The parts for my capacitor project had finally arrived and it was time to sit down at the workbench and think through just how all these components would fit into the Hammond enclosure. I knew it would be a tight fit considering there would be at least twenty two capacitors inside the box once it was complete and each time I installed a cap there would be less room to work with.
I managed to find twelve position rotary switches made by Alpha and a CTS concentric potentiometer with a 250K and 500K resistances which are the two most common pot values in electric guitars. Single coil pickups will generally have a 250K pot whereas humbuckers will usually have 500K. I wanted the left side of the box to be .022uf capacitors and the right side to be .047uf capacitors. It turned out I had eleven "sets" of capacitors in both of those values meaning each pair were made by the same company. There were more capacitors in stock but this covered many of the common capacitors being sold today and left me one position which would be "off". I used a multimeter to label each lug of the rotary switches and decided the order of capacitors ahead of time. The two Carling On/On mini switches would be used to select which potentiometer resistance was in the circuit and the other switch would send signal to either rotary switch (choosing capacitor values). At this point I was still undecided about using terminal strips or buss wire for the grounds which didn't really matter since the first steps of construction would be locating and drilling the holes for all the switches. I wanted to get the these mounted before finalizing the location of the input and output jacks.
Once all of the switches were mounted I could verify enough room existed to put the jacks where I wanted them and get to wiring up everything except the capacitors. I wasn't 100% sure where I could pull resistance measurements from but assumed it would be somewhere on the mini switches. I also wasn't sure I could have the multimeter plugged in while using the box with an amp since there would likely be direct current from the battery injected into the circuit.
The time had finally come to start soldering so after removing the terminal strips I used the existing holes to mount star ground tabs which could be soldered to. I took 18 gauge buss wire and ran it through the tabs and around near the perimeter of the box with the ends soldered to the potentiometer. The aluminum box is conductive but soldering directly to it would be difficult since the thickness of the metal would absorb the heat before melting a ground bead. There wasn't room to use the 200 watt soldering iron like on an amp chassis. I started at the input jack and wired up all of the guitar signal path including jacks for the multimeter probes so I could see if I had it right before the capacitors went in. I didn't want to have to fix mistakes once the enclosure was completely full with parts and capacitors but assumed there were corrections in my future.
Not only had I determined which capacitors would be soldered to each position of the rotary switches, but I also measured each cap for their actual value ahead of time. As I started wiring in capacitors I began to wonder about just how much noise would picked up by all of the wires in the box. Each capacitor would have heat shrink on the leads to prevent any shorts in the signal chain but that does nothing for noise. It didn't concern me enough to stop working before the capacitors were in so I wouldn't know until it was completely done.
It took a few hours to wire all of the capacitors and I was fairly lucky how the order of things worked with the space inside the Hammond enclosure. The buss wire really offered versatility for grounding since the leads of the capacitors differed in length. All in all the process went fairly smooth and I was anxious to plug in and give it a listen.
One snafu appeared right as I was finishing. Early in the process I had wired all of the switches and meter jacks and confirmed the meter would work to reference either 250K or 500K resistance. However later in the process and after a few "on the fly" design changes I noticed something odd....153K of resistance. That meant the two potentiometer values were now wired together in parallel instead of being separate. I took a break and went back to the design checking what had changed. The meter jacks were disconnected and I used some jumpers to correct the problem before soldering new wires back in.
On the initial testing with a guitar and amplifier I was mortified for a minute when I heard the noise. As I had suspected the meter was injecting 9 volts into the circuit and was making a lot of noise. When the meter was turned off the vast majority of extra buzz went away. However it seemed like there was still a little more buzz than I would have hoped for. I took a short break and grabbed some Mogami shielded wire and decided to replace the original cloth covered push back wire with shielded cable. I would use the center of the wire for the positive signal and the shield soldered to ground on just one end. This way any electromagnetic interference picked up by the shield would be sent to ground. So while it felt like a setback I was fairly happy with the design and probably being overly picky about the buzz. This was never expected to be exactly like a single tone control in a guitar. There are a lot of parts in a small space and plugging a guitar that already had a tone control would be just adding more load to the circuit before getting to the amp. What I wanted all along was a fast, easy way to listen to one type of capacitor to another and see which ones I liked best. After all the shielded wire was installed the design was very quiet and subtle differences from one cap to another could be easily heard. Currently there is one capacitor not working and I haven't taken time to try and reflow the solder or just swap the capacitor out with another one. Eventually the plan includes a video so people can listen for themselves to see which capacitors they like and before that I'll fix the one problem remaining.