I have built the drumBs Circuitbenders - drumBs analogue percussion PCB and they are great , but I also have the BMC18. ANALOG DRUM drums and I like the sound of the BMC drums better while not as fancy as the drumBs it is easier to plug and play [ get a decent sound ] out of the BMC drums . just my 2 cents your mileage my very …
3 Likes
Interesting! Where can I connect the feedback to if not the inverting input? In the MS20 I just saw that the feedback from after the first OTA goes into the inverted input as well ( is this also a bandpass signal there?) but the feedback after the second OTA (low pass) goes into the cap after the first OTA…
I need to read about state variable filters again!
Or I might try to use the diode filter by @moritzklein here! But I am not sure if it can simply be switched to band pass…
Too many decisions about stuff I still don’t really understand!! XD
That is also an important point!! Versatility is not everything! If it’s hard to get a good sound out of it, the design might be too complicated… in this case I find it strange though. The drumbs is almost identical to the BMC , but has the added noise section. The circuit for controlling pitch and the vca is different, maybe it’s that? 
1 Like
Not sure what “MPS” in your schematics refers to but the MS20 filter circuit, as used by Schmitz and LMNC, consists of two low-pass filters in a row. The actual MS20 has two copies of this circuit, with different configurations.
I found this:
By increasing the amount of resonance, these filters are able to peak at the corner frequency. However, by the lack of sufficient phase shift in the resonance feedback loop, they are incapable of self-oscillating
http://www.soundsemiconductor.com/downloads/AN701.pdf
The behavior of a filter can be significantly changed by feeding a certain amount of a filter stage’s output into the input of a preceeding stage.
If the total8 phase-shift in this feedback loop is an integer multiple of 360°, at one or more non-zero9 frequencies, and the total8 loop-gain at these frequencies10 is greater than one, then the filter will self-oscillate at these frequencies.
Very interesting stuff!! Still need to digest it xD
MPS is mega percussive synth 
https://www.birthofasynth.com/Thomas_Henry/Pages/MPS_project.html
Ah, ok. The description there explicitly refers to the LM13700 data sheet which has this circuit:
while the MS20 filter is similar to this one:
1 Like
Your circuit differs from Thomas Henry’s MPS design, your resonance control pot goes to the centre of the circuit where as his goes to the start. He states that his circuit can self oscillate and this is dampened somewhat by R46 on his schematic. I know Thomas Henry’s explanations can be long and difficult but there is a lot of information tucked into them.
Hm, I can not see the difference… the lowpass in his schematic goes through R87 to the inverting in of the first OTA and the bandpass goes through the resonance pot and then a 100k (R60) to the same in…
Something else I have not yet understood is that the resonance increases as I reduce feedback from the band pass out into the first OTA…
not that it is hard to get a good sound from the drumBs , I just prefer the sound of the BMC module . maybe that will change as I use them more my use is still limited .
2 Likes
psst … here’s an improved multimode version. i’ll do a video on it soon. https://tinyurl.com/yf9klq7z
you can switch on the different modes either individually, or mix and match them for whacky results.
7 Likes
Your RV1 is connected wrong, it should mix between band pass and ground. That was throwing me off.
Also why halve the resistor values? That will likely have the opposite effect to what you intended?
Ah! Now I see it! thank you!!
I got it correct in the LTSpice, though:
For the resistors: I just changed them around a bit, so see if it makes any difference, I will go back to original values probably.
Or maybe I will just go with the ladder filter by @moritzklein The video about the simpler version is awesome and I really like the sound!! Now as multimode it’s even better! Need to think about if I want to make all the modes available or just low and bandpass (maybe also highpass, if these crossover type DPDT switches are easily available) I don’t want to make my panel too complicated
2 Likes
you could also use a rotary switch that normals the unused positions to a common pin (which you’d then need to connect to ground), that should save some space on the panel – though they are typically quite expensive!
2 Likes
Good idea, but I noticed that you need to ground the inputs if you don’t use them, otherwise some wild things happen when you turn up the resonance (I checked with your simulation). I wonder if that is possible with a rotary switch… I can not decide if I want just a simple lowpass/bandpass switch or if I make three switches, like in your circuit. The easy/simple way or the crazy sound way?.. xD
1 Like
Still on this, but with a 3 year old and a 3 month old there is not much time I think, I will make three switches for the filter like in the simulation. Today I found this interesting distorting mixer:
https://learningmodular.com/exploring-the-moog-cp3-type-mixer/
which I would like to use to mix the VCO and the noise part. It uses 12V and -6V, so I need to find a LM7906, I guess, or try if it could work with -5V LM7905 (which I still have around from the VCLFO)…
1 Like
Some progress: got the ladder filter into Kicad:
Need to think about how to organize the schematic, because I will have to use a separate control panel, I guess… What is a good way to indicate that the pots will be on a different board and connected via headers while still showing that there is a pot there?
Also worked a bit on the panel again:
6 Likes
I think I have got most of it ready in the schematics now… Board interconnections are not yet done and I noticed that I can use one less opamp by using the second noise opamp for the decay vactrol control.
Oh, and I want to change the power section to use diodes as well, but I need to check how to do that again 
It’s not very similar to the drumBs anymore, only the VCO is still close (but on ±12V now and I added an expoential converter for v/oct), then noise from the bastl noise², the envelope from barton musical (BMC018, now with decay CV) a moog cp-3 mixer which can distort, diode ladder filter by @moritzklein (the updated version with HP and BP as well) and the VCA from BMC018 again.
Will this work? Who knows?! 
Will it sound any good? I have no idea! 
But I hope so and at least I learned a lot already!
Any glaring mistakes so far? Other feedback?
3 Likes
Yeah, go ahead, do that, make me look bad 
1 Like
Any reason for R34 ≠ R36?
They’re usually the same, to minimize offsets (an alternative is a small trimmer with the wiper connected to ground, as in the datasheet).
3 Likes