Think something must have been fried?

[rarara]

OK, as part of my ongoing quest to get a squelchy box I was testing the voltages to look for anything unusual. I also happened to check what J5 was looking like (5,0,0 ) and it was while measuring the J5 pins on the I/O board that I think one of the following happened:

either

1) the probe while touching the 5v pin made contact with another component (the only exposed thing close enough was one of the 10k resistors next to it, but have taken that out and measured and its fine)

or

2) the I/O board itself was resting on top of the main board - there are a ton of wires (I have the Bcbox I/O mod too) so this is usually no problem as it is cushioned on a veritable birdsnest. Maybe though some components touched and a short occured.

Result: I could very faintly smell something (hot plastic/burn?) but maybe was being paranoid. However when I put it back together and tried to play I get this:

a) doesn't sound right - notes flat/dull for want of a better word, a bit static/crackly
b) patterns don't play as they should
c) in keyboard mode only really C and D work, the rest don't light or sound (and I only get one note - need to leave KB mode and go back to hear another note). Similarly in pattern mode the accent/rest/slide/done are not working (play, next, up/down octave seem ok) - this probably explains why patterns are not working correctly - it seems like most notes, accents are not 'available'
d) accent know doesn't work (not surprising). decay, cutoff, res etc seem ok (considering the sound is poor)

Based on that where should I start looking???

IC12 and BA6110 voltages seem the same as they were, and the 5.33 and 11.98 are still fine. Something I did notice however when I had a pattern playing and I was testing some voltages to try and diagnose, IC15 and quite a lot of the other points on the PCB when touched by my meter caused a noise/crackle, and similarly testing the voltage of IC12 pin 8 made whatever note was playing drop by half an octave or so - strange. I can't remember if I previously tested voltages while listening to a pattern in my headphones so maybe this has always happened - pretty sure not though.

Please help - I feel like chucking the thing in the bin

[aminoacid]

if you senced a smell, there could have been a short circuit and a component whent burning.

look for burned components and electrolyts with expanded top.

how is your 5v? check both dc and ac on that.


if this is not just a smaller component failing, i hope youve got thoose ic´s socketed.

or perhas send it to some one for repair.

[rarara]

5V is still fine - only checked DC though. should AC be double or thereabouts? Visually checked all components and none look unusual.

Only a couple of ICs are in sockets the ATmega162 and IC15 I think. it seems odd that only a couple of the keys 'work'. Is there any listing of what the various ICs control?? or the voltages, similar to those available for the analogue components?

[aminoacid]

dont know is theres such list anywhere..

ic16 ic17 ic18 scans the buttons and sends the informations in embedded code to the micro. each chip scans 7 or 8 buttons.

ic9 recieves 6 signals from the micro and through a net of resistors it creates the analogue cv to the vco. its the ad converter.

ic12 is part of both accent/vcf env and vco slidecircuit. in both cases signal is carried out from the micro through some transistors and on to the chips.

could be that either or several of theese ic´s (and/or components around) have failed at the same time. or your micro, which is more the central piece has gone kaputt.

if you had a scope you could prob see what the micro is saying on each line. however buying a new micro and other ics are not so expensive and can be of good to have as spares when living dangerously.

[rarara]

by micro do you mean IC3? and needs to be pre-programmed? if so, where is the cheapest source??

[mmm]

Before you jump the gun and get a new processor:

This is very typical of a fault in IC 16 to 18:

c) in keyboard mode only really C and D work, the rest don't light or sound (and I only get one note - need to leave KB mode and go back to hear another note).

I've had this "stuck note" problem on a number of xoxboxes before. Unfortunately I'm not at home right now so any detailed description how to test will have to wait until I get back later today.

I believe the problems described as far as buttons and LEDs go can not come from a faulty processor.

[rarara]

interesting....ok, any info you could provide would be appreciated

[aminoacid]

they are called 74hc165.

try your local electronic shop, mouser, digikey etc.

[mmm]

IC16 to 18 serve the following switches:

IC16: Done, Slide, Accent, Rest, G, A, B, high C
IC17: Up, Down, A#, G#, F#, D#, C#, Next
IC18: the remaining stuff

You could also describe it like this:

When looking at the front panel: IC 17 serves the upper row between "Next" and "Up", IC16 serves the right "corner" (lower row from G upward and "Done"), IC18 serves the left "corner" (lower row from F downward and "Prev").

So this:

Similarly in pattern mode the accent/rest/slide/done are not working (play, next, up/down octave seem ok)

clearly points at a problem with at least IC16. This can affect all other ICs as well so first thing would be to swap out IC16 and see if things improve.

My journal indicates that I had to do this with several xoxboxes in the past.

[rarara]

excellent - ordered so hopefully will have them tomorrow. maybe socket them this time too

do you have any descriptions for other IC functionality? seems a really useful fault finding resource

[mmm]

do you have any descriptions for other IC functionality? seems a really useful fault finding resource

Maybe I'll write up something. The whole digital side is pretty straightforward.

[rarara]

thanks MMM - replacing ic16 worked. gets me back to where it all started though, little squelch and harsh saw. are there any ICs in the digital section which would influence these factors?

[mmm]

Before we continue: Is Accent working now again (not just the button - I'm talking functionality)?

[rarara]

yes, seems to work. apart from obviously raising the volume a bit, the decay is shortened for accented notes as expected.

[mmm]

So essentially the whole digital section works now? Pattern playing as they should?

I don't see how the chips of the digital section could affect the sound quality. This must be a completely different problem.