ΣSigmaStudio Ask AI Русский Log in

ADAU1452 Custom Board Bringup Issue

SigmaDSP & SigmaStudio Sylwester Apr 17, 2024 00:00 14 replies Original on forum ↗
Thread Summary

Thread Summary

The user is experiencing issues with a custom ADAU1452 breakout board where all registers read 0xFF and remote control via USBi is impossible. The problem was initially due to the reset pin not being pulled up. After adding a pull-up resistor, the registers now read 0x7F and 0xFF, but there is still no MCLK signal on the CLKOUT pin. The final answer suggests checking the MCLKIN pin for a clock signal and provides a project file for further analysis.
AI Generated Content
Subscribe
Sylwester Question
Apr 17, 2024 00:00

I've been working on a custom ADAU1452 for an upcoming product prototype and the schematic for my breakout board seems correct, but all registers seem to read 0xFF / remote control via USBi is impossible.

Not visible on this schematic is an SPI EEPROM (which is on a separate "stick" PCB).

Edit Notes
Added board layout

[edited by: Sylwester at 12:44 PM (GMT -4) on 17 Apr 2024]

Audio Signal Processors

adau1452

hardware

custom_board

bringup

SigmaDSP Audio Processors

Show More

Screenshot-2024-04-17-at-14.35.24.png
Screenshot-2024-04-17-at-14.35.24.png
Screenshot-2024-04-17-at-14.43.59.png
Screenshot-2024-04-17-at-14.43.59.png
Harishgowtham Reply
Apr 17, 2024 00:00

Hello Sylwester,

which protocol you are using, I2C or SPI?

Note that you have to connect to slave control port of the DSP to program the DSP. Please refer the pins below.

If it's I2C, then need to take care of I2C addressing. The state ADDR0 and ADDR1 decides the address. Let me know how you have connected it.

If it's SPI, then you can put a scope on the SPI lines and see what it sends out.

Since you are using our ADI's USBi module, it would eliminate a few perspectives. (Third party USB to I2C/SPI module may have difficulties in communicating)

Please verify the above points, I didn't have a clear look at your schematic yet, I will do it and comment on it soon.

Regards,

Harish

pastedimage1713362368076v1.png
pastedimage1713362368076v1.png
DaveThib Reply
Apr 17, 2024 00:00

Hello Sylwester,

I am going to add a few more questions to the ones Harish asked.

When this usually happens it means you are not communicating with the part and I know you understand that.. So the question is why? There can be many reasons and I will list them here as they come to mind. This is exactly what I do when faced with something like this.

This first one has bitten me enough times,... Does the board and the part have power? (I know, I would never be that dumb and do that.... LOL)

Do all the power supplies have voltage? IOVDD, AVDD, DVDD, PVDD?

Another good one is this: Is the USBi actually connected! Another LOL. but in your case you will have some sort of custom cable/connector arrangement so check that wiring carefully.

Is there a master clock coming in?

What is happening with the Reset pin?

These all can cause an issue with the part not responding and the data line sitting high and so you read all 0xFF's.

Do you see any clock out of the CLKOUT pin?

Let us know what you find.

Dave T

Sylwester Reply
Apr 17, 2024 00:00

IOVDD = 3.3v
AVDD = 3.3v

DVDD = 1.2V

PVDD = 3.3V

USBi is double checked and connected via short jumper wires.

Communication is SPI.

DaveThib Reply
Apr 17, 2024 00:00

Thanks for the info...

So do you have master clock?

what is going on with the reset pin? It goes off of the schematic so I cannot see anything about it. It should go high shortly after the board powers up. Do you have a reset generator? It is good to hold it in reset for a few hundred milliseconds.

Dave T

Sylwester Reply
Apr 17, 2024 00:00

The reset pin is on a pin header, which is connected to the reset pin of the USBi. There is no clock coming out of CLKOUT.

DaveThib Reply
Apr 17, 2024 00:00

Hello Sylwester,

That pin has an internal very weak pull-down so you need to add a pull-up resistor, 10K-ish, to IOVDD and it is a good idea to also have a 0.1uf or an 0.01uf cap on the pin.

If you put a scope or voltmeter on the pin then I think it is sitting low. Low is reset.

Dave T

Sylwester Reply
Apr 18, 2024 00:00

Hello Dave,

Thank you for the hint. After adding a Pull-Up on the Reset Line and pressing "Read All Registers" in SigmaStudio,
I am now reading 0x7F and 0xFF on most registers - is that correct? Looking at my Logic Analyzer, there is still no MCLK.

Sylwester Reply
Apr 18, 2024 00:00

In a 2nd read, the registers are 0xFF again.

1184.SigmaStudio_A.png
1184.SigmaStudio_A.png
DaveThib Reply
Apr 18, 2024 00:00

Hello Sylwester,

Sorry this is dragging on. If I were standing there it would take five minutes...

Is there a clock on the MCLKIN pin?

When you send you are still seeing no MCLK you are most likely referring to the CLKOUT pin that I had asked about correct?

Can you send over your project file? I have a hunch about something...

Thanks,

Dave T

Sylwester Reply
Apr 18, 2024 00:00

No worries. There is a 24.576mhz crystal on the mclkin pin. I did try with external MCLK from an oscillator before, with same results.
The CLKOUT pin has constant 0V.

My Project File is just a project with the default settings outputting a sine wave on out0.

DaveThib Reply
Apr 18, 2024 00:00

Please attach your project file.

Save me from typing for a long time to explain something that might or might not be happening.

Dave T

Sylwester Reply
May 5, 2024 00:00

Prj.dspproj

DaveThib Reply
Jan 20, 2026 00:00

Hello Sylwester,

Thanks for the project. It helps a lot. Your PLL setting could not be more wrong!!! This is why you are having issues.

This is a screenshot of your settings.

The main PLL is not correctly set for a 24.576MHz input. But, the clock generators could not be more wrong!! You are multiplying the main PLL output by 1 !!

This is all covered in the datasheet.

Here is the part of Table 28 in the Rev-D datasheet. For the main PLL you need to set the PLL Feedback Divider to 96

Since you have a 24.576MHz crystal and you desire a CLKOUT signal of 24.576MHZ as well then these will be your settings:

Then you also stated that you are hearing nothing out. Besides the sampling rate in the core being WAY off due to the CLK GEN 1 & 2 setting, your project is blank!!! Nothing is in the project so you have no DSP program running at all. You will never get audio!!

You have to pick up the audio from an input port, process it then send it to an output port.

Have you found my YouTube channel?

Here is one of my basic starter projects. You will have to change the PLL Input Clock Divider to divide by 8 and then change the MCLK OUT setting to get the CLKOUT to be 24.576MHz rather than 12.288MHz.

I have no idea what you are connecting to what port or the number of channels, the serial format or the sampling rate. So there is still a lot to setup. This project is written to work with our standard evaluation board.

ADAU1452 Basic Stereo Input-Vol-Meter-Mute Osc.dspproj

Dave T

pastedimage1768926549237v1.png
pastedimage1768926549237v1.png
pastedimage1768926752545v2.png
pastedimage1768926752545v2.png
pastedimage1768926969281v3.png
pastedimage1768926969281v3.png
Japon18 Reply
Jan 19, 2026 00:00

Hello gerber pcb zip?