Synth Tutorial, Part 6: Fatman

Sixth in the series of video tutorials on synthesizers, looking at the PAiA Fatman in detail.

Part 6: Fatman


I’m in the process of putting together a short series (maybe less than 10 parts) of video tutorials on synth music/electronic sound. In preparation for that, I want to talk about the gear that I already have, and give short demos of each of the major instruments.

Youtube video

Tuning the Fatman

I doubt anyone is ever going to need this file, but since I’m doing this work, I might as well record it to have something to go back to.

(My set-up. The A-300 is on the floor to the right, propped up against the table.)

The PAiA Fatman has been around at least since 1999 (copyright data printed on the main circuit board), and the instructions really haven’t been updated since then. So, I’m assuming that I’m in the minority of new customers that don’t know how to tune it based on “C” and “octaves”. The instructions for tuning the synth tell you to press the lowest “C” key on your keyboard, set the pitch control to match a reference note from a tuned keyboard, and then tweak the trim pots to properly adjust VCO2. Then, go to the highest “C”, repeat the process for 4 octaves up, do the same thing for the “C” notes in the middle for the middle octaves, and finally either do the entire process a couple times to get VCO1 and VCO2 in perfect harmony at both ends of the scale, or not. While there’s a suggestion for using an oscilloscope or frequency counter instead, there’s no mention of the actual frequency numbers that are needed for each “C”. Since I don’t have a scope, frequency counter or reference synth, I’d look to be out of luck.

(Japanino in freq. counter mode.)

However, I do have the Gakken version of the Arduino micro-controller, the Japanino, with the LCD shield. I can hack together a very simple sampling o-scope and a frequency counter, as long as I’m careful about not applying too large a signal to the Japanino analog in pins. I could save myself a bit of worry by replacing the 3-AA battery pack with a 4-AA pack and get Vcc up to 6V (some of the signals in the Fatman are 7V) and then put a potentiometer in series with the analog in pin as an amplitude control. But for right now, that’s not really necessary.

(Jumpers on the pot. controls are running to the Japanino.)

The Fatman only recognizes MIDI notes in the range from 36 to 84. If you look at a MIDI to frequency table, you’ll see that the lowest usable key is generally note 21, which is A0, or 27.5 hz. Anything below this is really hard to hear with the human ear. C1 is note 24, at 32.7 Hz. For the Fatman, the lowest key is MIDI note 36, which is C2, at 65.4 hz. The PAiA tuning instructions refers to this as “C0”. The highest used note is 84, “C6”, at 1046.5 hz. This then gives us the other C’s as:

C# Note# Hz
C0 36 65
C1 48 130
C2 60 261
C3 72 523
C4 84 1046

We don’t need to be dead-on with the frequencies, “close enough” is good enough. PAiA suggests as an alternative to a reference synth to use the online tuning fork. I think having the freq. counter sketch on the Japanino will be fine. What does make a big difference is having Offset at the middle of the range, pointing to Unison, and most of the other controls also at midrange just to get an audible signal to the speakers.

(Finished Fatman all boxed up.)

The Fatman doesn’t have its own MIDI keyboard, so I’m using my Roland A-300 Pro. This is a 32-key MIDI controller with +/- octave shift keys. The easiest approach is to hook the A-300 to the Fatman MIDI IN, then press the -Octave button on the keyboard and depress the lowest keyboard key until there’s no sound out from the Fatman. Work up the keyboard keys until you get the first playable note, and that’s going to be C0 for our purposes. Follow the tuning instructions for the Fatman. Then count up 12 white and black keys to C1, and adjust the R13 trimmer to give a frequency of roughly 130 hz. Use the +Octave button to shift the keyboard and press the notes to get the highest playable one. This will be C4, and count down 12 black and white keys to get to C3, and adjust R18 to 520 hz. Count down 12 keys to C2 and adjust with R21, and again to C1 and adjust R24. The last step is to set Offset to Unison and adjust R42 so that VCO1 and VCO2 are at the same frequency.

(Back panel. The only thing I wish were different is having an AC plug-in jack, rather than having the power adapter permanently wired to the box. Note the Pitch, Vel. and Gate RCA jacks. These allow you to send voltage out signals to other equipment.)

I initially tried to connect the Japanino A0 analog pin to the H pin of the output speaker jack, with GND to the G pin of the jack, but the scope sketch wasn’t registering any kind of signal. Instead, I connected A0 to the leftmost pin of the R56 control (VCO1/2 Mixer), and that gave me a large enough signal to work with, without threatening to damage the Japanino. I think having straight numbers out on the LCD shield speeds up the tuning process a lot.

(Bottom of the box. The bottom panel has gaps at the front and back for cooling air flow. The cutout in the upper right corner lets you get at the DIP switch for setting the MIDI port number.)

I think I’ve got the thing tuned for my purposes. There are a few mods on the internet that I may add later, so if I do go back and open the case up again to put in a mod or two, I can always do some retuning if necessary. The next step is to just find out what it can do as-is.)

Fatman Synth, Part 1

(All the parts right out of the box.)

Back in the 80’s, a company called PAiA came out with a hobbyist line of build-it-yourself analog synthesizer kits. I bought the smallest one, a little ribbon-controller unit entitled The Gnome, for about $35. It had pretty much the same functionality as Gakken’s SX-150, but with a soft vinyl ribbon strip. It was fine for playing with for a few hours, but after a while I put it in my closet, and eventually gave it away. It’s now a collector’s item on E-Bay. Anyway… I have something of a soft spot for PAiA, and recently I’ve been looking for a cheap DAW I can hook up to the Roland A-300 Pro for making music independent of a PC. Generally, a DAW (digital audio workstation) is software running on a PC, but there are a few hardware instrument synth boxes on the market, in the $1000 range. I wanted something cheaper, and I kept coming back to the PAiA Fatman kit.

(Everything excluding the case. Glasses not included.)

Finally, I ordered one off the PAiA website, about $260 including the main kit and desktop case, plus $60 for shipping to Japan. More than I wanted to spend, but what the heck. The kit has an optional 12 VAC power adapter that you can decide to not get for international orders. Because the west side of Japan runs on 110 60-cycle VAC, I decided to get the power adapter with the kit. Shipping weight was about 4 pounds. Finished package size is 29cm x 14cm x 7cm (roughly 11″ x 6″ x 3″). Everything is included except solder and tools (required tools include soldering iron, clippers, some kind of pliers and a small screwdriver. A ruler and wire strippers are recommended.)

(Circuit board with the 43 jumper wires and all of the resistors.)

There’s no suggested assembly time. Reported times are between 6 and 40 hours. It took me about 16 hours, not including that for calibration and tuning. The only way I can see to get a 6 hour time is to pre-cut all the wires and pre-sort all the resistors before starting the clock. The circuit board is single-sided, which means that there’s a need for adding jumper wires that normally would be plated on a 2-sided card. The kit can be mounted in a rack, or put in a desktop case, which changes the wiring needed between the knobs and the board. I picked the desktop case, with its larger surface area for getting at the knobs. The instructions are pretty straight forward, although I did have to flip back and forth a lot between the main manual and the desktop supplement when it came to the knob wiring. You start out by adding the 43 jumpers. Then, all the resistors. Although the instructions don’t warn you, you’ll end up with 2 extra resistors when you’re done, because they’re used later when you wire the knobs.

(All the capacitors installed.)

One power resistor is actually two 1-watt resistors in series and raised off the board to allow for air flow. Next, you add the caps, followed by the diodes and transistors. One power transistor needs to have the heat sink clipped on before being wired into the board, but there’s no diagram of the heat sink installation and I was left guessing as to what I needed to do for that. Fortunately, it wasn’t that big a deal to clip on. This is followed by the trim pots, the 2 IC sockets and the back panel connectors (3 RCA jacks, and 2 MIDI jacks – one IN and one THRU).

(Trim pots and sockets.)

The main EPROM and the 8031 microcontroller are the only two chips that go into sockets. The rest are soldered in place. This is the trickiest step in my opinion, because there’s no way of knowing if you’ve screwed something up until too late. The EPROM and the 8031 get plugged in at the very end. Then comes the case wiring.


The wires running from the board to the pots in the case are called the “flying wires”. Installation is in three steps. First, cut the flying wires, tin the ends and solder them to the board. Next, cut the second set of wires, depending on whether you have the rack or desktop case, and wire up the connections between the pots. Third, solder the floating ends of the flying wires to the pots. Steps one and three are where I made my first real mistakes – I soldered the wires so they pretty much stand up and down on the board and the pots. Bad move. This prevents the board from fitting in the case when it comes time to fold the wires in half and put the board in place. I had to go back and resolder all the pots so the wires run mostly flat against the inside surface of the case. Even so, the wires still don’t like being folded to bring the board up to the pots. Note that the DIP switch S2 is mounted on the solder side of the board for the desktop case.

(Steps 1 and 2 for putting in the flying wires.)

Put the EPROM and the 8031 in their sockets, check your work, then plug in and turn on the power switch for the all-critical smoke test. There was no smoke, and the power LED lit properly. Nothing overheated abnormally. The Fatman doesn’t have its own input system, so you have to plug in a MIDI keyboard to the MIDI IN jack. The MIDI Note On/Off LED would flicker properly when I pressed the A-300 Pro keys, but I couldn’t get the Gate LED to light when I set S2 for MIDI Port 1. That’s when I spotted the big solder splash shorting 2 pins on IC 7, a smaller solder bridge from one pin to an adjacent circuit path, and a big ugly bridge between pins 2 and 3 on the EPROM. I take a fair amount of pride in my soldering skills, so it really depresses me to think that I completely missed all three problems at the time. Especially the ugly bridge, because I don’t know how that could have happened unless it was loose solder on the table and it welded itself in place to the board when I set the circuit card down. Anyway, there doesn’t seem to have been any permanent damage caused by the bridges or splash, and the Gate LED lights properly when I send MIDI signals to port 1. So, that’s good.

(Step 3.)

The next step is tuning and calibration. This requires either using an o-scope, a frequency counter, or comparing the sound out against another musical instrument. And I have none of those things. I do, though, have the Japanino with the LCD shield, so I’m in the process of making myself a cheap and dirty signal tracer/freq. counter. When that’s done, I’ll return to tuning the Fatman.