Updated 8/27/07

Thinking About Making Pickups

Before starting this project, I studied how the original Hohner pickups were made (as much as I could see through the clear epoxy potting material). I was curious about the humbuckers on Clavinet.com, but I wanted to try making pickups myself like the originals if I could. I took physical measurements of the poles in my Hohner pickups and measured the electrical resistance of the windings. I was doing some research on the Web about making pickups and realized I needed a pickup winding machine.

Pickup Winding Machine

I discovered Jason Lollar's web site lollarguitars.com and he had a book he wrote on how to make a pickup winding machine. In the course of ordering it from him I found out that he is the one who makes Clavinet.com humbuckers. That was cool! His book arrived and had complete plans to make the winder. It told how to use a motor from an oscillating fan for the cam that distributes the windings evenly, and a sewing machine motor for the main spinning power. So I hit up a few flea markets and got those two parts. Since then (2003), I think he stopped selling his book. It's too bad because it is an excellent book. So I read his book and made the winder. The finishing touch I added to it was a home-made digital counter to count the turns of wire. A small magnet on the shaft of the winder passes by a Hall-effect transistor switch and increases the counter for each revolution. I sent him photos and told him what I intended to do. He liked the winder, but told me my pickups might have problems with microphonics and to experiment with them. Here are some photos of it.

Making the Poles for the Pickups

I used a hack saw and cut the poles out of cold-rolled steel stock (dimensions). There were two different shapes needed to fit the harp: the square-cut ones above the strings at the end of the harp, and the diagonal-cut ones inside toward the middle. I located some magnets that were slightly shorger than the length of each pole. I superglued the magnets to the poles. I wrapped the place where the wire goes with blue Scotch Safe-Release Masking Tape. I used #40 magnet wire to wind them. The Hohner pickups measured 175-ohms total, and that made them 29-ohms per pole. I made several sections before finding 131 to be the right number of turns to equal the resistance of the original pickups. I found some empty plastic extruded tubes that computer chips come in and cut off pieces for the troughs to hold the six poles to make a whole pickup. I glued them in and aligned them so the gaps were no more than the width of a thick string, then, I wired the six pole sections of each pickup in-phase like the originals. After testing the preamp (next), I finished the pickups by potting the windings with clear silicone adhesive so they wouldn't be exposed or wiggle around.

Beginnings of the Preamp

Clavinet.com has the Hohner schematic for the preamp, so I used that as the model for mine. The first thing I did was find a small audio transformer on eBay. I wasn't exactly sure what I was looking for but the one I found had a 200-ohm winding and a 64k winding. The 200-ohms seemed a close impedance match for the 175-ohm pickups. The ratio of 200 to 64k would give a voltage gain of about 300X. I didn't know what level of signal the pickups were generating, but with only 175-ohms (compared to maybe 5,000-ohms on a guitar pickup), I figured the signal level was pretty small, so 300X gain would be good. I mounted the first home-made pickup on the harp I got from Clavinet.com (I already had finished the keyboard and attached it) and ran a shielded wire to a little electronics experimenter's vectorboard. I put the resistors, capacitors and transistors on the board like the schematic showed, and connected a battery. To my delight, the clavinet sound came through the amplifier at a reasonable level.

Preamp Filters

Figuring out the filter circuits was going to be more difficult. To do that, I needed to do a strange experiment using the Hohner Clavinet I had. I disconnected the pickup wires at the input to its preamp and substituted a sine-wave signal generator. I set it for 1000 Hz and put an oscilloscope on the output of the Hohner Clavinet preamp and adjusted the generator so a signal of about 100 mV came out the Clavinet output jack. That was with all the switches on the Clavinet so that the filter coils were not in the circuit.

I set the Clavinet switches for each individual filter coil to be in the circuit one at a time, and I made a series of measurements changing the generator from 100 Hz up to about 20,000 Hz. Those were the end points where the preamp response rolled off to very low levels. I varied the oscillator and measured frequency points between those extremes where the level of the output changed significantly. I made a chart plotting those points for each filter coil.

Then I set the Clavinet switches so both filter coils were in the circuit simultaneously. Again, I varied the oscillator and measured frequency points where the level of the output changed significantly. I made a chart again, and came up with a drawing showing two bandpass peaks and a notch in between them.

A coil and capacitor together make a tuned circuit, and the peaks and notch are the result of the two Hohner coils working together. I got a hold of some coils used in organ voicing circuits and went back and experimented with my own preamp on the experimentor's vectorboard. I was able to find some combinations of coils and capacitors that gave roughly the same curves as the Hohner preamp. The coil values in the Hohner schematic show 2.0 Henry and 0.5 Henry values. The coils I used are not such wide variance. I don't know how many Henries mine are but they are 1700 turns of #39 wire and 750 turns of #36 wire in ferrite cups that are about one-inch in diameter and 3/4" high.

I left well-enough alone and incorporated those parts in my preamp. As you can tell, I didn't design any of this as an engineer might do with math! I did it empirically with whatever parts I could get. I think there are several different ways a person could do this. I can't really explain how it all works, but I knew what to do to copy it and get similar results. I think the two primitive movies I linked prove that the twangy, nasal Clavinet sound came through in spite of my unscientific method.

Testing the Pickups

At first, I wired the six pole sections of each pickup in-phase like the originals. I used it for a while with those pickups and noticed a lot of hum and buzzing unless I found the best place in a room for the clavinet to be sitting. I kept thinking of the humbucker pickups at Clavinet.com. I looked at the photo there, and it seemed that they were made from two long pole sections, each the size of three of my poles.

I read about guitar humbucker pickups and how they are made with two parallel bobbins wound out of phase. I didn't know how to do that with a long blade-type pickup like the clavinet had until I started thinking about the six poles as six separate pickups. It finally dawned on me that the Clavinet.com pickups might be wired with the two sections (halves) in a pickup out of phase to make it buck the hum. I made a special little test pickup out of two of my one-sixth size poles and wired them out-of-phase and, voila! The hum disappeared. Then I tore my old pickups apart and re-made them and alternated the phase of each pole section. That really did the trick and they work excellent now as far as bucking the hum. Those are the ones in the photos of my clavinet.

Jason Lollar was right, however. My home-made pickups/preamp does seem kind of sensitive to microphonic vibrations. That's something I'll work on in the future to fix. I still want to get some pickups from Clavinet.com and try them sometime.