synthgeek's blog

This is a toy drum machine with 8 playable pads (not velocity sensitive), 4 percussion-sound buttons, 30 preset rhythm loops, and a built-in speaker. There are two choices of drum kit for the pads, “classic” and “electronic”. I added a 1/4″ line-out jack (switching, so the internal speaker is cut off when something is plugged in) and a pitch knob.

Here’s a shot of the board, pre-bend, with the clock resistors marked:

There are two resistors there, on a daughterboard attached to the main board at a 90-degree angle- this daughterboard also contains the main IC/ROM, which is of the “black blob” variety. Adding parallel resistance to just the top resistor (thus lowering the resistance there) with a pot allowed only upward pitch changes, while putting the pot between the upper side of the top resistor (top in the photo, that is- it’s actually the bottom side of the device if you were looking at it head-on, like you were playing it) and the bottom side of the resistor underneath it (closer to the main board, not down from the circle in the photo- both resistors are inside the circle) allowed the pitch to be shifted slightly higher and much, much lower, with less resistance here giving a lower pitch.

The yellow wires which lead up & out of the photo are the output connections. I believe the 8-pin IC next to them is the amplifier.

I was unable to find any other bends that were very interesting. However, if one were to find a non-working one (or just not like it enough to keep it around), there are a few things that would make fine salvage. The drum pad triggers are piezo discs, there are a few slide switches, and the smaller board which is screwed in at the corner there, facing the opposite direction of the main board, has the two small slide pots which are used for the pad & rhythm track volumes. Also, the built-in speaker is around 5″, and not too shabby for a toy like this- might be useful for a small lo-fi amp of some sort.

Here’s a video:

Found this thing at a second-hand store cheap- since it’s made out of clear(ish) plastic, I could see a couple of resistors on the circuit board, and thought it would be worth checking out for the price, in case one of them was a pitch control. Score! It was.

Below is a shot (kinda blurry, sorry) of the circuit before bending. The resistor marked “R2″ is the pitch resistor. You can also see the backside of the built-in speaker:

The rest of the circuit wasn’t very exciting, the sound-making part was a blob-type IC. The other possible bend would have been to add external play triggering, but retriggering while it was already playing didn’t interrupt what was already playing, so that wasn’t super-exciting either.

What I ended up doing was taking out the existing resistor, replacing it with a 10k (it was 55k), and adding a 500k pot inline. Here’s a shot of the board after modification, notice the heat-shrink tubing around the resistor for added strength:

I ended up having to add a second 10k resistor to the opposite side of the pot, as 10k wasn’t enough resistance to keep the chip from crashing (easily fixed by cycling the power). Of course, a 20k resistor at the board would have worked too, but I’m fresh out anyway. I also added touch points (using screws) for changing the pitch- they also cause some glitching under certain circumstances, sometimes leading to crashes, which again can be easily remedied by cycling the power.

I also added an output jack (with a 1k resistor on the board end of the wire)- I had to go with 1/8 inch, due to size restrictions imposed by the enclosure. I also had to grind away some plastic which was getting in the way of the jack.

Some shots of the finished piece:

…and here are some sounds. First up, a dry recording, messing with the knob & touch points:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

Next, processed through the LM567 amplitude modulator in Phoenix, then a lowpass filter, and a delay:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

This one is just the Sorry Pawn with delay:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

I recently ordered a couple of CD2024 dividers from Jameco, but got Motorola MC2024‘s instead. Last time they sent me Motorola substitutions, it didn’t matter- but this time, the chip is entirely different: a dual VCO, as opposed to a divider. Turns out these might be useful, so I won’t bother sending them back, but I do plan to point it out to them.

This chip is designed to operate at radio frequencies (MHz), and a chart in the datasheet shows that the noise performance gets worse as you go lower in frequency. However, I didn’t let that stop me from having a go at it on the breadboard. Although it’s not perfect (the noise/distortion gets noticeably worse as you go into low frequencies), it is fun to play with.

Below is the circuit I came up with for using it as an audio-frequency oscillator. Note that the datasheet recommends +5V operation, so I also put that in the schematic- however, the version I built on breadboard uses +9V, and it seems to work. That is to say, it doesn’t seem to overheat or anything like that. Also note that the function of SW1 is to switch between CV modulation of the 2nd osc along with the first one (thus making it like a unison osc if using a dual-ganged pot, which I represented in the schenatic as VR2a & VR2b), or modulation of the 2nd osc from the first (thus making it a pseudo-FM osc).

Here are some recordings of my tests- the first one shows the oscs in “unison” mode, with CV modulation from a Schmitt trigger LFO:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

The next one shows the oscs in “FM” mode, with osc 1 being modulated from the same LFO:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

The third and last one is a combination of both modes, with the modulation source’s rate also being tweaked, often going into audio rates:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

After trying unsuccessfully to sell my Yamaha Electone 315u-i  electronic organ for a good long time now, I finally decided to take it apart and see what might be useful elsewhere.

The following view from the back shows some of the more interesting parts: a Leslie hookup, a wannabe-Leslie rotating speaker setup (in the boxed-off area in the bottom right), a spring reverb tank, and the main speakers. My initial thoughts are to use all this to make a combo amp.

The next photo shows the power board:

This one shows the microcontroller board, which contains the arpeggiator and auto-accompaniment components- I think at least the sequencer portion of the drum machine resides here too:

So far I’ve also found a couple other interesting things, including an IG02610, which is the same filter chip used in the CS-01; and several MN3006 chips, which are digital BBD IC’s designed for use in modulated-delay-based effects such as chorus. None of the IC’s are socketed though, so it could be tough to remove them from the board(s) without damaging them.

The stops and other lever controls all use Alps pots, though all the stops are stepped, which will make them less useful. Plenty of wire in there though! The keys and associated circuitry could also be useful.

The same dead guitar amp from which I had recently harvested a number of pots and dual op amps also contained two CD4053′s (well, one was an equivalent with a different part number, but still…). After getting Loid transferred to the new container, I had stuck one of these on the breadboard for playing around with, and finally got around to putting it on perfboard to become a permanent addition. This is another module which is just wires going right to the pins, and fit perfectly on one end with 4 rows of 3 screws (control inputs A, B, and C, which switch the X, Y, and Z in/outs between two other sets of X, Y, and Z in/outs- I chose to ground the inhibit pin).

Here are a trio of recordings made with the 4053 in various combinations with Loid’s other modules:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

So, the coffee cans-with wires thing turned out to be less than ergonomical. I’ve gone with a new approach, involving plastic Wal-Mart containers and metal screws for a “bolt-bay” patching system:

The mixer & R/2R remain uncased, and the 4019/Switchgate combo remains in the old modem case. I’m finding this much more fun to work with.

Here’s the first video of Loid in action:

My friend Kris recently gave me a couple of old joysticks, the first of which I’ve modified to be used as a CV controller.

The photo below shows the inside of the joystick, before any modification:

As you can see, the X and Y controls are actually slide pots (150k linear, to be precise). Using the bottom pot in the photo as a reference, the far left pin is the wiper. The first thing I did was remove the PCB and the PC connector cable. Then I drilled holes for the jacks, and wired it so 9V comes in via a 1/8″ switched jack, and goes to the far right pin, and it’s equivalent on the other pot, with the middle pin going to ground to form a voltage divider. Since I used a switched jack, I could also make it capable of running off a battery for portable control of stuff like the Modutronic Messmaker.

Here’s a video of the joystick in action with Phoenix:

I have plans to also add some internal LFOs which could be switched in with the buttons, but this is definitely a fun start!

An R/2R ladder is a simple circuit, often used as a Digital to Analog Converter (DAC). There is a good introduction and how-to guide here.

Here’s the circuit on perfboard (the op amp is an LM358):

It may not be immediately obvious how such a thing might be useful in a noisemaking device- I know it wasn’t to me. They say a picture is worth 1,000 words, so here are some screenshots from s(M)exoscope… the first one is a shot of 4 of Loid’s oscs, roughly tuned with two at one frequency and two at twice the frequency:

…and here’s a couple with them completely detuned:

Here’s the obligatory audio example- this is just those same four oscs being fed into the R/2R, whilst I twist knobs:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

After posting about the Switchgate at Deathlehem, the always-helpful Rich suggested a method for extracting a short pulse from an incoming signal- see the thread here. Though that original suggestion didn’t work, it got me headed down the path of breadboard experimentation that led me to this:

I added two of these to the same board as the R/2R:

Switchgate is a simple dual gate based on a 556 in monostable mode (with trigger input), with the monostables triggering two switches. The monostable outputs are also accessible via their own outputs.

Here’s what half of it (a single monostable/switch circuit) looked like on the breadboard:

I put a 4019 quad AND/OR select gate (wires straight to pins) on the same perfboard with the dual gate VCA, and put them in an old external modem case:

I’ve also added a 4051 multiplexer wired straight to the pins to the Melody Generator/ring mod can.

Here’s a recording of the Switchgate/4019  in action with the other modules- the Melody Gen and 4051 were left out:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

Loid is growing quickly… after the first two modules, I decided to build another oscillator board. I had been looking at two things I’d come across at the Deathlehem forums: YAVCO and Ryk’s 4023 Pulse Width Oscillator. Ryk’s osc sounded really cool in the description, but that left 4 unused oscs in the 40106- so, I decided to see what happened if I combined the two. It worked great on the breadboard, so onto perfboard it went. I basically built the 4023PWO, then added two more oscs, which are controlled from a dual-ganged 100k pot. For the second of those two oscs, I added a 50k pot in series, so it could be offset from the main pitch. For the 4023, I wired up the unused NAND gate to be used with external signals. I also added a DPDT switch so the oscillators could be switched between standard powered operation, or the YAVCO-type pseudo-CV, with a SPST in series for the 4023 so that it’s power could be cut completely.

Here are two recordings from the post-build test- the first is showing the whole module in use, with the extra NAND gate being fed from the onboard oscs and one of the oscs from the first osc/divider board:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

The second one is just the PWM osc, switching between “normal” power and the CV input here & there:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

Next, I decided I needed some kind of mixer/output stage. I had recently scavanged a 10-way DIP switch, so I decided to build a 10-input unity gain mixer with a dual op amp. It’s very similar to Circuit 2 from this page at all-electric.com, without the capacitors in the feedback paths. I also used 100k resistors.

The mixer, output jack (with signal LED), and power distribution board are mounted on a plastic tuner-face scavanged from an old stereo. The power board is just a bunch of two-pin headers pulled from computer equipment mounted on stripboard, with the power coming in from a 1/4″ jack mounted on the face of Phoenix. The power supply I’m using there is big enough to handle all of these and several more modules.

While testing the mixer on breadboard, I had come to the conclusion that building this synth in the standard modular way, with nice panels & jacks, would be pretty expensive. With just the two oscillator boards, the Melody Generator, and the mixer, I was already looking at 39 jacks! I’ve seen other people just building everything on perfboard with no panels/case at all, using various headers & board-mounted pots, but I don’t really have an area where I could put such a thing and not worry about something happening to it. After a little thought, I decided to go with a method I’ve used before: plastic coffee cans. I just drilled holes in the sides for the input/output wires to hang through, with heat shrink used in places to keep them from going back through. Each coffee can comfortably hold at least two modules.

This photo shows the Lunettoid modules I have finished so far, with a 4051 on the breadboard:

The can on the left holds the two oscillator boards, the one on the right holds the Melody Generator and two pseudo-ring mods, one using a 4011 (see this page for a schematic) and the other using the TS12A Chopmodder I also used in Phoenix.

Here’s a recording made with the patch shown in the photo:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.