MRCC Freezes when KS37 is connected via USB

[TakeCareOfMyMidi]

Hey guys,

i recieved my MRCC today and it's such a wonderful machine, thanks for creating such an useful device =)

However i'm facing a problem:

I'm using a Squarp Pyramid which is connected via USB to the MRCC as a main sequencer.
I have two Keystep37's which are connected via USB to the MRCC as midi controllers. 
Midi In of each one (sync and start/stop filtered) is going into the Pyramid, Pyramid's midi out is going into the Keystep37's for tempo sync and start/stop (notes, cc and program change are filtered) and also unfiltered to my synths ofc.
The MRCC is always freezing after a while of playing with the KS37's. The issue can be reproduced very easily. Just use the arp on hold on the KS37 and let it play for a while.

Things i've tried:
- Upgrading the FW on the MRCC -> same
- Upgrading the FW on both KS37's -> same
- Using only one KS37 instead of two -> same
- Powering the KS37's with a dedicated power supply instead of power over usb from the MRCC -> same
- Playing around with midi filtering settings -> same
- Connecting the Pyramid via midi -> same
- Connecting the KS37's via midi -> solves the issue, rock stable for ~30mins of flooding midi into my synths via both arps and recorded stuff from the pyramid directly.

It seems like there is a bug or sth like that.
The problem is i really need all the midi inputs (quite large setup) and don't want to buy/use an extra device. A big reason for my purchase were the 4 usb inputs.

If you need further testing, just tell me, i'm happy to help to solve the issue. =)
I'm based in germany, regarding the time zone.

I hope you can reproduce the issue and can fix it.

Best wishes,
Dennis

[Jesse Johannesen]

We have used the Keystep with the MRCC throughout it's development flawlessly and have never heard (to my knowledge) any reports of it acting up, however I have had conflicting reports regarding the squarp pyramid, so that is my guess as to the offender, (but it sounds like you may have ruled this out?). Can you recreate the issue with just one KS37 and nothing else connected. I've sent Steve a message to see if we have a KS37 in house to test with, if we have one and can recreate the issue it should be fixable in theory, if not the chances of tracing the issue gets a lot less likely.
If DIN inputs work but you need more, it seems like a good use case for the Xpander. I hate to suggest buying more things to solve issues, and I'm sorry to be doing it now, however in my experience it is is almost always worth it if the fix makes things "just work" rather than spending hours debugging. I'll report back to you once I know if we have one in the lab and we can discuss it more then.

[Jesse Johannesen]

I just confirmed that we use the KS37 as our "daily driver" in the lab and have had one hooked up to the USB on the MRCC for more than a year straight. Darryl is testing the use case you described now to see if we can recreate it here. I'll let you know how it goes.

[TakeCareOfMyMidi]

Honestly it seems a little odd to me that you never heard about the issue, see: https://conductivelabs.com/forum/archive...-1751.html
Also that the KS37 is your lab's daily driver but it's not listed in the usb compatibility list seems strange to me.
This thread also does not shine the brightest light on the devices USB host capabilities: https://squarp.community/t/hapax-and-mrc...outer/8351

However, i'm here to help solving the issue, not to complain about things! =)

I did further testing without the Squarp Pyramid and only one KS37 and could reproduce the error:

KS37 connected to USB host port A, Novation Peak on midi out port 10. KS37 and MRCC are running the newest firmware.


1. Configure the KS37 dipswitches on the back to recieve clock over USB, set the SEQ/ARP switch on the front to ARP and the DIV knob on 1/16
2. Turn on the MRCC and load the factory default preset with no routing
3. (optional?) Route the USB host port A to midi out port 10
4. Enable MRCC's internal clock generator and set it to a high BPM like 170 (to get it crashing faster)
5. Route MRCC's internal clock to the KS37
6. Press start on the MRCC's internal clock, KS37 is synced now
7. Press the HOLD button on the KS37 to hold the arpeggiator
8. Play a chord on the KS37 so that the arp is playing notes fastly

It usually takes a few seconds (up to a minute) and the MRCC will crash. Maybe there is a internal clock loop going on that will confuse the MRCC?
This is a very basic scenario that you surley tested in your lab, did you? haha, no offense =)

I'm happy to help with further testing if needed and i rly hope that you can reproduce and fix the issue regarding your lab's daily driver.

Best regards,
Dennis

EDIT: I do not want to sound bitter or ironic, but suggesting to your customer to spend more money to work around a bug is something i would not do. However you can send me a Xpander for free if you want. That will also solve atleast my issue, haha.

[Jesse Johannesen]

We're still doing some tests, but would like to find out how is clock being provided in this setup? Is it coming from the KS or through the MRCC? What is the source?

---

We have been able to reproduce the result, by the way, but only when sending the clock to the KS37. We're going to try a few things, I'll let you know what we find.

[TakeCareOfMyMidi]

Hey Jesse,

yes, like i wrote above, the MRCC is providing its internal clock to the KS37 in this constellation.

I know many people with dawless setups an this is the most common scenario:
Providing the main sequencers clock to the KS37 for syncing and starting/stoping the ARP/SEQ. However in this case the MRCC doing the main sequencers job of providing the clock.

I'm really happy you guys could reproduce the issue =)
Thank you for trying so quickly and good luck with finding the nasty bug.

Best regards,
Dennis

[TakeCareOfMyMidi]

Hi again,

any news? Is this something you guys can fix by a firmware update?
I'm asking again because i only have a few days left for sending the unit back if it cannot be fixed.

Thank you in advance!

Best regards,
Dennis

[Jesse Johannesen]

Hey Dennis, sorry for the delay. We were able to recreate the issue consistently when providing external clock to the KB37, but from looking around on the internet it seems like the KS37 tends to lock up from time to time, so I'm thinking that is probably the issue here.
https://www.google.com/search?sca_esv=66...2&dpr=3.13
I think when it crashes it's causing the MRCC to crash as well in this case. I don't think there is anything we can do firmware wise to solve the issue, but maybe Arturia can be asked about it? Have you reached out to them at all?
I think the reason this hasn't happened for us before this is because we're always sending clock from the Keystep and not the other way around.

[TakeCareOfMyMidi]

Thanks for the reply.
Hmmmm, i never ever had the KS37 locking up when providing clock from my PC/Pyramid/Modular/... Not once!
If i follow your google link it seems like there was indeed a problem which was patched by a firmware update -> see release notes.

Sure, i can ask Arturia about the issue.
Using the KS37 for the main clock is not an option because i have two KS37 and i want my main sequencer to provide the clock. It can surely be a workaround but not in bigger setups like mine.
I will fiddle around soon and give some feedback.

Thanks anyway