Post Go back to editing

SigmaStudio communication with EVAL-AD1938AZ

Category: Datasheet/Specs
Product Number: AD1938


My team and I recently purchased an EVAL-AD1938AZ board and we are having trouble connecting it to our DSP.

Our DSP is the MCHStreamer from miniDSP. The goal right now is to have an analog signal as an input to the EVAL-AD1938 via one of the 3.5mm audio jacks and then have the converted digital signal sent to a PC via the MCHStreamer.

Right now, every time we try to write to the EVAL-AD1938AZ using the SigmaStudio software, it seems as if the write was successful, but then when we pull the current settings from the board, it seems to have reverted back to its default state.

Does this behavior sound correct or when we get the current state of the board should we be able to see the same values that we wrote to the board?

As far as we can tell we have the EVAL-AD1938 configured the way we would need it to be in order for the MCHStreamer to be in control of the MCLK and the AD1938 to be the receiver of the MCHStreamer's MCLK, but we aren't getting any audio from the ADC output.

Thanks for any input.

  • Hello thaines3,

    I am not sure if you are actually communicating with the part. It does use SPI so there is no handshaking. 

    Watch the capture window in SigmaStudio. Click on a register to change a setting and you will see the data going out to the eval board. 

    Then click to read all settings and watch the data coming in. There is a good chance that it is all zeros or more likely all "F"s. That means the data is either all zeros or all ones which means you are not really talking to the part. Either the part has no power or the jumpers for the SPI port are not configured correctly etc. 

    Take some pictures of the eval board to show me all your settings and connection to the board. I also gave our central Apps person some more details late yesterday when he contacted me about your case you submitted. 

    I gave him a lot of instructions. 

    It is great that you are using SigmaStudio. 

    Dave T

  • Alright, thanks for getting back to us on this.

    I think we figured out the problem, we needed to connect the common "0V" from the power supply to the green post on the dev board



    We are getting several new LED illuminations, such as Error on D9, but we're pretty sure that this is just because the CPLD is bypassed.

    The picture below shows these LED's

  • Hello thaines3,

    Thanks for reporting back to us on what you found. Yes! 100% you need to make this connection otherwise the center ground point is floating and there are many current paths that only use the positive or negative supply so there is no return current path back to the "center" voltage of 0V. 

    I would also add that you need to connect this center ground point to the chassis ground of the power supply. The board will work without it however, you will end up connecting this to other devices like a computer or an audio source/destination that is referenced to chassis ground. Therefore, you will end up using that external device as the connection and main current path to ground for any noise and that might cause all sorts of unpredictable results and usually it is digital noise or garbage in the noise floor. It also can lead to damage if you happen to have static electricity on you there is no place for it to dissipate without going through the power supplies or the external devices. 

    Here is a picture of my bench power supply and you will see the green jumper to connect the chassis ground to the center ground point. 

    This was a good find. Things should work better now. By the way, the "Error" LED has to do with the SPDIF receiver. With no valid signal being sent to it this LED will turn on. You need not be concerned unless you are using the SPDIF of course! 


    Dave T