Search the Community

Hi, as part of my organ conversion I'm reading in the manuals with a DIO matrix module. Because of the weird pin layout of the keyboards and there also otherwise being no reason to use the keyboard drivers, I am using a simple DIN matrix (NGC file is attached). There are actually two DIN matrices. One for the pedalboard which is a 2*16 matrix and the two manuals which is a 8*16 matrix for both. All are inverted due to "wrong" diode directions. There is also loads of other stuff further down in the config file like buttons as well as switches and solenoids + some logic connecting the two. The problem that I'm having is that on the second DIO Matrix I occasionally get a glitch that on the first multiplexer all 16 keys fire a NoteOn event but never a NoteOff event (the lowest 16 keys on the first manual: IDs: 1033 - 1048). The glitch happens irregularly, at the moment maybe once every 4-5 hours (about every 15 million iterations of the midibox). Earlier in the year it was something like every few minutes though I was running a different firmware back then. I first suspected an electrical problem as I employed a different technique at the beginning of my project (crimping) which I later found to be unreliable and changed it (using ribbon cable + IDC connectors and soldering the wires with heat shrink over the solder joints. I re-did all the connections with IDC connectors. That didn't change the problem. I then remembered that the first time I powered up that DIO matrix module I had the DIs and the DOs swapped and thought I might have damaged the chips, so I exchanged the whole board: No change. I have now basically run out of ideas and wondered if any of you had any ideas: electrical/configuration/operating system/could it be software? What is surprising me is that even if the SRs return xFF for a tick and cause 16 NoteOn messages to get fired, wouldn't a reading of x00 the following tick cause NoteOff events to get fired? It seems weird that the NoteOff events simply never seem to get sent. I'd be very grateful for any suggestions as the only thing I can now think of is to go on a wild goose chase and trying to figure out if the special case of my configuration with inverted DIN matrix and the number of shift registers being limited can very rarely cause some kind of exception in the software. Best Regards Mathis ORGAN.NGC