Midi in not working properly on shruthi_xt

Hello again,
some might know me from atmel not programmable :slight_smile:
Now I am back with something else very odd:

Midi in signals do not seem to do anything, now i know this is a bit hard to debug.

I will tell you what i expected: Connect my midikeyboard to the shruthi and simply play some notes, of course this didnt work, usually it has something to do with midichannel setting, so i checked htat, it said the shruthi workes on ch0 for receive all channels so i did put that and nothing happend, so put it to ch1, or tried a lot of stuff but nothing happens.

So i got checked the 137, everythign seems fine, even did unplug it resoldered it, checked the resistors, finally got the oscilloscope out and checked the midi line, you can see some strange things happening there i can tell you that.

But for those who are into measurment, it is kinda hard to probe without an earth, so you always see a giant voltage fl.oating on top of your wanted signal.

But anyhow i think i did see some midi signals on top, now things get weird:

When i tough the optocoupler with the tip of the connected probe ( set on 1:1 this is important i guess for the 50ohm termination) i do send midisignals, i can see the little symbols, they come very random, and i do not even have to actually send something.

So this seems all very strange to me, this screams to be an grounding issue.
I just connected another midikeyboard to it, and no i cannot send anything when touching it with the tip, so maybe my electribe ( which i used as midi keyboard for the shruthi) has some serious midi out issues, which i almost cannot believe cause it works well with everythign else i got.

So i am pretty confused, which is normal when running into problems, i guess i will find a solution in the next days, or someone of you got a solution.

I also read some former threads about this, i already watched all the pins, and i am pretty confident the opto just doesnt transmit any signals, maybe there dont even arrive signals, main reason its hard to test is my current workspace is so little its impossible to press a midikey on the keyboard and watch the output at the same time on the oscilloscope :slight_smile:

So i hope a simple solution comes the way, feel free to share.

Thanks a lot.

update #1:

I did try it on a bare pcb with only midi and 6n137, i did test it first without the atmel chip, i must say the midi signals looked pretty nice, with the atmel socketed again, i strangly only get signals during the negative half wave of the floating signals,now i do not know if that is some kind of measurment error, or if i can take that for granted.

yeah well, just did try to do it at my working shruthi_xt , i can make the sound stop when touching the whol ic, the opto, with my bare finger. That is by the way my most usual testing way :slight_smile: of course only when the voltage is touchable low, 5V well is in my touching range :slight_smile:

And yeah the whole circuit just reacts, so something cannot be right, gotta check the ic again or some solder spots do not know, but i guess its something hardware. but strangly my other board got it also when the atmel chip is attache,d hmm that makes me think.

k ok ok kk

:slight_smile:

half past12 midnight, got it working, as usual, maybe bad chip, i usually do not have problems with esd, but sometimes, amybe this time, no clue, or some pad connections into the socket, but i got some midi sequencer runnign from my korg electribe.

Ok now very nice, now i can start learning about the wavefoprms and watching them on the oscilloscope, sadly I got pure analog here, so no waveform memory yet, i try to do some nice pics and sounds in probably the next weeks.

I like the music :slight_smile: