Need help troubleshooting

[ignotus]

Hi all,

A few years ago I built myself a Megadrum on experiment board following one of Synthex's old all-in-one designs (attached below). It has been working fine for over three years (except for a small hiccup that was solved with a resistor), but for a few days it has been acting strange and I just can't find what the problem is. The first 8 vu meters are hovering at around maximum (nothing connected), with the first 2 (pedal input and footsplash) randomly going up and down. The other 6 inputs just give out a stream of machine gun triggering. I've taken out the multiplexer and the diode array corresponding to these inputs, changed them for different ones and checked all over for shorts or breaks with a magnifying glass, but still no progress. I am aware that troubleshooting a circuit on perf board is quite tricky, but what I'm asking is if anyone can tell me what part(s) of the circuit I should be focusing on, as I'm sure there is something physically wrong with the board -it all started when light taps on the enclosure would trigger these channels, only now it stays like that- as opposed to something wrong with the Atmega (I reloaded another firmware just in case). If I take out the multiplexer and diode array nothing changes either. I include the pcb layout as a reference.

Thanks in advance.

[dmitri]

According to http://www.megadrum.info/sites/default/ ... _kicad.png:
1. Test for +5v on the pin 30 of Atmega.
2. Test for at least +2v on the pin 32 of Atmega.
3. Check all resistors connected to the pins 28, 29 and 32 are Ok.
4. Check R26-28, R1, R4-6

[ignotus]

Hi Dmitri, thanks for jumping in.

1. 5v ok
2. Around 2.5v
3. and 4. After having a look at the schematic you provided in the link and the design I used, there are discrepancies between the values of some resistors - those going to pins 34 - 40 of the Atmega are 4k7 in the schematic and in my board 34, 36, 38 and 40 are 100 ohms. Could this have anything to do with it? Also, the resistors going to pins 30 and 31 measure 3.15k instead of 4k7.

[dmitri]

After having a look at the schematic you provided in the link and the design I used, there are discrepancies between the values of some resistors - those going to pins 34 - 40 of the Atmega are 4k7 in the schematic and in my board 34, 36, 38 and 40 are 100 ohms. Could this have anything to do with it?

I'm not sure what resistors you are looking at but I referred you to resistors which are in parallel to the Atmega ADC inputs and should be 4k7. If they are 100 Ohms than it would not be the cause of your problem, on the contrary, the signals on the ADC inputs would be very weak.

Also check:
1. For proper connection between pins 1-3 of Atmega and pins 9-11 of MUXers.
2. That pins 6,7 of MUXers are properly grounded.

[ignotus]

Apologies; I made a mistake when stating which resistors went where... the thing is, after a closer look, pins 34 to 36 are not even connected anywhere on my board, while in your schematic they are, and all in all, I can see differences all over the place regarding resistor values, resistors themselves (which I haven't even got)... looks like my board is completely outdated. Could newer firmwares (well mine isn't that new - 20100516 - but then the board was made in 2007) have problems with old schematics?

[ignotus]

1. Connection between pins 1-3 of Atmega and pins 9-11 of MUXers is ok

2. Ok too

[dmitri]

1. Connection between pins 1-3 of Atmega and pins 9-11 of MUXers is ok

2. Ok too

Then it is either a missed break in one of the areas I suggested to look at or a faulty Atmega.

[ignotus]

Once again, thank you so much Dmitri for your help. Out of sheer desperation I resoldered the 4k7 resistors going to pins 30 and 31 as they weren't giving the right values and...voilá! They appeared to be properly soldered but there must have been a microscopic crack there for ages because now the module's performance is much, MUCH better! I thought it was my diy snare's fault and/or bcause of a poor module assembly - I never managed to get proper rolls and had redesigned the snare a number of times, tried loads of different materials for cones - and now it suddenly works beautifully!