The Ace Tone rhythm producer is an early attempt by Ace Tone / Roland / Nihon Hammond to create a programmable drum machine. The sounds available are very ‘Rhythm Ace’ like, and the usual preset rhythms can be found on the front panel, but what sets this unit apart is a programming matrix that allows limited control over the patterns, something that was very hard to come by in those days.

There are two control lines, and for each one you can activate Bass Drum, Snare, Cymbal, Clave and Low Conga. Once you’ve selected the sound you want to control, you create a pattern from 8 on/off switches, and then use the ‘measure’ buttons to activate this pattern in the first or last 8 beats of the preset rhythm over which it will play…

A close-up of the front panel, including the programming matrix.

Since each preset pattern is 16 beats long (except the 3/4 beats), and the matrix only 8 beats long, you’ll either have to repeat your programmed pattern twice per preset ‘loop’, or have it come on for only half of it. Still, you can independently control two sounds by using each control line, which allows for a good amount of variation. If you want to have a pattern that changes over the whole 16 beats, you can program the first 8 beats on the first control line, and the last 8 beats on the second, selecting the same sound, for example cymbal, for both of them… But that means you’ll only be able to control one of the sounds.

Another trick is to use one of the control lines to ‘mute’ a sound, by activating it but not putting any beats in. It’s a quick way to remove, say, that ‘clave’ sound that ruins the preset you’re trying to use! Sometimes, just replacing one of the voices with a different pattern can completely transform the beat, allowing to move away from the usual ‘beguine’ ‘bossa’ ‘waltz’ kind of thing…

This drum machine shows it’s possible to create a fully-analog pattern matrix, even if, in this incarnation, it’s a very limited one. If it could be done for the Rhythm Producer, there’s no reason why an expanded version couldn’t be built, with full 16 beat matrix for all six sounds, all this without using a single IC chip or digital circuitry. High cost is probably why Ace Tone decided to settle for the Rhythm Producer design, and not something more flexible…

Technical Info / Service Notes

The voice board has many small potentiometers that can be used to tweak the various sounds:

• BD frequency
• BD duration
• BD amp
• Low Conga frequency
• Low Conga duration
• Low Conga amp
• High Conga frequency
• High Conga duration
• High Conga amp
• Snare amp
• Cymbal amp

Unsteady tempo issue

When I bought this drum machine, the seller warned me that, although all the sounds and patterns were working fine, it had difficulty keeping a steady beat, hence the low price… Strangely, I have the exact same problem with another Ace Tone drum machine from the same era, the “Rhythm Fever”. I decided to take on the challenge and try to fix this issue on the Rhythm Producer, which, if I succeed, will probably enable me to repair the Rhythm Fever as well.

I’ll use this page to share my thoughts and my progress on this troubleshooting and repair, hopefully it will help someone else experiencing tempo problems with their vintage drum machine.

* * *

After receiving the FR-15 through the mail, I plugged it in and, as I had been warned, after about fifteen minutes, the tempo began to fluctuate strongly. Not your usual ‘analog, voltage controlled, slowly drifting tempo’ kind of thing. No, this thing is unusable. Sometimes the tempo slows down so much the beat almost grinds to a halt, and then five seconds later it’s at 130 bpm…

I did a bit of research on analog clock/timing in electronic equipment, and found that the most common clock source is based around the 555 IC chip. I looked around the insides of the FR-15, but couldn’t find a single IC chip! (this thing is old technology!) So the clock circuitry had to be made out of discrete components.

I read somewhere that R-C circuits (resistor – capacitor), along with a few transistors/diodes/etc., can provide a clock source. I don’t have enough electronics knowledge to locate, identify and troubleshoot this kind of circuit easily, so I decided to follow the wires leading from the ‘tempo’ pot, and that led me to a section of the circuit on the preset pattern board that’s very different than all the rest (see below). It has a transistor, three caps, a couple of diodes and five or six resistors. It also has two miniature pots that, I found, serve to adjust the upper and lower range of the ‘tempo’ knob.

This is where the wires from the ‘tempo’ knob lead. Two miniature pots for the tempo knob range, three electrolytic caps, two diodes, one c945 general purpose transistor, seven resistors including a high wattage one (3W / 1.5Ohm)

Different view of circuitry around tempo knob

So far, I’ve tried replacing the three electrolytic caps, but the problem didn’t go away, on the contrary it seems to have made it worse. I’ve also switched the transistor with another of the same type from elsewhere on the circuit board, to no avail. I haven’t replaced the resistors yet, because none of them appear damaged and I’m not quite sure if it’s worth it.

I suspect that the problem lies in this area of the circuit, but it could be somewhere else… I thought about the main power supply for the drum machine, but the sound quality isn’t affected by the tempo problem, and the tempo fluctuations are sometimes so drastic that I have a feeling a slow voltage drift isn’t the issue here…

So this is where I stand now; I’m not quite sure which part of the clock/timing circuitry is the culprit for the instability: caps, resistors, transistor, diodes, something else?

