Troubleshooting microcontroller based audio device

[artran]

Hi,
I have an Arcam rCube iPod speaker that no longer does anything useful. It has an external line-in but I can't get it to switch to it. I have obtained a full schematic from Arcam's website and done some troubleshooting and it looks like all of the peripheral bits like keypad/display, DSP, and amplifier are working as I was expecting but the microcontroller seems to be totally inactive. For example I can see I2C signals from the keypad but I can't see the micro initialise the 74HC595 that is attached to it to extend its IO.

This has made me focus on the micro to see if I can identify any issues with it. So far I've verified its supplies seem OK - it takes a 3v3 supply for IO and generates a 1v8 supply for its own core. When I look at the crystal oscillator I'm not sure if what I see is correct; it appears to be running in that I can see a 12MHz sine wave on both of the crystal's pins but the voltages on the pins seem to be lower than I expected. One side has 0.8V pk-pk and the other side has 1V pk-pk. Does that sound reasonable or not? If it's fine is there anything else that you can suggest I should look at?

The micro is an Atmel AVR (AT32UC3B) in case that's relevant.

[Manul]

Crystal amplitude seems ok. Look MCU reset circuitry, try to analyze, maybe firmware is waiting for some kind of signal to continue booting and never gets it? For example "power good" or similar critical signals.

Probably there should be no I2C signals if MCU is not functioning, because almost guaranteed that keypad is a slave. So it should mean that it is running. Well, as far as my experience goes.

[artran]

Thanks for confirming the crystal seems OK and for the suggestions on where to look next. There are a couple of candidate signals so I'll check them later.

I initially thought that the I2C might be showing me that the micro was alive but the keypad has its own controller so I think it can generate I2C and an interrupt that go to the micro. There is another I2C and two SPI buses but they seem to be constantly idle.

I'll report back on anything i find.

Cheers,