Mod that thing and take it way beyond the 101…
A couple of things you can easily mod:
- Faster ON portamento
- LFO as master clock for sequencer/arp
- LFO reset input
- a whole bunch of I/O for use with modular synthesizers
- VCO synchronisation input and linear VCO FM input
- different filter types (yup, including 6db, 12db, 18db and BP)
- change envelope speed (see also bryan’s cool measurements here: https://www.muffwiggler.com/forum/viewtopic.php?f=4&t=236634&start=25)
- amplitide modulation for the VCA
- VCA overdrive
- and a couple of improvements, such as de-clicking the VCA on gate and a cleaner audio output
You can find some sound demos of mods here.
MS-1 Modification Updates
I wrote a quite comprehensive guide for modding the MS-1 September last year. One or two things in this guide need updating (mostly corrections of typos and the updated RD8 trigger converter, see below), so eventually there’ll be more. Here, I’ll focus mostly on a couple of things that are not yet in the PDF guide.
MS Bouncing Castle Mod (March 2021)
Just followed a convo on MW and felt prompted to finally follow up a bit on the more widely shared impressions that the MS-1 could do with a bit more rubbery, spongy qualities in terms of how its envelope, filter, and VCA interact. tested around with introducing a little bit of slew here and there (I just like putting in random caps, don’t I :), and got the following, which declicks the VCA a bit and also slightly alters the impact of the envelope on filter cutoff.
What you do is simple and, if you don’t want to change the Envelope timing capacitor for faster “snap”, can be done from the back of the PCB alone. Wire a 220nf cap between the VCA select swithc (GATE/ENV) and ground. Also wire a 560nf cap between TP7 (that’s the envelope output) and gnd for introducing a bit of slew on the ENV. The slew on VCA softens the note onset and lenghtens release (even on GATE) a very tiny bit, so you get rid of clicks. This additional bit of “smear” on ENV to filter further countercats the staccato tendencies of the MS.
I also recommend changing your env timing capacitor to 2.2uf since this gives you quite more snap and to use a VCA direct output /described below) rather than the normal output, which can crunch in a quite displeasing manner when VCO levels are up.
While you’re at it, also reduce the VCA SIGNAL IN attenuation a bit, i.e. drive the VCA harder with the filter output.
White noise that sounds more like on the SH101
The white noise on the MS-1 is brighter than on the original SH101, so if you want to get it so sound closer (i.e. a bit duller), you can just solder a 100nf capacitor between the cathode (negative leg) of C63 and ground. Although I passed on my SH101 yonks ago, I adjusted the amount of dullness with the Roland SH01A “OG” noise as a reference. If you want to retain the option of “modern” vs. “vintage” (as on the SH01A), you can just wire an on/off switch between ground and the respective leg of your additional capacitor. I personally keep the duller version permanently.
Replacing the portamento with analogue portamento ON
First of all, never mess with the existing portamento switch since this is partly coupled to the octave switch and cutting traces there wil mess with the CPU control! Best install an additional pot/switch.
Portamento is handled by the CPU on the MS-1. Although you have familiar controls (switch and time pot), these control software, not an analogue circuit, which would be neglegible if portamento on the MS-1 were of a decent speed.
An initial fix I found was to introduce a little slew (by means of a 0.1uf capactitor) at IC23, where the pitch CV from the CPU to the VCO is filtered and buffered. For completion’s sake, here’s a picture for the point where you can do this (connect to only one of them). The disadvantage of this method is that you have only one porta speed, which corresponds to the value of your cap.
Analogue portamento ON with variable speed
For variable speed, you need to do a little bit more – remove one resistor and implement a little active circuit using an op-amp, two resistors, a potentiometer and a capacitor. I implemented the circuit mapped here: http://audiodiwhy.blogspot.com/2019/11/glide-circuits-part-i-portamento-basics.html. For values, I used a 250k pot and a 2uf film box capacitor, which I find gives me a really good range and action. (Actually, I found a 1uf cap also really cool but this incidentally gave me the old Behringer pot action with not much happening in the first 5% or so, lol..)
First, you need to break the CV connection from keyboard and sequencer. The pitch voltage from the CPU is filtered and buffered at IC23 with R242 an ideal candidate for easy removal. You then need to wire up your slew circuit as shown in the audiodiwhy page (or any other variant, there are some out and about), and feed the output of that through a 47k resistor back. Slew in is connected to the right hand terminal of where R242 used to be, slew out to the left hand terminal. Supply voltage for your opamp can be had at C2 of the bender PCB, which carries +14.11V.
Getting carried away I went so far as to replace the existing portamento pot for the analogue version. Unfortunately, the porta on/off switch cannot be replaced that easily.
Autoporta (still needs testig)
The circuit linked above gives you always ON portamento. For legato (autoportamento) you need additional circuits. Helpful info about this can be foudn here: https://www.muffwiggler.com/forum/viewtopic.php?t=231189.
External LFO Reset Input
This is a classic as offered in the circuitbenders.co.uk SH101 mods. You can wire up a connection to an external input that resets your LFO. This deos not sync your LFO to an external clock but just restarts it’s phase when triggered, which is cool for rhythmical effects and also affects the seq/arp when you have the LFO master clock modification.
For wring this up, connect the tip lug of your jack socket to a diode (anode to jack socket), the cathode of which goes to a 100k resistor, which, eventually goes to the point marked above. Usually, you would send +5V triggers into this but gate signals also work. NB that as long as voltage is high on your connection, the LFO is “frozen” and starts cycling again when voltage goes low. Here’s how it sounds:
Different LFO Waveforms on VCO Pulse Width Modulation
This is a classic mod for the 101 and somehow I just got round adapting it to the MS-1 now in late 2020. It’s easy to do and widens your sonic palette quite a bit.
