago that was.

I wonder how many of us on Planet Analog got into electronics because of (or at least as an early indication) music and the attempts to record it. Unlike you I had no expertise to call on and living in a technical backwater in deepest darkest Africa did not help. I started out recording with a microphone and when I hear certain oldies (like “Have I the right?” by the Honeycombs) I can still hear my kid brother screeching. After that I tried recoding by connecting across the speaker of the radio (there was no audio socket) and into the mic input (early reel to reel- only had a mic input) with varying degrees of success.

The tape recorder was replaced and the new one came with an audio cable, and that worked OK (again I thing across the speaker, no stereo yet) until the cable broke. Hand made replacements nver worked well becasue it seeems there was some kind of impedance matching withing the cable. Of course the noise from the potentiometers always found its way through as well.

Finally I caught up with technology when both the radio/record player and the tape recorder had matching input/outputs.

4. samicksha
August 12, 2014

I remember potentiometers were used to control picture brightness, contrast, and colour response but just wanted to check how well they work in motion controls systems.

5. David Maciel Silva
August 12, 2014

Very good article AnteDeluvian. Some electronic issues never go out of style lol! Pot is certainly a component of many applications. I confess that I've burned several in one setting, in my time in high school. This configuration is very common since operational amplifiers connected to a high input impedance to current flowing through the POT is very small ….

6. GSKrasle
August 13, 2014

How curious! Last night, I was playing my favourite video-game ('Excel') with a model of your circuit (c), which is a bit more complicated than you'd think. The additional resistors change the function Vo/Vi vs. X (X being the wiper index) from simple linear to a ratio of polynomials. I was trying to approximate a perfect LN function by manipulating R1,2,3 (where R4+R5=Rp=const). I got real close using the 'goal-seek' utility, though I suspect using circular references and iteration could be used too.

I HOPE my algebra is correct: anybody else care to try their hand at the five-resistor circuit? It's harder than it looks!

The motivation is to approximate the behaviour of an expensive 'logarithmic' or 'audio-taper' pot with a cheap linear one. (Cheap 'log' pots are poor in their approximation, and, particularly, in their matching.) Remember that, because 'most everything biologicalous is logarithmic or exponential, volume-control, brightness, etc. is always best-handled logarithmically.

Wikipedia has a pretty good reference on pots, and there's this: http://www.bourns.com/data/global/pdfs/OnlinePotentiometerHandbook.pdf

But let me add some practical caveats re using them in the real-world:

Don't ignore the wattage ratings, and remember that if you are using only half the resistive element, the rating is half as large! Watch what happens at the extremes!

The extremes of the adjustment range are usually messed-up. The wiper might jump to a short to the end-contact, it might reach a limiting value that's not 'zero,' or it might go open or flakey. Take that into account! (This is one reason to connect the wiper to one end.) Many pots that have been used for experiments have damaged elements at the extremes (see above). Often, you can FEEL the messed-up-ness.

Pots are expensive. Specialler ones are generally expensiver, exceptions being certain ones used in consumer gear, including some 'tapped' ones, which, incidentally, offer a way to do thermal compensation.

7. antedeluvian
August 13, 2014

GSKrasle

anybody else care to try their hand at the five-resistor circuit? It's harder than it looks!

Given my history with Excel, I have regularly thought I should develop a model of this circuit, and every time I have decided it is too complicated.

I got real close using the 'goal-seek' utility,

I am very impressed on your use of the gaol-seek, and perhaps that is the way to go- rather than develop a generic solution use the solution you want and solve backwards.

But let me add some practical caveats re using them in the real-world:

And especially thanks for the caveats. There is no substitute for experience!

8. GSKrasle
August 15, 2014

Well, I HOPE the math is right, but here are my solutions to making logarithmic and exponential tapers from a linear pot. It's rougher than my usual work, but all I really wanted to do was to work-out the math. The solutions appear to be close to ideal, but were found with “Goal Seek,” not analytically.

But I can't seem to get a picture to paste!

This is the equation I used (check my work?):
Vo/Vi=((1/RP1)+(1/R1)-(((RP2+RP1)*((RP2*R3)/(RP2+R3)))/(RP1*RP2*(RP1+((RP2*R3)/(RP2+R3))))))/((1/R1)+(1/R2)+(((RP2+RP1)*((RP1*RP2)/(RP1+RP2)))/(RP1*RP2*(R3+((RP1*RP2)/(RP1+RP2))))))

In your five-resistor circuit above, if Rp = 1k, the best-fit to a “log” or “audio” taper is when R1 = ∞,  R2 = 944.7, R3 = 0. “antilog” requires R1 = 126.6, R2 = ∞, R3 = 9.8.

Just a fun little math exercise!

9. antedeluvian
August 15, 2014

GSKrasle

10. D Feucht
August 17, 2014

Aubrey,

These are good basic pot concepts. One way to think about why 3(c) is better than 3(b) is that the divider off the wiper attenuates not only the supply voltage across the pot but also the voltage noise variations caused by the wiper motion. Thus the resolution is increased by 1/ the divider ratio.

“I have only seen figure 3c in a 1987 Maxim seminar …” Take a look at older Tektronix instrument circuit diagrams. This is the usual method of applying pots in Tek equipment, and in particular, in oscilloscopes.

11. antedeluvian
August 18, 2014

Dennis

These are good basic pot concepts. One way to think about why 3(c) is better than 3(b) is that the divider off the wiper attenuates not only the supply voltage across the pot but also the voltage noise variations caused by the wiper motion. Thus the resolution is increased by 1/ the divider ratio.