Optical Potentiometer for a DIY Amplifier

This instructable needs a short foreword. The title states that the pot is for an amplifier and not for any amplifier, but for a DIY amplifier - this is because I developed it for such an amplifier, but you, of course, might find a better use for it :)

Nevertheless, it has certain characteristics, which make it interesting for audio applications, and DIY amplifier might have enough space for it. Above, I used the verb "developed" for a purpose - even though the idea is quite simple, I never saw it on the Web nor elsewhere. I didn't perform a deep patent search and I wouldn't be surprised if someone will point out that a similar idea was patented back in 19xx or so (an update - I found a commercially available pot for an exorbitant amount of money, which uses the same principle, but is composed of a single resistor and a single photoresistor, so it can't replace an audio pot; another article describes a pot with a reflective stripe, which is also not what is needed). In any case, feel free to use it for non-commercial purposes. The others are welcome to contact me :)

What is a need for such a device? I guess, most of you have faced with a pot of some audio device, which started makind a hissing sound in the loudspeakers when you turn it and you know that some day this pot will start losing one channel and then you'll have to replace it. In the world of HiEnd audio, these pots have been replaced with the 24- or 32-position rotary switches, which ensured a proper contact. Nowadays, the volume is programmed electronically and the modern amplifier has an encoder attached to a volume knob, so it's more of an input device rather than a real potentiometer. Still, there're a lot of devices, which use a "normal" audio pot to adjust the volume and this instructable is for the people designing or repairing such devices.

The idea is as simple as the electronic scheme above. Imagine that we have a potentiometer, which is composed of two photoresistors Rph2 and Rph1. If we had a light source and if we could redistribute the flux between these photoresistors, we would change the Vout/Vin ratio in the min(Rph1)/(min(Rph1)+max(Rph2)) ... max(Rph1)/(max(Rph1)+min(Rph2)) limits.

To adjust these limits and the taper, one can use the fixed resistors R1 and R2, which will be discussed below.

One can imagine several schemes of flux redistribution between Rph1 and Rph2, but since we are speaking about an audio pot, it would be logical to bind this redistribution to the rotation of the volume knob. The first idea, which comes to mind, is shown in the figure. The Rph1 for the left and right channels are mounted on one side of the PCB, two other photoresistors are mounted on its backside and the PCB itself is mounted in the rotating enclosure, illuminated from one side. This setup ensures a smooth taper because of the cosine of the angle between the photoresistor's normal and the direction to the light source. It's needless to say that for a given position the illumination of L and R photoresistors should be equal. This is achieved by using a scattered radiance or by a source placed at a sufficient distance - the choice is up to you.

In audio equipment, the log pots are used more often than linear ones because the response of human ear to the loudness of sound is logarithmic. Making a custom-curve pot is difficult for an amateur unless he/she uses an aforementioned rotary switch, but an optical pot provides you endless opportunities. Imagine, for example, a tinted film moving in front of Rph1 and Rph2 - by changing the density, one can fit any curve.

But, the setup described in this instructable aims to reproduce a logarithmic curve or some smooth curve, which would resemble a logarithmic one. The Figure shows the results of simple calculations for a couple of pairs of (R1;R2) coupled with real-life photoresistors available at RadioSchack or elsewhere. As one can see, the optical pot has two drawbacks w.r.t. resistive one: (1) the sound is never fully cut and (2) the full volume is never reached. However, the curves give you an idea of what can be done to improve the situation - either the number of photoresitors has to be increased (check the curves with N=2) or a brighter source should be taken (check the curves with F=2). The sharp transition near 90 deg is not that sharp in the reality because the light is scattered and the simple model does not include the scattering or the shadow of the enclosure wall. In the real setup, the pot is characterized by a S-curve, which is not a logarithmic one, but which is pleasant for the ear.

If you want to reproduce this variant of the optical pot, you will need:

(1) a prototype PCB of about 3x5 cm

(2) 4 photoresistors, which change their values from 10K to 0.3K under normal conditions (small flashlight)

(3) a set of 1-5-10-20-50-100-200K resistors (you can choose a pair from the curves above, but you might also want to experiment a bit)

(4) a length of one-inch PVC pipe

(5) a solder, a multimeter, and a saw.

(6) wires, screws, hot glue

(7) a flashlight or any other source of light (I used the ballast lamp of my parafeed tube amplifier)

Once you've collected these items, the rest is straightforward - cut the PCB to fit the PVC pipe, solder the photoresistors and regular resistors according to the circuit diagram, cut two windows in the PVC pipe, put the PCB inside the pipe, solder the pot wires, mount the pipe and connect it to the volume knob (I did it using a flexible hose to have a smooth rotation), mount the light source and point it at the photoresistors.

It'd be a good idea to foresee a possibility of replacing R1 and R2 without unmounting the whole system. If you find that the sound is too loud in the leftmost position of the pot, decrease R1 or increase the light flux.

If you did everything right, you'll have a nice control of your music and the potentiometer's hissing, clicks and other sounds will be gone forever. The pot itself does not introduce any coloring to the sound, and there's no crosstalk from 50-60Hz light sources, either. The latter is explained by a "running mean" nature of photoresistor - it takes a while before the change of the input flux turns to a change of the resistance. This creates a small delay of volume change, which is a feature and not a bug. The volume increases smoothly and the overall impression of this pot is quite good. Hope it will be your case, too.

That's it, thanks for reading :)