UPDATE — Tempo problem solved

After more research on the net about analog oscillators, I was able to locate the source of the timing clock in the Rhythm Producer. As it turned out, I’d come very close on my first repair attempt, but hadn’t replaced the faulty capacitors, which were just a couple of centimeters away from the ones I replaced.

After much internet reading, I realized that pre-IC timing circuits are made of, among other things, a pair of transistors and capacitors connected in a somewhat symmetrical way. A visual inspection of the drum machine led me to such a configuration, very near the area I’d been working on previously. Using a brand-new logic probe I’d just acquired, I was able to get the clock signal from that part of the circuit, which confirmed my hunch.

The two caps I found in that area were brown tantalum capacitors, and that set off alarm bells: a week before I’d found two very similar caps in another Ace Tone product, the PS-1000 monosynth, that were completely shorted. I pulled the two tantalum caps, found that they were a bit resistive, replaced them with new electrolytics, and now the drum machine is super stable.

Location of the two faulty capacitors, after replacement

If you see any capacitors such as these in an Ace Tone instrument, don’t trust them! They’ve caused me trouble on another Ace Tone drum machine, the Rhythm Fever.

Guilty!

Multiple output mod

I’ve done quite a few multiple output mods on drum machines already (check out my pages on the Roland CR-5000, National SY-50, Korg Mini-Pops 120 & Roland PB-300), but this time I decided to go for a more sophisticated design; the previous mods I’d done only separated individual voice, taking them out of the ‘main signal path’ of the drum machine, which means that once the drum machine has been modded, you can’t go back to a ‘one jack, full mix’ drum machine output. I don’t mind that much, because I can always recombine them in a mixer or through my sound interface in the computer (which is perfectly fine for studio work), but this time I decided to do it differently…

The following design allows for separation of the main drum voices, but each instrument only ‘comes out’ of the main output if a jack is inserted; thus, if only the main output jack is used, the drum machine behaves exactly as it did before the modification. This is useful is you have a limited number of inputs on your sound interface, if you want to isolate only some sounds, or if you want to keep that ‘all-in-one’ feeling of the original box. It’s not that much more complicated than the other mods I’ve performed, but you do need twice the amount of wiring, and jacks with a ‘shorted pin’ to keep the sound flowing when no jack is inserted.

I spent a bit of time signal tracing inside the drum machine, until I was able to isolate each sound just before it gets ‘mixed in’ with the other. I found that I couldn’t really separate the snare sound from the cymbals, because both of them use the same noise generator, so I had so settle for these outputs:

1- mix out
2- bass drum
3- snare drum + cymbal + maracas
4- low conga
5- high conga
6- clave (which is actually two signals mixed together, see below)

The last component, right before each signal is joined with the others, is a capacitor, so I decided to pull these out of the PCB, and use that space for wiring the jacks to the circuit. This made it easy to solder all my wiring to the jacks, but I was left with a problem: where to put the caps I’d removed… In the end I decided to solder them directly to the lugs of the jacks, which I think isn’t ideal (mechanical stress from inserting and removing the jacks could, in the end, mess up the connection), but hopefully it won’t be too problematic.

Location of the various voices. The green bars represent the capacitors I removed.

After I’d removed the caps, I soldered wires in the holes I’d just freed up on the PCB, and connected those to the tip pin and shorting pin of the various voice jacks (not forgetting the capacitors), and I was good to go.

Prior to that, I had to drill new holes in the drum machine’s enclosure to make room for 4 new jacks (I ditched the footswitch jack and converted it to an output), which was pretty easy using a step bit. I soldered everything in, turned the drum machine on and it all worked on the first try… nice!

Voice board, after soldering all wires

Output jacks after modification. Notice the caps that were taken from the voice board.

Output jacks after modification

Specifications

Gallery

Preset selection buttons + pattern board

This is where the wires from the ‘tempo’ knob lead. Two miniature pots for the tempo knob range, three electrolytic caps, two diodes, one c945 general purpose transistor, seven resistors including a high wattage one (3W / 1.5Ohm)

Different view of circuitry around the tempo knob

Front panel and control lines

Voice board before modification

Location of the two faulty capacitors, after replacement

Location of the various voices. The green bars represent the capacitors I removed.

A close-up of the front panel, including the programming matrix.

Preset pattern board, bottom view, before multiple out mod

Front panel and control lines

Output jacks after modification

Voice board, after soldering all wires

Output jacks after modification. Notice the caps that were taken from the voice board.

Gallery contd.

This is where the wires from the ‘tempo’ knob lead. Two miniature pots for the tempo knob range, three electrolytic caps, two diodes, one c945 general purpose transistor, seven resistors including a high wattage one (3W / 1.5Ohm)

Different view of circuitry around the tempo knob

Front panel and control lines

Voice board before modification

Location of the various voices. The green bars represent the capacitors I removed.

A close-up of the front panel, including the programming matrix.

Preset pattern board, bottom view, before multiple out mod

Front panel and control lines

Preset selection buttons + pattern board

Location of the two faulty capacitors, after replacement

Output jacks after modification

Voice board, after soldering all wires

Output jacks after modification. Notice the caps that were taken from the voice board.

Main voice board of the Rhythm Fever. Almost identical to the Rhythm Producer and FR-6 boards