Therefore, I want to understand it and have some questions. Why does it use the 6V? I guess it is to be able to run from positive only supply, but I am not sure… also I can not see if the opamps are connected to ± 12V or to 12V and 0V (with 6V as “GND “)… if I don’t want the roll function, can I just remove that part? What do I need to change to make it run without the 6V?
So far I think that the upper two lm13600 parts are for filtering the noise, the lower one is for the VCO for the “percussion “ and the rightmost is the overall VCA?
I am happy for more interpretations/explanations also, if you have some stuff on basic drum synthesis, please send
Once I understand this, I would like to make an annotated schematic to help beginners like me to understand!
The latter. There’s nothing connected to the -12 V. This apparently is based on a 1983 design using a single power supply. So the 6 V rail is needed everywhere.
Thanks! Do you think I could then use tl072 instead of the lm324, because with the 24V all would be “farther from the rails”? Or do they have a different purpose?
The 6V rail is already in the schematics, more or less like that.
Surely it’d be possible to adapt the circuit to ±12 V. But some component values might need to change. On balance, building it as is probably would be easiest, especially if you want to use the PCB.
You are right, this would be easiest but I would like to make a new, modified version with this a a template… I did not write this explicitly, I just noticed, but this is my plan understanding the original and then designing my own version with some modifications (so far: no roll, maybe leave out one of the potentiometers, ±12V, maybe distortion,…) I also want to compare with some of the classic 909 or 808 circuits and see what the differences are… yeah, basically I want to learn how to build a drum module “from scratch”
I started simulating in LTSpice, but I am not very good at it.
Here is the part that I think is the VCO (bottom right part, stuff around IC4A in the schematic), but LTSpice says “timestep too small”… maybe anyone with more experience has an idea?
Okai, I replaced the LM324 with a LT1014, because the LM324 spice model did not really work for me…
I also had to ramp up the voltages from 0 for the simulation (startup in the directive)
Blue is Voltage the at the output of the LM13700, green is current at the LM13700 (amp label).
I made a parameter study, checking different “potentiometer settings” (you can not directly make a potentiometer in LTSPICE), but the wave barely changes… Also it looks more like a saw than the expected triangle wave…
Where is my mistake?
Any ideas anyone? @analogoutput maybe? Since you do some of this modeling stuff as well?
Okai, I am getting something in the simulation as well! Had to adjust some values but now it looks good. Also the filter works in simulation!
It’s funny, because I think the makers of the drumbs (syntom) put some effort into making ±12V circuits to work with only 12V and now I put in work to make it go on ±12V again xD
I also plan some modifications, namely external noise (I want to keep the internal, but when I plug in something else, that should be used via switched jack) so with some other form of noise it might be suited for hihats or cowbell or whatever Also distortion! But I don’t want to have too many knobs, so I am considering if I remove the “pulse” knob. In the demo videos it is not really shown how it affects the sound (the pot is turned a little here, but I can not hear a difference). @devicex you have the modules, right? Would you miss the pulse or can you describe how it changes the sound?
sorry I didn’t get back sooner was working late today . I haven’t had a lot of time messing with these modules but from what they say and my very limited experience the pulse is the click or hiss at the beginning of the sound so yeah I guess that would be the " noise " still seems like a lot of work to change something that already works .
Oh, I am still simulating in LTSpice, but it looks good so far! I have a VCO and a noise source which I can mix and then get that into a state variable filter (lm13700 design). Very similar to the Thomas henry mega percussive synth. This then goes to a VCA which is controlled by a simple decay envelope. The envelope also modulates the vco freqency and filter frequency (with attenuation pots, want to replace them with attenuverters). I will share the current schematic and example sounds (simulated) later! with the noise added the simulation is very slow and also turning the pots in software is a bit tedious so I will have to go to breadboard soon!
(For people considering moving in the opposite direction, here’s a video from Julian Ilett where he’s building the original Syntom, the predecessor to the drum voice drumBs is based on. There’s a link to his PCB in the description.)
Yes and no I mean it’s still the VCO , Noise and filter and envelope from the original drumbs, but with slight changes… (I guess you can say that about all synths xD )
Okai and the distortion added… (which I still need to do in the schematic/simulation) I mainly replaced parts which I could not get working in the original version in LTSpice (probably because I got some stuff wrong when changing from the 6V as GND to 0V as GND) and then I wanted a square wave and attenuverters and the lowpass, because I saw that on the DFAM let’s see if this will end up as a working module… I need to start breadboarding!
edit: here are some simulated sounds. The high pitched noise in between has to do with the simulation having some trouble to get started.
I’d have to simulate this to be sure, but I think you will find that topology fights against self oscillation, disconnecting the feedback from the negative input of the first OTA should change this. I could be way off the mark though as I’ve studied a few lm13700 filters I’ve yet to actually build one.
Not sure what you are trying to achieve with this filter design, for drum circuit it’s probably overkill, but many LM13700 designs swap the internal transistor buffers for a tl072, there is also the option to buffer the feedback path with diodes … see the Rene Schmitz ms20 filter design if this interests you.