Rollz-5 Schematics

rollz-5-pcb-frontThis is the Rollz-5+ PC Board by Meng Qi.  The circuits follow fairly closely the original paper circuits, designed by Peter Blasser.  I am building it with modifications of my own to make it more modular.   Here’s the back side of the board.

rollz-5-pcb-backI’ve drawn schematic diagrams for all of the circuits and added my own modifications to the drawings.  The Rollz-5 circuits are early, non-voltage-controlled versions of the ones that ended up in the Plumbutter.  While Plumbutter schematics are available (see previous link), I haven’t been able to find normal schematic diagrams for the Rollz-5.  Peter B. provided the paper circuits only, which are pictorial graphics of the circuit layouts.  I drew my schematics referencing those graphics and checking against the Meng Qi board shown here and against the Plumbutter schematics.  (There may be mistakes on my diagrams, so be warned.)


rollz-5-avdogAVDog is made up of three internal ‘modules’.  An envelope generator, made from a filter set to a very low frequency, drives a transistor VCA that gates a simple audio oscillator.  ‘Inpulse’ triggers the envelope.  I’ve added four features.  A dual-ganged pot replaces the ‘x’ resistors, so you can control the period of the envelope.  The oscillator has a separate output jack, plus a switch into the VCA so you can cut it off.  Finally I added an auxiliary input to the VCA.

Also show is one of the 3-Roll Rollz circuits (which is unrelated to the AVDog).


rollz-5-2-roll-lfoI bread-boarded up the Roll circuit and played around with it.  I made 2-, 3-, 4-, and 5-roll versions.  I found that the odd number ones oscillated at ultrasonic frequencies (~100 KHz), and I didn’t want those.  I found that a 2-roll produced a nice LFO.  I added a positive pulse output (like Plumbutter) and an LED driver by Ken Stone.  The two ‘nodes’ still have jacks.  I’m building all eight Rollz identically, but with different value capacitors.  These are the rhythm generators.


rollz-5-gongGong is a low-pass filter, combined with a circuit to ring it.  Interestingly, this gong circuit is different from both the paper gong and the gong in Plumbutter.  The PC board has places for two 2M trim pots, but I will use a ganged panel pot, so that you can adjust the filter frequency.  I’ve also added an auxiliary input, so you can patch any of the other modules through it.


rollz-5-x4-ultrasoundUltrasound is a switch-capacitor filter with an internal high frequency oscillator.  If you build this PC board, you should add the missing 22K resistor to ground to each circuit.  I’ve added an output jack with 10K series resistor for the H.F. output (this is the yellow banana jack on Plumbutter).

 Output Mixer and Voltage Regulator

rollz-5-output-mixerPlumbutter has an output mixer with pots for Gong, AVDog, and Ultrasound that mix each one of the pair to left and right outputs.  Rollz-5+ has 12 audio modules (4 each of those just mentioned), so I designed this output mixer to be similar to Plumbutter.  There are six ganged pots.  Each pot sends a pair of the same type of module to both right and left.  The connections to the mixer aren’t shown on the individual schematics above, but they will tap off just ahead of the series resistor going to the corresponding output jack on the panel.

I’m including a 7809 voltage regulator.  This takes a 12V wall wart supply input.  I am not building the Vactrol voltage starvation circuit that’s on the PC board.


I’m planning to build this behind a clear acrylic panel, so that the PC board will show.  The pots will all be panel-mounted.  More details to follow.



17 Responses to Rollz-5 Schematics

  1. Pingback: New Instruments!! | Aidan Richard Taylor

  2. Pingback: Rollz-5 Assembly Progress | Richard Brewster's Synthesizer

  3. Tuukka Jääskeläinen says:

    Hey there! Awesome that you’ve documented all this, thank you for that!

    Just one question. Why are there still those “node”-jacks in the rollz schematic, and what happens if you put some voltage in it?

  4. Richard says:

    The node jacks on the rollz are the brown jack patch points. Located at the base of transistors, they can act as both output (from previous transistor) or input. Since connected directly to the transistor, patching here could result in damage, but there will be no harm if patching from other outputs on the Rollz-5, because of the 10k resistors on those.

  5. Tuukka Jääskeläinen says:

    Thank you for the reply! So would there be any way to secure these inputs/outputs so that you could supply voltages in there from eurorack clock sources? 10K resistor before the jack?

  6. Richard says:

    If you want to patch into these from Eurorack, yes, a 10K resistor in series would be a good idea. But I don’t know if patching from Eurorack into these nodes would be very useful for anything. I suggest you go to the Ciat-Lonbarde discussion on MuffWiggler and ask this question.

