• MatriX
  • Matrix continuously sending CC64

Hi. I noticed this sometimes, but i’ts getting worse.
Loopa and SEQ4+ are wired both into Matrix, issues appear only on SEQ4+, no matter if I switch the input ports.
Now Matrix is almoust continuously sending a “CC64” signal to the SEQ. They are still working together but the connection sometimes is lagging by beeing flooded wth this signal. It is displayed on the screen, it shows up in SEQ midi monitor as B601 or B701. It also can be triggered by the slightest touch of the joystik. Not by touching any contact on the board (resoldered all contacts on the board already). In test mode all joystik actions appear ok.
In test mode, another observation annoys me: every exactly 5 seconds the connection indicator switches from green to red and back, doing this the monitor shows “sysex”. Although in working mode the connection seems to be stable. As i sayd before, it only appeares on the SEQ connection, Loopa works fine.

    @Synthfox thanks for the report!

    This may be possible if there’s some noise coming from the joystick - if the ADC sees two different analog-to-digital converted values in short time, it will send a CC update.

    As you wrote it got worse over time, i think your joystick pots may be gathering dust.

    You may know this behaviour from some synths (Alesis Andromeda A6 being a prominent example), where scratchy pots can send lots of unwanted updates.

    Also, as the MatriX joystick is much more sensitive than the sliders, this is more likely to happen here.

    But: as we want to do better than the A6, i am thinking of adding a software filter to the SEQ v4+ mode of MatriX - this problem should be entirely solveable in software - i’ve put this on the list.

    The test screen showing reconnects is normal - this better visualizes the 5 second “heartbeat” between SEQ v4+ and MatriX which would otherwise not be shown - we’re implementing TK’s BLM16×16+X protocol which works identical - during normal operations there is no visible dropout as the reconnect happens instantly, while the MatriX is still connected to SEQ v4+.

    Best regards,
    Peter

      Hallo. That this 5 second signal may be the “handshake” was withinin my considerations. But the joystickis brand new, had this behaiviour from the beginning. A slightest dip on the board nearby the stick triggers this noise too. Has the Loopa this filter already, cause no strange behaviour occurs when switched to the Loopa input?
      Whishing you the best:
      Reinhard

      @Synthfox - no problem, it makes sense that the LoopA mode does not show the symptoms, as on most pages it usually does not transmit the joystick position data as CC data but only moves the active window around. There’s also a bit of “deadband” configured, so that the window does not drift - this should filter the noise.

      But you could check LoopAs “KEYS” page where you can map the joystick to output MIDI controls (i.e. pitchbend) and see if this also transmits a lot of messages?

      For the SEQ v4+ mode, I’ll add a software filter and will let you know when the new version is available - this should help also for other users, that may be affected - in the end i think just something adds a little bit of noise on your ADC - it might be different cable lengths, a noisy PSU or some other unknown external factor.

      In any case, you are right, this should not saturate the MIDI bus with CC messages, so i’ll target this as an important update and report back to you once i have the new version available.

      Best regards,
      Peter

        14 days later

        @Synthfox - as promised, wrote a filter routine to get rid of the unwanted CC message storm in SEQ v4+ mode of MatriX, if the ADC is just slightly jittery - here you can download the new version v1.00b6 to test on your MatriX:

        https://www.midiphy.com/en/matrix/

        Thanks for reporting back after testing!

        Best regards,
        Peter

        4 months later

        Hello Matrix Friends,

        since a few months i have the same problem. The firmware upgrade of Matrix and Seq4+ doesnt helped in my case.
        I have tried today to clean the poti’s with Tuner-Spray. That doesnt helped.
        Does anyone can help?