While PWM in the 101 and MS-1 per LFO are driven by the triangle wave only, it’s fun to have the other LFO waveforms available as a modulation source. For doing this, you need to lift R117 (82k) and wire TP4 on the back of the PCB (this test point carries the selected LFO waveform) via a 82k resistor to the right hand terminal on the PCB where R117 used to be.
Even more fun is to have a switched jack socket in between TP4 and your new R117 (TP4 connects to normal pin, new R117 to tip pin), so you can plug in external modulation sources.
Clocking Sequencer and Arp from the internal LFO
This was a bit of an on/off thing, so apologies if you already implemented an older version and the newer one is cooler/more elegant. Both work.
One massive oversight in the sequencer implementation of the MS-1 is that you cannot use your LFO as a clock for the seuqencer/arp as you can do on the SH101 or Roland’s boutique SH01-A. If memory serves, an earlier firmware of the MS allowed you to play the LFO in sync, but this was undone for the sake of two separate tempo controls. Behringer might change this or not; I’m not realy willing to wait/trust them in this regard any more.
When I started playing with clock-related aspects in the MS, I first used the square LFO, buffered and rectified, on an extra output socket, which could be plugged back into the external clock input, which worked/works very well and has the sweet side effect that you have a clock to drive other gear as well.
In the meantime, I found an easier version of this: you can tap the LFO clock directly either at the right hand terminal of R61 (next to the LFO LED) or from the lowest pin of the external clock input jack socket shown on the picture below. You can use these signals on an extra output socket (best buffered with an op-amp, or at least protected with a resistor and/or diode), which gives you clock for feeding back into the MS or clocking other sequencers/drummachines.
Version with switch (requires some trace-cutting):
Finding a solution here was one of the things that make me feel really dumb and took a lot of trial and error (shoulnd’t though…) All clock signals in the MS-1 go though the external clock input socket. The signal from the CPU/Midi is carried through the normal pin of that socket (i.e. mechanical contact to input pin when no cable is plugged). Oddly, the clock from the LFO (yes it exists!), is present at the lowest pin of that socket. What you can do is to cut the trace between CPU signal and normal pin (the continuous black line from the pin in the middle) and wire up both ends of the conncetion you just broke with an ON/OFF switch. When the switch is off, the clock from LFO determines the seq/arp speed, and when you swich on (i.e. restore the original conenction), your seq/arp speed dial determines the speed. When you plug external clock in, external clock speed sets the pace, irrespective of your additional switch.
Recently, I got into Kicad more, so one byproduct of toying aroudn with schematics is that I traced the MS-1 filter. So, here you go:
Interesting to see that they used fet buffers AND the LM13700 buffers. More like the SH09 filter than the 101, no? At any rate, as for last millimetre sound differences, I personally would rather look into the VCA implementation than the VCF in this particular case.
Different filter types
The VCF on the MS-1 is a redesign of the SH101 circuits, based, like the frequencycentral modules, for instance, on two LM13700s. This makes it easy to tap each filter stage output at the buffer outs of the LM/VM13700. In the video above, I took the 6db out, attenuated it a bit and fed it into the VCA input in addition to the 24db output. More details in the PDF and soon here too (with pics).
Main out crackle fix (final op amp too driven too much)
Some mods of your VCF or VCA can increase the output level, which, under circumstances can drive the final op amp of your MS a little bit too hard and result in clipping. It appears that some users of unmodded units also report on crackles, especially when all waveforms in the waveform mixer, including noise, are on full tilt. You can counteract this by reducing the VCA output volume a little before the signal enters the main volume pot and the final op amp after that.
On the bender board, C3 carries the VCA output signal to the maikn vol pot, which then feeds into the op amps underneath R2, and R3. If you take out the anode (lower leg) of C3 and insert a resistor the lifted leg and its terminal on the PCB, you attenuate the signal. As for concrete resistor values, check 5k and if this is not enough, increase the resistor values. Since I drive the VCA in my unit quite hard with various filter mods, I use higher values. As for implementing this mod, best take out C3 entirely, and wire everything up from teh back of the PCB, so you have these parts more easily accessible.
Using direct VCA output for cleaner signal
The final mixer stage of the MS-1 (after the VCA) is not exactly clean. If your unit is like the ones I had so far, you can hear some hiss on notes when all waveforms are on zero level (and filter is down, if memory serves). I notice something similar on my MFB Dominion 1, yet far quieter and only on gate on/off spikes.
Testing a VCA direct out for different purposes, I found this noise gone, so I assume the hiss happens somwehere between the VCA and the LM4558 op-amps in the final stage.
For a direct VCA out, clean and strong (that takes VCA overdrive much better than via the usual audio out), just wire a jack socket to TP6 on the back of the PCB, done.
RD8 to MS1 trigger converter (obsolete as per new MS-1 firmware)
Behringer MS-1 users reported that the RD8 does not allow for triggering the MS-1 sequencer in the good old Sh101 fashion (where a trigger input forwards the sequencer by one step), which was a bit of an oversight by the respective engineers. I didn’t have a MS-1 at the time but became curious and invented a little passive trigger converter that was quickly adoped, adapted and improved as regrards capacitor values (I had used a 1uf first in my hypothetical mockup, but 0.1uf works best). Basically, you can prolong the trigger spike sent from the RD8 by using a diode and a capacitor, turning the spike into an RC envelope, which the MS clock input “understands.” A nice side effect is that this also works for triggering envelopes on my Model D, so band-in-a-box there you go! (BTW thankfully the TD3 cooperates with the RD8 triggers as they are out of the box.)
When using a small ceramic disk capacitor, you can even put this inside a cable. Be mindful of the direction of your diode though (black marking in direction of MS1 input).