Troubleshooting Midi. Stopped Working. Problem Solved!

[Pablop]

Hello people,

I have what was until yesterday a little fully-working MB64.

Midi has stopped working, and, following the troubleshooting guide, I see I have no 5v at PIC´s Tx and Rx. (TEST CORE4, all other +5v are OK)

I´ve already checked all solderings, voltages in core, etc. and, everything is ok. The test results are making me believe what I suspected, that is maybe I burned the Tx and Rx on the PIC.

What do you think? Is there anything else to do before buying a new PIC?

I forgot to say that all other functions (menus, pots, LCD) are working.

Tanks a lot

Pablo

???

[Pablop]

I´d like to add some more details, maybe they help...

All +5 voltages are ok but those to Tx and Rx.

Ground to Tx: 3V

Ground to Rx: varies a little around 0.5V

Just to make it clear, the application is up and running, showing in the display the menus, pot movements, etc.

Thank you

PD: the led test doesn´t work (no flicker, no light at all)

[TK.]

Hi,

the Rx pin has a pull-up resistor to +5V, accordingly you should read +5V so long no MIDI data is received (you can doublecheck this by temporary removing the optocoupler)

So, it's very likely that there is a problem with the power tracks at the right side of the core PCB

With your multimeter you can follow these tracks to locate a crack, a bad connection or something similar - once the voltage turns from 5V to 0V, you know, where it is

Best Regards, Thorsten.

[Pablop]

Hi Thorsten,

First of all, THANK you for your answer, and what´s more, for this incredible project.

Following your guide I made the 5V appear on Rx. Apparently there was a broken joint on the pull up resistor.

But Tx is still at 3.2V. With my scarce electronics knoledge I understand Tx should deliver the 5v to that "line". There are not too many connections. So I checked whether there were any unwanted contacts ( I used an ohmeter) with all adjacent tracks, and found nothing.

Any feedback will be much appreciated. ( I have to say I´m very impressed with the quick answers in this forum!  ;) )

Pablo

[Pablop]

Hello,

Made another test:

took out the pic and measured voltage from ground to TX pin. It read .05V, what I think is normal when I have an open circuit, and that´s what we want here, isn't it?

This I did because I thought that maybe some of the 5v delivered by the pic in that pin could be going somewhere else, in case of a non-wanted contact between adjacent tracks.

Hope that´s useful.

Thank You

[Pablop]

One more test.

This maybe stupid, I don´t know.

I measured

GND to TX 3V

5v to TX: -0.1

Maybe that´s normal, because the TX part isn´t connected to other parts of the circuit?

Well, maybe someone can lend me a hand before I go purchase another pic ;)

Thanx a lot

Pablo

[Pablop]

Another one:

I bent a little the TX pin so that it stays away from the socket. Point: to isolate the voltage of that pin from the rest of the circuit.

It keeps reading 3v in my digital multimeter (it shows a little variation +-0.2V). What I found SO wierd is that a little analog multimeter reads 0V instead of 3V!

This is puzzling me more and more..!

Appart from the funny thing about the analog multimeter (I´m curious, though, how can this happen), question is:

Should I blame the PIC?

Thank you all

Pablo

[dj3nk]

A ANALOG multimeter has its own resistance what means you have to calculate back.

A digital multimeter has an enormous resistance of some MOhms, so it shows you directly the real value.

greetz

[Pablop]

YES!!

;D

I solved it!

It seems I had burnt the TX of the PIC..  >:(

Apparently, a bad soldering had shortened the TX and RX pins. The PCB was made by the cheap photocopy method, so the tracks are not so neat. That makes soldering a little difficult.

Once I solved that, the RX had the 5v needed, but TX was still at 3V.

I bought a new PIC, and everything is correct now!

By the way,

I first thought that if I copied all of the data in my old PIC to the new one it would run. But it didn´t. I had to make the usual procedure, burn the bootstrap loader with the programmer and the rest via MIDI.

Why´s that?

Thanks

Pablo

A ANALOG multimeter has its own resistance what means you have to calculate back.

dj3nk: I´m sorry if this sounds stupid, but I cannot get how could I calculate it back when the reading is 0..

Maybe is a too cheap and bad meter?

[dj3nk]

Fine! Thats good.

Look at your analogmeter. Anywhere on the case the internal resistance should be printed.

But 0 is too low to read :). Your analogmeter acts like an additional resistor in the circuit. And its really uncomfortable to calculate back. But with R=U*I you can do. Write it down with your readings. Dont forget your analogmeters resistance! Now, to get the REAL U you have take away the "additional resistor" from R ..... now you can calculate how much volts there are. But you see you have to know the other exact values to get usable results. So spending some euros or dollars to a nice digital multimeter should be ok if you dont have one ^^.

greetz