DIY stripboard Roland System 100 ADSR envelope

System 100 ADSR front panel

The original envelope on the System-100 keyboard wasn’t really one of my favourite things about it. There’s precious little resolution on the decay slider for tuning in those tight bonky basslines, so I tend to end up moving the slider by infinitesimally smaller amounts to try and get the sound I want.

I should point out that it’s a different circuit to the possibly legendary System 100m ADSR envelope (“the snappiest adsr eg of the world…”). For once the hype is deserved, my stripboard 100m 140 clone felt super-snappy, with satisfying control ranges.

System 100 ADSR schematic

Let’s build the System-100 version anyway, maybe we can use different potentiometers or swap the timing capacitor size for something a bit smaller.

The rare-ish programmable unijunction transistor N13T1 was swapped for a 2N6027, although it has since been possible to get hold of the originals on eBay. The 700 LFO worked better with the N13T1, so that might be case here. I used 2SA733 and 2SC945 transistors as per the schematic, albeit the -GR variant rather than -Q, with 1N4148 diodes standing in for the 1S2473s that are splattered throughout.

System 100 ADSR stripboard

The original never quite fully opened the VCF so I’ve added a single opamp to optionally boost the level from 6v to 10v.

Here’s a video of the envelope in action – I left the text on there to keep the camera from trying to going apeshit trying to autofocus on the trace. Not being a storage oscilloscope makes it trickier to track slow moving signals but you get the idea.

In this test I’m using a 3.3uF tantalum for the timing capacitor as standard but smaller pots than the original: 500k for the attack, 100k for the decay (which is too small, really), and 500k for the release, all audio taper. At about 0:43 the sustain pot is turned to 100% which causes the voltage to ramp up slightly rather than staying level. I’ve not looked very hard into fixing this – I’ve just been turning it up to just below maximum.

Here’s the layout and the DIY LC file (without the 6v to 10v boost) should you feel in the mood to torture yourself with some stripboarding.

Roland System 100 ADSR envelope stripboard layout

Having been through all that, if you want tight and snappy then the 100m envelope is probably a better bet.

Due to other things (work, house move, more work…) the System 700 envelope PCBs I made have been left untouched in a box since March, I’m looking forward to getting those going and comparing them to the System 100 and 100m… eventually.


  1. 7th September 2019Analog says:


    could you send me the envelop Schematic for the system 100 envelop?
    the one you made in your cad programs, that is probably better and easier to read.

    kind regards

    i have the Frequency Central Envelop, 2 x … but i don’t get a decay or sustain like you in the video. for me it’s like decay and sustain are non existent. i tried to scope again today, but it looks more like a Befaco Slew limiter. Attack and Release.
    nothing close to what i see on your scope.


  2. 7th September 2019Analog says:


    could you send me your System 100 envelope schematic?

    i have 2sc945, 2sa733, 2sk30.

    Kind regards

  3. 7th September 2019ua726 says:

    If you’re talking about the Frequency Central stuff I’ll presume you mean System 100M – the System 100 had a different envelope design (as I mentioned in the post).

    The schematic in the service manual is pretty clear, I didn’t annotate my version of the 100M envelope schematic particularly well (and it has some extra bits for LED indicators) so it’d just be confusing.

    There’s a couple of versions of the service manual around – this one here is as good as it gets.

  4. 10th September 2019Analog says:

    Hey, thnx.

    i didn’t knew that.
    could you write me an email, please.

    so i can send you a PDF with the design i made in EAGLE … to see if i have done it right.
    i have name all the parts the same, including R1, but i did mirror the whole.


Write a comment: