Midi IN still not working

nunja

Hello everyone !

First of all, congrats to Julian for this wonder. I really was impressed by the sonic possibilities of this beast.

I write here because my LXR midi in is still not working and I am looking for a solution.

I carefully tried solutions exposed in this thread: http://forum.sonic-potions.com/discussion/104/inpossible-to-sync-the-lxr-with-my-daw-with-midi-inout-and-a-led-problem-/p1 but nothing worked for me so far.

I tried 3 different firmwares:

• latest 0.34
• marimba 0.33
• old 0.23

with RXF midi filtering turned to ALL and BPM to 0

In all cases, no MIDI Sync IN and no midi notes IN (I checked the channels on global and tracks). Nothing.

USB is perfectly fine thus. Sync is OK, I didnt test the notes IN.

I had no mean to debug it and I guess it is an hardware issue. I have some programming skills but it won’t help there so I finally managed to remove the LCD screen to see what is going underneath.

Now the my LXR is running headless so I guess I didn't screwed things when I removed the LCD.

I am looking to know how to test the midi in connection with a multimeter to see if its good on this side but I am almost noob in this domain, help would be really appreciated.

I tested the optocoupler according to this tutorial: http://www.antimath.info/electro/how-to-test-an-optocoupler/, it reacts like expected so I dont really know where the problem comes from and I am a bit lost.

Any clue on what I should check / test / replace here to get my MIDI IN working ?

I would love to have my LXR integrates my setup and start to develop on it a little bit



Nunj

julian

did it ever work or do you have these problems since you build it?
on page 3 of the schematic you can see the used parts for the midi in:
http://www.sonic-potions.de/sites/default/files/documents/FrontpanelSchematic.pdf

did you check all the solder joints on these parts?

nunja

Thx for the reply ! Actually I exchanged it with another peace of synth, I didn't built it myself but the guy told me that it never worked. Solder joints seems to be okay but i need to check again.

nunja

Just checked the solder joints, seems to be okay to me. I joined photos.

There is a way to test if a component is broken ?

borfo

Triggering from external midi has been messed up for me for quite a while - worked fine in previous versions of the firmware...  It started messing up when the RudeOG firmware's multichannel/multi-note midi feature was incorporated.  My LXR is pretty much unusable when trying to trigger from an external midi sequencer.  Also, recording from an external keyboard doesn't work for me at all now.


Would it be possible to add the original external midi triggering implementation back into the firmware?  (where each drum sound is triggered by one particular note, all on a single midi channel)  ...single notes on a single midi channel is a lot easier to use with external sequencers.  Maybe this could be in addition to the RudeOG multichannel implementation (maybe the single channel implementation could be implemented on channel 10 or something, and all other channels could be used for RudeOG's implementation?)

brendanclarke

I think if nunja can't even get a sync signal going, there is definitely some hardware issue going on.

That said, I think you're right - the current system of having two different channels (global and voice-specific) that both produce chromatic playing is a bit silly. If you want to try my branch, I've tweaked this a bit (pretty simple to do) so that if the currently selected track is assigned a note other than 'any', and you send a note-on with the global channel, the lxr will try to match the note up with any other voice that isn't assigned 'any' as well.

So if you assign all your voices to different notes, you should be able
to play and record on the global channel as before. Apparently notes
only get recorded when they appear on the global channel - I'll have to
take a look at that some time.

https://github.com/brendanclarke/LXR/tree/instapattern
or just the firmware
http://www.brendanclarke.com/files/LXR/

Note that the MIDI CC assignment is completely rewritten to be by voice channel in this branch (I couldn't get a lot of the cc's to work in the old version anyway). But you can absolutely record with them! Other than that it should work as in .34, the other stuff is added global options, just be sure to back your card up, I'm still testing things

nunja

Just ordered the parts on Mouser, I will change the midi in/outs, diodes and resistors and with chance I hope it might work.

brendanclarke, do you know if there is any way to debug the MIDI IN on the LCD screen in the code ?

Something very simple like displaying a number (eg: midi message number) when midi message is received by the chip ?

brendanclarke

I haven't noticed any midi debug mode, but if you are getting midi data from the USB but not from the physical MIDI jack, it is definitely a hardware problem. There shouldn't be any difference in how the LXR responds to MIDI from the jack and MIDI from USB.

My advice is, if you have access to or can borrow an oscilloscope or data logger and to put it on pin 37 of the mainboard connector (see the pdf Julian attached), you should verify whether you are sending anything to the mainboard at all before taking the front panel apart further.

Sometimes you can kind of detect the voltage change from midi data with a standard multimeter if you send a LOT of data, too Start at the MIDI in jack and work down the chain of components to see where it goes wrong.

Check your resistors carefully, don't just de-solder and swap parts wholesale unless you have a good idea of what's gone wrong.

Good luck!

nunja

Thx ! I will investigate more in depth following your instructions

nunja

Just did the test with an average multimeter. Pin 37 shows a steady 0.8V then ramps from 3V to 0.8 periodically on a very regular basis. Like a pulse. I used the backup sysex data of my a4 to send it to the LXR because its huge (2.2M). I guess its the best I can load hehe. Exactly the same when I test the pin 6 of the optocoupler (at reduced scale, from little bit less than 3V to a steady 0.7).

So I guess the problem is further, but I cannot say much more without looking at the actual signal. 

I will test the software extensively again, but I have no more clue, the solders looks OK on the board.

julian

is pin 6 of the optocoupler conencted to pin 37 of P1 (the mainboard connector, its marked with midi in)
also check the soldering on P1 for bad joints or shorts.

the midi jack is directly connected to the mainboard.

nunja

Just did a continuity check between pin 37 of P1 and pin 6 of optocoupler. everything is allright, solders also. maybe the signal is bad with the optocoupler ? I have a new one I will try to change it.

nunja

diode and resistors are correct also

nunja

When I check the pin 37 with voltmeter, I got a steady 4.70V when no sending midi in. then when I send lot of midi data (lot of noteon and noteoff in all scales), it varies a lot from 0V to 4.70V. So i guess the issue come after, in the main board ?

nunja

Last iteration for today :

= Midi data seems to be received correctly on pin 37 P1

- steady ~5V when no midi in, varying

- varying V from 0 to 5V when incoming loads of midi data

= There is no midi packets incoming in the UART (Uart.c): 

- case if (USART_GetITStatus(USART2, USART_IT_RXNE) != RESET) is never met

Do you think I have to change the main board ?

brendanclarke

Yeah, that's what you should be seeing voltage-wise. Very odd. can you post hi-res images of the mainboard, especially around the edges of the cortex and the header pins? It seems unlikely that it's something wrong with the cortex itself, if all else is well.

nunja

Okay I will take photos with magnifying glass and upload here ! 

brendanclarke

you can also test continuity between the cortex pin and the mainboard socket, that might give a clue!

nunja

How ? I mean which pin on the cortex ? there is a direct midi in pin to the ARM ? it's the world of SMD here it freaks me out hehe.

brendanclarke

Yup, should be direct, to pin 26 of the cortex:
http://www.sonic-potions.com/sites/default/files/documents/MainboardSchematic.pdf

Just put one probe on that pin 37 and test the cortex leads (gently) with the other. One of them should beep, if it doesn't there's your problem.

nunja

Okay I will test it ! thanks a lot for your instructions !

nunja

Mmmm, no pin passed the conductivity test on the arm with the pin 37. Not sure which is the pin 26 but i checked 4 times each pin to make sure. I got a short bip once but never again. I post some photos but I am not sure it won't help. I will have to order a new mainboard Anyway not bad I learned a lot in the meanwhile )) thanx you for your help.