        Best regards and thanks for reading.
        Marty

        MartyMcFly Hi Marty,
        no worries, we should be able to find out what’s going on - i do suspect a slightly noisy ADC (i.e. due to external factors or even the PSU itself), it could be well fixable in software, but i’d have some questions up front:

        a) can you confirm your MatriX is running 1.00b6 (it shows the version number when booting)
        b) is it only happening in SEQ v4+ mode? What happens when you connect a LoopA and go to “keys” mode - do you see some messages sent as well?
        c) could you check out the Test Mode screen of your MatriX? What are the readouts of the joystick and the sliders showing? Are any values “jumping”?
        d) which PSU type do you have? Are the CC64 messages happening at every MatriX brightness level? If the PSU is loaded only partly (reduced brightness), the analog-digital-converter might be more noisy.

        Thanks a lot and best regards,
        Peter

          Hello dear Peter,

          thank you very much for the fast response.
          a) i have the 1.00b6 installed this morning. hoping to loose this strange behavior.
          b) it is happening in both modes.
          c) in test mode there are several values jumping in different times
          d) i use the psu you told me… :-) it was working without issues for several month…
          and the problem persists on any brightness levels.
          e) in test mode with no other hardware connected the jumping values persists.

          best regards
          Marty

          i also rotated the plug of the psu in the primary side for testing with no result.

          absolute strange is that in the moment i cannot reproduce the problem ?!?!?

          while further testing the problem appears again.
          perhaps it is the psu…. changing to 100 percent brightness seems to reduce the problem.

          MartyMcFly @Marty - if it worked nicely for a time and it now doesn’t, i’d also first suspect the PSU - over here, three ThinkPad PSUs (with a similar form factor) failed in a timeframe of just three years.

          How big are the value jumps in the test screen and are all sliders and the joystick affected similarly?

          You could also try to completely unplug the MatriX from SEQ v4+ and LoopA and check if the problem is gone - a bad PSU on either of those might also cause it, maybe Andy has further suggestions as well.

          Best regards,
          Peter

            Hello Peter,

            in the moment in the test screen the jumps are just one plus or minus. and it jumps when i connect the seq4+.
            with no connection to loopa and seq4+ the jumps are gone.
            and yes i also updated the seq4+ to the file on the matrix page…
            in the test screen the blm connected disconnects every 5 secs.

            best regards
            Marty

            @MartyMcFly - Have a good new week start!

            Could you try to find out, if it’s coming from the LoopA or SEQ v4+ connection and potentially change the USB power supplies you use to power those? If they are noisy USB PSUs, this would explain the behaviour.

            But even if SEQv4+/LoopA USB PSU changes won’t help, it is good news, if these are only +/- 1 value jumps - we can filter this in software without problems. The faders and joystick have way higher resolution than required for normal MIDI CC output, so you’re not losing resolution - if the SEQ v4+/LoopA PSU swap does not help, i’d add this filter to the list for the next MatriX software release.

            Best regards,
            Peter

              Hi Peter,

              after many testings here are some news…
              From the LoopA Input it is enough to plug the cable into the Matrix… with loopa keeping disconnected.
              It is a long wire…

              On the Seq4+ Connection the problem begins with connecting the Seq4+ to the computer… No Matter wich USB Hub i use. I tried several hubs.
              While connecting only the Seq4+ to the Matrix and powering the seq4+ via an powerbank it works fine.
              But - that cannot be my solution because i hardly use the usb midi connections as well…

              It seems that the Matrix gets the noise over the BLM Ports.

              Best regards and thank you very much for your patience and help.

              Marty

                Hi Marty,

                thanks a lot for your tests - there would potentially be hardware options to improve the ADC situation - Andy sent this list for reference:

                • touch up the soldering around points the headers,
                • disconnect 0V from axes-neo or
                • add the BAT diode

                but as the ADC noise level is quite minimal and as i think you would not want to open up your MatriX, let me add a software filter for you - i’ll report back, when the software is done - please give me a few days, lots to do at the moment.

                Best regards,
                Peter

                  hi dear Peter,

                  thank you very much for the work in the software!
                  I would be happy to get this solution. ;-)
                  I will be patiently waiting for it!

                  Thanks a lot!

                  Best regards
                  Marty