  7. Jesper says:

    Hi Richard!
    I’ve replaced gong trimmers with 1M dual gang pots (with 22k resistors in series – one for each trimpot). When compared to the gongs with trimpots the dual-ganged gongs do not get as low and has lesser decay. There’s also some volume loss. Could this be cause I’m not using 2M’s, or could I have wired the resistors all wrong (resistor goes between pcb and lug 1(A & B)..?
    Any help is greatly appreciated.

    all the best

  8. Richard says:

    Jesper, I just double checked my build. I used dual gang A1M (audio taper) pots with 47K series resistors to replace the trimpots on the gongs. The series resistor value limits the upper and lower frequencies of the gong. See what I said about frequency ranges here.

    Mine went down to 95 Hz (1M + 47k) and up to 2.1 Khz (47k). With 22k series resistors, you would see higher values for both of these, which is what you reported. If you want lower frequencies, change the 22K for 47K or even 100K.

  9. Jesper says:

    Okay! Yes, the gong freq spectrum is definitely higher. I’ll try out some different resistors and compare with the trimpots.
    I’ve been using linear pots.. Do you think they could be the cause of decay loss?
    Anyway I’ll change these and let you know the results. Thanks a lot!

  10. Tuukka Jääskeläinen says:

    In the Plumbutter schematics for AVDog, there are some weird components that look like two circles intertwined with each other. Any idea what components are those supposed to be?

  11. Richard says:

    The two entwined circles is Peter’s shorthand for (I think) a two-transistor current circuit. I remember seeing a catalog of all such idiomatic symbols he uses, but I can’t find it right now. His Wesleyan Thesis contains many of them.

  12. Neil Baldwin says:

    Hi Richard,

    I have an almost-complete Rollz-5 (revision 3 board) that is giving me some trouble. I’ve built a few of them before successfully but this one is baffling me.

    First of all there seems to be a weird power issue. Using a decent 9V supply, the av dogs won’t trigger unless there is something fed into the ultrasounds. If I remove the 4066 ICs the av dogs work OK. Further to this, if I reduce the power input (down to, say, 7V) then the av dogs work. I’ve done the 22K to ground mod (without this the ultrasounds don’t work at all).

    Second, the ‘envelopes’ on the gongs are ridiculously short. I’ve looked at your schematics but I can’t see what part of the circuit forms the envelope. I’ve built it with ‘standard’ components, no modifications.

    What is strange is, if it were just one or two of the av dogs (and gongs) that were misbehaving I’d put it down to an error on my part (bad solder, short, mistake etc.) but it’s ALL of the av dogs and ALL of the gongs.

    I’m very baffled!


  13. Richard says:

    It may be either a bad IC, or a problem with the etch on the board.

    Unplug all the ICs. With power off, check the VCC (ground) pin of each IC for connection to ground.

    Then check power up with 9V battery and check that the mido voltages are 50% of the power supply, i.e. 4.5 volts with a 9 volt battery. AVDogs and Gongs have separate mido voltages.

    Also check that the power pins of all ICs are at +9 volts.

    Adding back one chip at a time, power up and check the power pin on the chip for +9V.

    Finding something wrong during this procedure should point you in the right direction. Try a different chip if one doesn’t work.

  14. Neil Baldwin says:

    Thanks, Richard I’ll give that a try.

  15. Neil Baldwin says:

    Hi Richard,

    I tested the ground on all the IC sockets – all good.

    I tested the VCC on all IC sockets – all good.

    I tested the mido voltages on the av dogs: from your schematics I determined this should be present at pins 3, 5 and 9 of the av dog TL084s? The voltages here were all around the 8.5V mark. 4.17V was present at pin 10

    I tested the mido voltages on the gongs: from your schematics I determined this should be present at pins 3, 5 and 10. These were all correct (about 3.9V anyway – the voltage from the power supply is about 8.7V)

  16. Neil Baldwin says:

    Quick update – once I’d installed the ICs back into the av dogs the mido voltages were all around the 4V mark. I guess the TL802s are part of the circuit that scales down the 9V?

    Anyway, I put the ICs back in and all voltages remain as they were with the ICs removed.

    One other thing I forgot to mention: with the power supply set at 9V, triggering the av dogs produces no audio unless I patch a signal into the ultrasound on the same half of the board. If I drop the power input to 7V triggering the av dogs produces audio as normal without the need to patch anything into the ultrasounds. There’s something very odd going on!

  17. Richard says:

    If power looks OK, then I suggest looking for some board fault or construction error. Double check values of all resistors and capacitors. It’s probably not a board fault, because of the global behaviors being the same. Did you use CMOS ICs from the same batch as those that worked on the other boards? Some CMOS chips just don’t behave correctly in low frequency circuits.

Leave a Reply

Please use your real name instead of you company name or keyword spam.