Modding an Old Rhythm Accompanion Machine (Elka Wilgamat) into a Drum Module

[ninqe]

Hello fellow Elektronauts,
I just started a thread on Reddit (https://www.reddit.com/r/synthdiy/comments/urkik9/elka_wilgamat_turning_an_organ_accompanion_into_a/) regarding this Mod, but maybe someone on here is interested to chime in and share their knowledge with us

A friend and I got hold of a super cheap Elka Wilgamat, a 70s Rhythm and Melody Accompanion with some nice and crispy drums, a pleasant bass and suuuper cheesy melodies.

After having it tested properly we decided to give it a shot to search for Individual Outs of the Drum Elements and by Chance also found the Trigger In of the Bassdrum, which led to the Idea of having all drum parts and maybe even the other elements aswell to be triggered externally and picked-up individually.
Whereas we were lucky with the Bassdrum the other parts (at least the Snare Drum and High-hats) seem to be a bit harder to grasp.

For the BD trig, we simply added a Jack that is connected to R7 and GND on DM1 (see picture 1). For the Individual Out we found resistor R8 to be the one we need to aim for. (Both traceable on picture 2 see orange highlights)

After we were successful with hunting down Trig and Indiv. Out of the BD, we went on to what seemed to be the SnareDrum (SD), hooked up a Jack for the trig to R6 and GND (on DM1 - Picture 1/2) in the hope to trigger a snare drum which is picked up by an Indiv. Out connected what seems to be R52. Here the first problem emerged.

If one compares Pic. 1 and 2 you will find that within the circuitry of the SnareDrum on Pic 2 there is a Resistor labeled R52. However this one is not to be found on Pic.1 but rather 2x R50. Therefore we went on, tried our luck and connected the Indiv. Out Jack onto the lower end of the vertical R50 (Pic1). Triggering went just fine, here it seems we found the right connection. However on the Output side, it seemed as we were only picking up the Trig. Signal as it did not sound like a Snare but rather like an Impulse.

Taking a closer Look onto the schematic on Pic 2 one can see that R6 is connected to a further PCB and thus we were wondering if the problem is buried here. Maybe someone on here has some suggestions on how to approach this?
Furthermore, as all the rhythms contain at least one or two HiHats - which are not to be found upon DM1 we were wondering if these might be triggered by R4 and R5 on DM2 - but if so, where could their Outputs be?
I’d be happy about any feedback if someone is interested to chime in and share their knowledge with us

[CompNutation]

Cool project! Watch out for the lead solder tho…you seem pretty knowledgeable but thot I’d mention it just in case

[ninqe]

actually rather new to modding but some experience with repairs, seemed like an interesting project given its rudimental design! Still lack of skill regarding reading schematics is the biggest hurdle

[Bunker]

From a cursory scan I think the general arrangement is as follows.

The small sub circuits in the boxes with the drum names generate the drum sound with resonant bridged/twin T filters. So for example in the bass drum that would be formed by C2, C3, C4, R5 and R6. The same topology can be seen in the other drum sub circuits. The outputs of these look to sum via a res and cap onto a common output bus (the left hand one of the two vertical lines) which feeds into the lower right hand side of the drawing at point Rp. The other circled and numbered R connections look like trigger inputs.

If you look for transistor Tr14 you’ll see the collector leg isn’t connected. This is classic white noise generator territory. On that basis we can see that the drum sound is generated by the resonant T circuit and the white noise portion is generated here. If we assume the trigger inputs are the circled numbered inputs then if we follow from point R6 we see a trigger applied to the snare drum filter and that trigger is also applied to the base of Tr7 via C1. My guess is the sub circuit containing Tr7 a simple EG/VCA for gating the white noise for the snare sound. There are a couple of other similar circuits and these look to be triggered by inputs R4 and R5. If you follow the output of the noise generator it connects to the common node of C10 and C11. To the right of C11 the R4 and R5 circuits connect and are applied to the base of Tr11. Similarly the output of the R6 circuit connect to the left side of C10 and onto the base of Tr10. This is effectively the VCA part of the circuit gating the white noise. As there are no obvious drum sub circuits related to the R4 and R5 triggers I’d say these are your cymbal and hi hat triggers.

That’s my take on it but I’d love to Hear other folks input? If you look up bridged T filters and the Swing VCA some of what I’ve said might make more sense. Roland used these resonant filters and simple transistor VCA’s in their classic analogue drum machines and to my eye something similar is being used here.

[Bunker]

Forgot to say, all I’ve said refers to the schematic drawing rather than the overlay one.

[Bunker]

For reference these are the circuits I described above.

[darenager]

Do you have a oscilloscope? If so just play the rhythms then “sniff” for the trigger signals, ascertain the voltage and pulse width, then just make a simple circuit to attenuate any external trigger going in, usually a resistor and zener diode will suffice.

If you don’t have a scope you can make a little ghetto trigger sniffer using a 100r resistor and a LED, connect the resistor to one of the LED legs (doesn’t matter which) then add a wire to the other leg and a wire to the resistor, connect the wire of the cathode of the LED to ground on the PCB, then use the wire from the anode to probe for the trigger points. When you see a flash in time with the particular drum sounding it should make it easy to find the triggers. Use a slow tempo to make it a bit easier.

Re Leaded solder, on mods to old gear never use unleaded solder, it is not only crap (IMHO) but the alloys used can react with leaded solder and cause shorts - so called “tin whiskers” which is a metallurgical phenomenon that results in growths of very thin filaments between joints, kind of like rust/mould.

When using leaded solder just don’t put it in your mouth, and wash your hands when done.

Edit: when using any solder ensure good ventilation, ideally use a fume extractor, it is the rosin/flux in any solder which is the problem, not the lead.

[darenager]

Found this, looks like a different version than yours, but it should be fairly easy to translate the mods.

Wilgamat III Schem Markup.pdf (2.0 MB)

Source: Elka Wilgamat midi ? - MOD WIGGLER

[johnl]

uniPulse | tubbutec has mention of another Elka device.

[Bunker]

Sorry if I’m teaching you to suck eggs but I just thought I’d expand on how the drum and snare sounds translate into the final instrument sound.

The resonant filter is designed so that when it gets triggered it ‘rings’ or oscillates at a specific frequency for a specific duration. The frequency and decay (duration) are set by the caps and resistors. It stand to reason that if the resistance values are changed the frequency and decay can be altered however there is an interaction between these settings and values have to be chosen carefully (or lots of experimentation). Because the decay is set by the circuit there’s no need for a VCA.

The noise generator is ‘free running’ in oscillator terms; for an articulated sound we need to generate a simple decay envelope and apply that signal to the control input of a VCA that controls the noise level. A bonus is that we only need one noise generator that can be fed to multiple sources and can be filtered and contoured for different noise based sounds.

If you want to get into the nitty gritty of generating analogue percussion sounds and controlling them I’d recommend searching out the TR 808 service manual. The notes and diagrams are comprehensive and informative.

[ninqe]

Ah really on time that this thread gained traction today, as we decided to spent some time apart from jamming on it again today!
First of all, thanks to all for sharing your knowledge.
I’ll briefly recap on what has happened today:
As it might’ve been noticed, the last time we tried to tap the trig in and indiv. out, we thought that we’ve been succesful, but it turns out we’ve been listening to our own generated Impulse and mistook it for the BD of the Wilgamat.

So first thing today was to take off the indiv out. that was positioned on r8 in the Bass Drum circuit.
Then we placed it onto c7, triggered the circuit on c1 with a DFAM and got a pleasant result.
We repeated this procedure for the Bongo 1&2 as well as the Claves. taping the signal on the C’s before they hit the bus, with success.

I’ll need to make a mouser order to get hold of some necessities and therefore the plan for the next weekend is to work on R4&R5 for the Noises and to figure how to most reliably get a way to trigger both parts of the Snare Drum and tap both parts simultaneously.
Here, @Bunker, I hope I got you right in this respect, for R4,R5 and the noise-part of R6 we would simply trigger a gate rather than the white-noise itself, right?

[ninqe]

Sadly no oscilloscope at hand!
But it seemed we got away quite lucky but the idea with the ghetto trigger might be something that we are going to try if we are up to finding the trigger ins for the bass and accompanion voices ^^

[ninqe]

This one is actually also really insightful, three more instruments in there as well as well as a dedicated cymbal generator ^^

[ninqe]

And also regarding this and your first post! What would be your opinion, what mods could we look out for?
By chance I already enlarged the case of the Wilgamat so there would be some space for some further addition besides the trigs and outs, would be handy to have some further CV in’s (would these necessairly lead to adding vactrol’s into the circuit) or at least some variable resistors at interestng intersections

[Bunker]

I’m glad you’ve had some success; it looks like you’re on the right lines. The other instruments should be available by repeating the process of tapping the outputs before they hit the bus.

Regarding the triggering. I think you may have misunderstand what I was saying. If you look at the drawing there is the R6 port that goes into the snare drum generator. Before it hits this circuit there is a line that goes to a second R6 port on presumably a different board (the dotted lines usually signify that circuit is on a different PCB). That R6 input feeds the gating circuit for the white noise. So when you apply a trigger to either R6 port the drum and snare circuits will trigger at the same time. What you’ll need to do tho is tap the drum circuit output and the snare noise circuit output individually. The two elements of the sound get mixed on the bus along with the other instruments so to isolate the snare sound you’ll need both outputs and a simple mixer stage where you can set the ratio of drum to noise.

That leads me on to mods. Once you have got the multiple individual outs you could easily add a volume level control for each instrument. You could add a mixer stage and have normalled jacks in the input path so you could send an individual instrument out and still create a mix output with individual level control. This can all be implemented with a handful of components on some stripboard if needed. I’d be tempted to send each instrument that you tap off the bus into a buffer stage before you don anything else. It’ll isolate anything after the buffer from the Elka so any slip ups/shorts/voltages etc applied in error don’t fry something on the Elka boards. Belt an braces but for a handful of op amp stages and a bit of stripboard it’s a no brainier imo. Plus the Elka might be funny about having those individual outs too heavily loaded or running into too low an impedance. The buffer stage will negate any issues.

I see you’ve been using the DFAM to generate your triggers. If that’s the case and it’s working then great but if you do get the chance to use a scope if would be interesting the see what pulses the Elka generates to trigger each instrument look like. They may vary between instruments and without seeing the control circuitry I wouldn’t know whether it was a positive or negative trigger and what the voltage and pulse width is. I’d deffo want to be making sure any triggers I apply are the same level, polarity and duration as those generated by the Elka itself, not least because it could affect the articulation of the instrument but it could also damage the Elka if it’s too high a voltage or the wrong polarity. You may be able to see the pulse level with a multimeter but a scope would be ideal. I’d recommend getting hold of one if you can, you’ll wonder how you ever managed before!

Another simpler option would be to fit normalled jacks after the cap at the bus connection. With no jack inserted the Elka functions as normal but insert a jack plug and you break the connection to the bus and send the signal out via the jack. You could possibly add your buffer and level pot before the normalled jack so the Elka has individual level controls that output to its own bus via the normalled jack.

[Bunker]

@ninqe

for R4,R5 and the noise-part of R6 we would simply trigger a gate rather than the white-noise itself, right?

The R4 and R5 inputs look like trigger inputs to gates on the Elka. The noise generator feeds three of these gate circuits in parallel (gating isn’t strictly the correct term more like simple envelope generators and vca’s but you get what I mean?), the R4 and R5 gates and the R6 gate which is for the noise element of the snare. So just by applying a trigger to each of these you should be able to trigger their outputs and tap those outputs in the same way as you have for the kick drum.

[CompNutation]

@darenager

When using leaded solder just don’t put it in your mouth, and wash your hands when done.

It’s super harmful to kids under 6, so I would say you need to be extra careful if you have kids near your workspace and not leave dross lying around and wash down surfaces. When I’m desoldering it usually ends up all over the place.
It is way better for soldering though!

[darenager]

Ah yeah, I would not let any solder come into contact with kids or animals/pets, or any of the other potentially deadly stuff like flux, chemicals, components etc.

[rockpapergoat]

good to know… i have fond memories of soldering stuff as a kid and enjoying the aroma. that might explain a few things now.

[ninqe]

Thanks again, for your suggestions!
As it turns out, we will be able to borrow a USB Oscilloscope from a friend of mine the upcoming weekend. We will then try to as explained by @darenager and you to tap into the OG trigger signal to see and confirm what kind of trigger we would need!
Both of us haven’t used a scope before, but I’m positive that we will find our way

The suggestion with a buffer and mixing stage both sound reasonable and the option to have volume control seems like a pleasant addition as well! Hereby we will also try something we both haven’t done so far, but we are doing our best to inform ourself properly and will try to integrate something in this direction to have it done to our best 'hopefully by-then knowledge’ ^^

Regarding the Mix-Out, I am wondering if it wouldn’t be easier to just use the existing main out? Surely it is not a drum-bus solely then, but by now, we are not strictly aiming to have Indiv. Outs for the Accompanion or Bass stage included in the wilgamat (all of the three sections have a dedicated Vol. Fader on the UI). Whereas especially the latter would be something nice, the whole process around gaining access to it (especially having reliant pitch ctrl.) deprives even more from our knowledge

For the sake of completeness, I’ll add the service manual of our model to this thread for reference later on:
ElkaWilgamat.pdf (5.4 MB)