Yes, you heard it right, you can use an audio speaker as stacking stage for high magnification work. It was during the Chinese new year when almost all businesses are shut down for almost two weeks in China, my speaker for PC was behaving strangely, after all, I have been using the pair for about 7 years. So, I started investigating what was wrong. By touching some wire, the speaker made some noise, and since I was testing some setups for high magnification stacking, the movement of speaker cone caused by noise immediately drew my attention: can that movement be used for stacking? I was pretty sure it can and I posted the idea on a forum which drew some attention.

After some more experiment on my own with above circuit, I think it is a very viable solution.
How exactly does it work? Well, as we all know, when electronic current representing audio sound waves passes through a speaker, the speaker cone will vibrate and cause air to vibrate, too, thus making physical sound wave. However, if a DC current is fed to the speaker, it will only cause displacement of speaker cone, up or down and is proportional to the amplitude of current (may or may not be linear). Ah, that is interesting, right?

Now, from the right side of the above circuit, it is just a resistor and a speaker. As we all know, the main component of speaker is a coil, which will be resistive and inductive. However, if we do not feed the speaker with audio signal, only a DC current, the inductive part will be almost zero, so the right hand side of the circuit is almost like a resistive linear circuit because the speaker can be thought as just another resistor under DC.

Having established that the right hand side of circuit is close to a resistive linear circuit, changing current passing through it simple — just change the voltage across it, albeit slowly so that the change will be almost like DC. Therefore, having a variable voltage circuit is enough to experiment with the idea of using speaker as stacking stage. And that is exactly what the circuit on the left hand side does — using a LM317 to make a variable voltage source. Why is it necessary to use LM317? Why not just use a potentiometer? LM317 configured as variable voltage source can drive different loads at set voltage without issues while a potentiometer circuit does have issues.

This is actual implementation of the circuit. I did not have through-hole capacitors, so I did not use them. It is highly recommended to use multi-turn potentiometer so you can fine tune the current. The voltage readout is a convenient way to see actual voltage

Here are some formulas for above circuit:

Vout = 1.25*(1+R2/R1) and since R2 is a potentiometer, Vout will be variable.

In actual construction, I used 10K potentiometer for R2 and 1K for R1, so the variable range is 1.25V to 13.75V, Note, the input voltage must be approximate 3V higher than maximum output voltage, so to achieve 13.75V, the input voltage must be more than 16.75. In my construction, the maximum is caped at about 9V. I chose R3 to be 200 ohm and the speaker I used is a 4 ohms one.

After building the circuit like above picture, I was surprised that it works really well. Here is a video showing it in action.

As you can see from above video, there is very little lateral movement, meaning very little or virtually no “wobbling” associated with most mechanical devices. And since the displacement is controlled by current flowing through the speaker, there is NO backlash either.

There are many advantages using voice coil as stacking stage. However, here I need to point out that a speaker is not linear across its full range due to its construction, but it is easier to find a linear region for our use and particularly for high magnification work where total displacement needed is usually less than 1 millimeter.

More experiments are coming . . . stay tuned.