i'm having troubles in using the adau1446 ADC. I've read its datasheet and AN951 about GPIO, but i haven't found anything helpful.

I have ADC input 0 (MP0) pulled down through a 100K resistor, pin MP0 is set for aux adc input, other neighbour pins are set as digital inputs as well.

The DSP is internally working at 192KHz and if i set a readback cell on the ADC output i read 0.9942-something value (?).

I measured the voltage on the ADC input pin and this is about 2.2V with the pin pulled down (??), no other connections are made to


I then tried to change all GPIO pin working modes, the pin is always stuck at some positive voltage (2..3.3V) , except if i set it as an output. In this case the voltage read is zero.

So i set MP0 back as an  analog input and tried to set the other three ADC GPIO pins (MP1 to MP3) as analog inputs: as i was swtiching the input configuration of the other pins, i had the voltage read on pin MP0 falling down to about 0.5V when i set all GPIO pins MP0 to MP3 as ADC inputs.

First question: is this normal ?

With 0.5V on MP0 i had a readback value of about 0.5 (?)

Second question: as the maximum voltage input is 3.3, which i suppose to be the maximum value corresponding to a readout of 1.00, shouldn't it be 0.5/3.3=0.15 ?

I then tried to ground the pin directly, no voltage reading on it, but on the dsp readout i was reading about 0.17..

Third question: why i'm not reading zero?

Note: in all cases, except the first when i was reading 0.999something, the values are quite unstable, no matter on how i set the ADC input filtering.


I carried out these tests on two proprietary PCBs and i got the very same results but i didn't tried on the evaluation board yet.




  • Update:

    tried the same program on the evaluation board and it works wonderfully.

    Now i've to know the reason for that messy behaviour on the final board.

    I can explain part of the unstable readout could come from noise

    on the AVDD line (but it is only 1% which is pretty normal when an smpsu is used),

    but i've to know why i read a positive voltage on the line when configured as an input (?).

    Next update:

    tried again on the production board and found the cause for unstable reading and offset reading when

    input is directly grounded.

    The weird thing is that the MPx pins are at +2.34V when configured as inputs (??).

    Connecting a scope to MP0 input i see this DC voltage with a little 192KHz ripple when

    MP1 to MP11 are set as inputs without debounce.

    Setting MP1 to MP2 as aux adc inputs, i see on MP0 a 48KHz sawtooth: i guess it is caused

    by the aux adc multiplexer which reads this 2.34V value on MP3 (it is set as an input).

    Indeed setting MP3 as an aux adc input, i've no voltage reading on MP0 and almost no ripple.

    I found this problem when a multiplexed ADC was feeded by a high impedance source: the stray

    capacitance on multiplex output or adc input, is the cause for a voltage put on any input to be

    propagated to the other inputs if their driving sources have not a low enough impedance.

    In this case the voltage on MP3 was propagated to MP0.

    But this cannot be the case, since i have a readout of about 0.1 (on the evalboard i correctly get 0.000 when

    MP0 is grounded) when i tie MP0 directly to ground; it cannot be the resistance of the connecting trace,

    as, on the evaluation board, adc inputs are driven by a 100 ohm resistor (the trace resistance is orders of magnitude lower).

    At this point i'm thinking that those 1446 are defective, but it seems strange to me that different components on two different boards

    are defective in the same way.

    What are your thoughts on merit?



  • 0
    •  Analog Employees 
    on Aug 2, 2018 2:45 PM
    This question has been assumed as answered either offline via email or with a multi-part answer. This question has now been closed out. If you have an inquiry related to this topic please post a new question in the applicable product forum.

    Thank you,
    EZ Admin