Progress on that slowed after i was satisfied that a pattern could be quickly erased to a given length. I wasn’t up to speed at the time that the UI enabled pattern erasing in a simpler way than i had appreciated (i.e. that priming step mode record would wipe and re-build up steps, this was ultimately convenient enough)
I’m recovering the machine i did this on and awaiting a new hard drive but I do not recall that the device needed any priming per se when sending a sysex string. iirc the mcc was sending a few strings out at launch or repeatedly and i wasn’t sure what the device was responding with (may have been a firmware version test or similar) - i think it was just a fair amount of tedious reverse engineering (i was also aiming to set up the cv stuff iirc) - i’m not sure if my strategy was to send empty (zero vel) sequencer steps back into the wiped sequencer - it was pretty much the last time i tinkered with it when i was writing above - i may get back to it one day
Perhaps check that it’s not a matter of mixing up hex / decimal depending on how you spied on the data or how arduino is programmed
Sorry i can’t help more, i guess it’ll click eventually - as ever, try to test the arduino output against a simple use case (maybe not with ks) to test out the midi/sysex coding/hardware etc or explore whether it needs to be din/usb or whether ks needs external power
so aim to monitor the arduino output in a midi monitor to compare with the output from ks and so on
arturia support may have info