Post Go back to editing

ADAU1781 Clock Output

Hi I'm trying to use the Master Clock output on the ADAU1781 eval board. I have all things set up as they are described in the startup guide, and my MCLK source switch is set to OSC. I was checking the MCKO pin with an oscilloscope and nothing was appearing, until I zoomed in on the y axis to a 20mv scale. Clock output appears but is under 20mv. Is it normal for the clock's amplitude to be so small?

  • Hello,

    That is definitely not a valid clock signal; you should be seeing something closer to 3.3V. Could you attach your SigmaStudio project so I can verify that there's nothing wrong with your code?



  • Hello,

    Which MCLKO pin are you looking at? J21, J14 or the MCLKO of the part?

    As you see you must set J14 to the "B" position to get it over to the header or you can tap off of pin 3 of J14.

    Here is a screenshot of the schematic:

    Dave T

  • I am trying to use the ADAU1781 as an I2S codec only, so I do not currently have any sigma processing uploaded. Is it necessary to upload a sigma dsp program to configure I2S?

    I am trying to get the clock out from the MCKL pin on the J4 header. The MCKL Source switch is set to OSC. On MCKL pin 1 I see a signal with a frequency around 12MHz an amplitude a little under 20mV.

    Also noticing some strange stuff on the other pins. On LRCLK pin 1 I get a signal that looks like a distorted lower amplitude version of the signal on MCKL pin 1. On MCKO pin 1 I get 2 volts DC.

    What does it mean to put J14 into the "B" position? Is that relevant to the J4 header or is it something separate?

  • Hello,

    Verify that R12 is stuffed. This resistor is a 49.9 ohm resistor in series with the MCLKO pin on its way to header J4. I am pretty sure it is stuffed when built but this is the only thing that would keep the signal off of the header. If this is OK then the issue is with setting up the part in software.

    You should use SigmaStudio to program the part. It makes it really simple to setup the registers. You will have to setup a simple pass through program. Take the inputs and feed it to the outputs, then SigmaStudio will allow you to compile and load the program. Then you can control the registers in real time and look at the settings etc. When all is setup properly then you can capture the register settings for use by your system controller.

    The MCLKO pin will not be working if the PLL is not setup properly etc.

    Another benefit of using SigmaStudio is that you can send the file and I can try it here.

    The "B" position of J14 is what is used to send MCLK to the digital mic header. If you are not using a digital mic then it is not important.

    You mentioned MCLK pin 1. Which pin? Is that the MCLK Source switch? Then pin 1 is the external MCLK.

    What you are probably seeing when you see this very low 12 MHZ signal is just in the noise floor and it may depend on where you ground your scope probe but often it does not matter, you see it all over the board. This board was designed a while back and we have gotten better at layouts but in the case of 12MHZ it easily bleeds to other places. So this is not a signal, just noise is my guess.

    Sorry, for the delay in my response.

    Dave T

  • Now that I uploaded the sigma studio program I am seeing some solid clock signals. I now get 48kHz on LRCLK 3MHz on BCLK and 1.5 MHz on MCLK (in PLL mode when I select MCLK Out Enable). R12 is stuffed.

    For the codec to respond to I2S is there something more I need to configure within Sigma Studio? I cannot find a mention of it anywhere in the register control window, or in the schematic window. It seems odd that I would leave input connected directly to output when they are actually connected to an external device.

    Also by pin 1 I meant that on the J4 header there are 2 columns, labeled 1 and 2 at the top. By MCLK pin 1 I meant the pin in the row labled MCLK and the column labled 1. All the outputs I am seeing are in column 1.

    Thanks for your help!

  • I experimented on my own a bit and now I am seeing a signal that looks like output bits on the ADC pin when I route one of the inputs to digital output zero in sigma studio. Is this the correct way of doing things?

    Also, how would I then route input from the DAC pin on the board?

  • Hello Emmett.

    Sorry for the delay. I have been somewhat overwhelmed and trying to catch up.

    Can you send over your SigmaStudio project? It is difficult to see what you are doing. Plus there are register settings that can change the digital signal flow.

    So you can take the ADC output pin and connect it to the DAC input pin which will route the ADC back into the serial port. Then based on what the settings are, you will either directly go to the DACs or back into the core where you can then send it to the DACs.

    Dave T