So i managed to build a functional tester, including the display… until i accidentally shorted -12v and 5v :(((. Most probably atmega644 was dead so i burned the hex and fuses on a new one. Since i was afraid all the ic’s could have been damaged i replaced all of them. Now i got signal on all the outs, can’t read the ins, nothing on the display. there is a faint change in cotrast when i scroll the encoder but nothing readable, also with the contrast trimmer on max. please help and thank you for your time
All ICs have been damaged indeed, including the display controller. You need to replace the display.
Thank you pichenettes, I’ll change it tomorrow
Funny O_O message before death btw.
Changed the display, still nothing The other strange thing is that if i press any of the switches with the corresponding led and after this I turn it off, after i power it up again it’s starting with that last led\\switch used or the last 4,5,6 (each time randomly) and after a second it’s starting on led1 as it should boot. Changed all the active parts less the vregs and the diodes since i get 5v and ±12v. have no idea whats wrong
It has a built-in oscillator IIRC. Does that work?
The 20 Mhz crystal? Also replaced.
Ah, no I mean the audio oscillator for testing
I have good signal an all four outputs, blindly scrolling the encoder through the menu seems to be ok, things like freq. and waveshape, clock/gate/cv, don’t know about the inputs obviously.
Have you tried adjusting the display contrast trim pot? When I first assembled mine it looked like the display didn’t work, but I only needed to adjust the contrast to make the text visible.
done that, it was adjusted for the replaced display, even if i’m adjusting still nothing
Well, then the processor is obviously running just fine. The display worked before and you probably (hopefully!) substituted the defect one for one of the same type. So the only thing I can think of now is a bad solder joint or some IC not sitting properly in its socket.
Does “nothing” mean even the backlight isn’t on, or it’s on, but no characters visible? Have you measured the +5V on the LCD pins, or maybe the backlight resistor is damaged (or incorrect for the new display)…
the backlight is on “there is a faint change in cotrast when i scroll the encoder but nothing readable” only ▒ ▒ ▒ on the first row. I got 5v on pin 2 and 15 on the LCD pins. Moreover if I access any of the switches and then i restart it has the last three led lighting or the last accessed switch and after 30ms it’s on led 1 as it should be.
Okay. ▒ ▒ ▒ on the first row means that the lcd is not initialized.
Let’s see what the LCD needs to be initialized:
1. It needs power. => Check 5V (you did that already, so thats okay) and check GND as well!
2. The power must be stable. => Check with an oscilloscope that the power rails are fine during the bootup-sequence. If you can’t do that, try another power supply. Or try supplying the unit from a bench supply directly. E.g. inject 5V at C26. It shouldn’t be problematic if the -12 and +12V rails are missing when you do this test.
3. It needs a special sequence of bits on its D4-D7, RS and EN pins. => Check continuity between the pin on the lcd module and the corresponding Atmega pins. Measure the pins on the atmega directly, not the socket. Sometimes the chip doesn’t sit in the socket properly (as stated before)
4. The sequence of bits must have the right timing. => Check that the processor is running at the right speed. Check the crystal (what value does it have?). Check the fuse bits (does it use the crystal as a clock source?) Check the audio out of the internal test signal generator (oscillator). Does it have the expected pitch?
If all of these are correct, I don’t see why it should fail.
PS: 30ms sounds like a normal bootup-time to me. the LEDs are driven from a 74HC595 shift register. It may have some random content when you power it up. That content may show up on the LEDs while the processor is booting and vanishes once the processor starts updating the shift register.
Great! i got info on display. It seems i was testing the display with jumper wires and there was a cold contact on pin12, i feel that there is still a latency in response when i press the switches, same business with the LEDs. Could this be because i replaced the crystal? or the 644 clock?
Thank you @TheSlowGrowth your troubleshoot guide is coherent and complete but unfortunately i don’t have access to a scope. Is there any other way i could measure the chip timing?
The only issue apart the booting latency is now that it doesn’t read gate in, also if I trigger/gate the audio envelope with a gate signal it does no audio output. it can however read midi in order to trigg notes and gates
me stupid, replaced 3904 with 3906, wrong drawer, case closed