I recently built a circuit using the ADAU1701 in selfboot mode, which is basically the circuit found on page 3 of AN-923, with the WB trigger tied to GND (no WB required) and RESET being connected to a power-on reset supervisor IC. Components in the PLL filter circuit differ slighty from the suggested values, the resistor is a 470 ohms and the suggested 56n capacitor consists of 33n+22n in parallel, because those were the values I had on hand.
I am using both ADC inputs, 2 DAC outputs and a couple of GPIOs to connect incoming I2S data, a D/A selector switch and to control LEDs and a relay. To burn the EEPROM (24AA256, 32kByte) I use the generated E2Prom.hex file and an EEPROM programmer + software (PonyProg). Programming of the EEPROM itself works fine, no issues.
When I first powered the circuit (EEPROM installed), I noticed that the GPIOs were not behaving as they should (the 'high' or 'low' states i found had nothing to do with the program, one of the was putting out 1.2V?!) and that no signal was put through to the DACs. I suspected the oscillator not to work properly (the crystal has a CL of 10pF and 22pF capacitors going to GND), but I have a nice 12.xxxMHz sine on the MCLKI pin.
Next thing I did was to check my schematics and found that the +3V3 connection to IOVDD was missing on my PCB! So after soldering a small wire from +3V3 to IOVDD and its decoupling capacitors I repowered the circuit, but found the same problems as described before. I programmed the EEPROM with a small test program using no GPIOs (all configured as debounced inputs), that should output a 1kHz sine to DAC0. No change. GPIOs still behave as described earlier and no signal on DAC0. Manually resetting the device shows no effect.
It seems that the program is not loaded from the EEPROM, as things turn out the same even with no EEPROM installed.
I read the voltages as follows (referred to GND):
- IOVDD, PVDD, AVDD: +3.26V
- DVDD: +1.75V
- CM: +1.52V
- FILTD: +1.64V
- FILTA: +1.62V
- ADCs and DACs are properly biased to 1.5V
Is it possible that the missing +3V3 connection to IOVDD caused damage to the ADAU1701?
EDIT: It seems that the GPIOs are permanently configured as inputs, since their 15k pull-up resistors allow enough current to flow into my NPN transistor buffers to turn the relay on, for example. I guess that's why I was measuring 1.2V on the GPIO driving the relay.