me is extending my old Core32 SEQV4 for moaaar MIDI ports…
Are there any known issues with the old IIC MIDI modules (MIDI out only) in conjunction with an old Core32/STM32 SeqV4 (running @ latest firmware, compiled with old toolchain, not newlib-nano enabled)?
I installed four IIC MIDI modules, jumpered them to different IIC addresses and the SEQ only recognizes the first one, sporadically also recognizes the second one (in about one out of ten boot times). The 5V voltages seem ok (stabilized 2A 5V PSU, got around 4.75V at the MIDI IIC modules).
I swapped through the IIC MIDI modules (installed every one as single first IIC1 module), and they all work individually.
The IIC bus is also wired correctly, imho… Has anyone got any idea?
It is not really important, I managed to get IN3/OUT3 and IIC1 running and thus already have two more out ports, am just curious, what is going on .
are you up to date bootloader/software wise? I seem to remember something breaking IIC if you had a firmware with old BL (mine is STM based with 4xIIC and works fine)
Thanks for your answer - yes, the bootloader is up to date.
It is really not important, please don´t pull your hair - will upgrade to the new STM32F4 based core and check IIC MIDI again, when it is available - 192KB of RAM will allow for more SEQ goodness, maybe even 1024 step patterns as a Christmas present?
All I can say is that it used to work on my side - that was before I made the switch to LPC17 based core, so I cannot test anything with current firmware versions.
It is good to know, that it should work, maybe it is because of my adaptations to get the VFDs running which required some (probably improperly done :)) driver hacking or the long SRIO chains I am using, or the old compiler toolchain i am using (did not manage to update yet, am compiling on FreeBSD, which complicates things :))…
But no worries for now - will wait for the new core and then update everything properly :-). Am really looking forward to the STM32F4, awesome specs, cheap price
thanks for your time and your thoughts! I have four separate (old) MIDI IIC modules from SmashTV, all jumpered to different IIC addresses.
When I install two modules on the IIC bus and power the SEQ up, normally only one (IIC1) gets detected. I can successfully output MIDI over it. When I remove this module from the IIC bus (and don´t change the jumper setting on the other module, its IIC address is not 0x10), the other module gets detected anyways and appears as IIC1 module.
Now here comes the mystery: when the SEQ has been powered off for a prolonged time (say at least one minute), sometimes two IIC modules will be detected, but only, if there are only two modules on the IIC chain. If I have all four installed, only one will get detected (and is usable). If I do quick power offs/ons (less than 10 seconds of waiting), always only one module will be detected. Strange, isn´t it?
Please don´t waste energy on that matter, it is not important, I just thought maybe someone has a clue and says… of course, you have to do this and that and it will work
Just checked the IIC modules, everything seems to be present, capacitors C1-C4 with the right values are there…
As a further test, I just connected the MIDI IIC modules on my spare LPC17 core, which is now running the same version of SEQV4 (compiled with the same toolchain but with LPC17 target). The LPC17 core is USB bus powered and has nothing else (no frontpanel or displays) attached.
Result: exact same behaviour - it sometimes detects both IIC 1 and 2, most often only IIC 1 (checked via terminal). I used the same IIC bus cabling (same wires and connectors).
Then I uploaded a default build of SEQ V4 (and the newest bootloader just to be sure) downloaded directly from ucapps.de - no change…
It is getting late, thanks for your help, while testing I discovered that MIDI port rescan “Newbie Mode” easter egg in MIOS Studio, that made my day, thanks! :).
Will report back with new findings, at the moment i would say something with my IIC cabling seems really messed up
Ha! Somehow I knew that something could be incomplete at my side when you are not able to get these modules running!
I added the jumpers to the out_only schematic. The original “out_only” module was only intended as a workaround for a UART bug in the PIC18F4620, but today the module is mainly used in conjunction with MIOS32, and people prefer to use more than one…
Best Regards, Thorsten.
P.S.: music session: let’s plan this for mid december once I’ve vacation
- will upgrade to the new STM32F4 based core and check IIC MIDI again, when it is available - 192KB of RAM will allow for more SEQ goodness, maybe even 1024 step patterns as a Christmas present? 
