Adau 1452

Hello Sergica,

There are so many ways to do this, I do need some info.

So, based on which input is present currently, you want that input to be routed to the output?

will both SPDIF and I2S be present at the time or only either input is present at the time?

How do you want to compare your two inputs to route one of them to outputs?

Below is a basic technique which detects which signal path is present (has high signal energy) and routes that to the output. 

If both inputs are playing at the same time and at almost same level, then the two inputs will be switching back and forth between them and playing at the outputs, which is undefined. So, that situation needs different methods.

This above circuit is just a basic start; it can be optimized and customized based of applications.

Regards,

Harish

No comments...thank you so much for the instructions.
My goal is to be able to input from both analog (CD player) and Toslink (streamer).
Either one sound source or the other, not both at the same time.
I followed your diagram, and it works perfectly for my purpose.
I'm so grateful to people and spaces like this. Thank you.

Harishgowtham please help with the follow up questions. Thank you

Hello Sergica,

Sorry for the delay in response.

Sergica said:

1. Does muting unused outputs from hardware configuration help at all?

I didn't quite understand this; there are no options to mute the outputs in hardware configuration, it only has options to 'tristate unused output channels' in SPORT section.

Moreover, the unused output channels don't need to be muted in the schematic, because if you didn't include the input or output channel blocks (unused channels) in the schematics, then those channels will not get into the core. which means, if you are using I2S in serial input port 0, then you don't need to worry about the unused 14 channels, it can be disabled (uncheck the box) So, you don't need to worry about the unused channels, it won't create any noise. I hope I understand your question. If not, then feel free to reach out.

Sergica said:

2. Is it possible to implement a block or set to minimize clicks during compilation?

You can include a slew-based mute right before outputs and enable the muting before compiling the project, once the project is compiled and running then you can unmute the outputs. Both sigma studio and system controller can do this job, you can implement via which one is available in your system.

Sergica said:

. I tried powering the board with ADAU from a USB power cable, but it doesn't boot. Does it need prior power negotiation, and is that why I need a 4-wire USB cable? Thanks

I need some info.

The ADAU1452 processor’s worst-case maximum power dissipation is 960 mW, while a standard USB 2.0 port can supply up to 2.5 W (5 V × 500 mA), which is sufficient to power the processor alone.

However, ADAU1452 evaluation boards typically require higher current (around 1.5 A at 5–7 V) because they include additional components such as codecs, op-amps, LEDs, and EEPROMs. The total current demand depends on all these peripherals.

In your case, we’ll need to calculate the total power requirement based on the ICs and components used on your custom PCB. Could you please share your schematic so that we can estimate the power consumption accurately?

Thanks lallison  for the reminder.

Regards,

Harish

Hello...thanks for the information.

1. I understand...it's not necessary to mute it, and in any case, it doesn't add anything, okay.

2. I'm applying multiple slew for my 4 output channels...I leave the volume control at the minimum (down) but I'm still experiencing the same clicks when compiling.

3. My board is the low-cost Chinese one with an ADAU1452 chipset. If I power it with a 1A, 5V smartphone charger, I have no problems.

But I tried powering it with a Meanwell 5V 2A power supply, cutting the USB cable and leaving only the two power wires, and disconnecting the data. In this case, the power supply goes into protection mode and the board doesn't boot. Or if the board does boot, it only works for a few seconds.

Hello Sergica,

For your second item about still getting the clicks when you program it. Are you talking about from SigmaStudio? Because you do not know exactly what the DSP was doing before you wrote the new program. If it is from a powered down state then it could be due to how the codec is programed and how it comes up. It should come up muted. 

As far as the power issue. It sounds like this Meanwell is not doing well at all!! I have actually experienced issues with Meanwell supplies. Look at the output on a scope. Is it the correct voltage? Is it putting out a lot of noise? I do not know what this data wire is you cut? Data for what??? I highly suspect the power supply is not capable of supplying 2A. 

You could try this as an experiment with the power supply. If it indeed can supply 2A at 5V then use ohm's law. 5V/2A = 2.5 ohms. Connect a 2.5 ohm resistor onto the output of the power supply and see if it can actually supply 2 amps? Keep in mind that the resistor needs to be rated for at least 10 Watts!!  I2R= 10W. Do not run it for long but watch the output on a scope and see if it keeps it at 5V and does not produce a lot of noise. I am betting it cannot hold up to that test. Shoot, try 3 ohms. That would be 1.67A. 

Harrish was doing great helping you but I felt I should chime in...

Dave T

Hello.
On my Asian-made board based on the ADAU1452 and CS42448... The two boards, ADAU and codec, are interconnected via a 30-pin header. Pins 2 and 30 are connected to GND. As I understand it, two GND points are not enough to compensate for ground loop issues. Is this correct? Would it be advisable to add more GND points to improve performance? Does the problem seem to stem from not having separate AGND and DGND? My system generates a slight hum and hiss, not audible at the listening position but noticeable a few centimeters from the drivers. It has improved with balanced cables, but I would like to try to improve it further.