Noise production when DSP is active on dev. Board

Hello JDWBE,

What is connected to the analog input of the eval board?

Have you used my started project where there is an internal oscillator in the DSP program that you can use as a signal source? Do you hear it with that program?

How are you listening? How is the amp powered? Try to plug in headphones directly to the eval board and see if you hear the noise?

There are many variables here. There can be places for ground loops and digital noise to get in. Often I have found that noise from a computer will be similar to what you describe. I often use my MAC as a sound source and sometimes I get a ground loop and the result is not always a hum but can be a buzz from the digital signals in the computer.

So eliminate the amplifier by using headphones and see what you find.

Thanks,

Dave T

Hello

Many thanks for you quick response.

My source is a HP notebook, powered from the battery. The only difference with the 2 boards: ADUA1701MINZ => powered from the USB notebook, ADAU1467Z powered from the wall adaptor delivered by the DEV board.

I start from the project inside Sigmastudio+, can you give the link to your project please? (The we can talk about the same things. )

When I dive in the settings, I got an orange button “PLL unlocked” standard stock values used here, this is maybe also not good? (I’m not yet so familiar with AD DSP.)

Hello JDWBE,

Please make sure that you answered all questions from Dave. it helps to resolve the issue faster.

Please find the Basic pass thru project for SS+ below. There are two inputs in this project, one is from the ADC and another one is internally generated sine tone. You can switch between these inputs and see how it behaves. There are also level meters, so you can monitor the input and output levels.

As Dave asked, please connect a headphone in the output jack of the Eval board and see if you are still getting the same noise.

For this project, please connect the input to 'IN1' (ch 32-33) and output to 'OUT1' (ch 0-1) jack.

Basic_passthru.ssprj

JDWBE said:
When I dive in the settings, I got an orange button “PLL unlocked” standard stock values used here, this is maybe also not good? (I’m not yet so familiar with AD DSP.)

I don't think it's a problem; you can verify this by reading back the PLL_LOCK register. You can open a Read/Write window by right clicking the DSP IC >> Open settings>> ADAU1467_0 >> Read/write.

One thing to note here is that this window accepts the register value in decimal, so you should enter 61,444 for PLL_LOCK register which is 0xF000 in hex. You can see that it returns 0x00,0x01 which is PLL_LOCKED condition. Another thing is once you read back the PLL_LOCK register, the orange button becomes green, and it shows PLL_LOCKED. Have a look below and try reading to confirm PLL is locked.

Regards,

Harish

Hello

Thanks for the response.

I tried with headset and I hear the noise in the headset.

I tested also without any connection at the input, only dev board and AMP connect. The AMP and Devboard are powerd by the same PSU (R&S LAB power Supply) When the DSP is running / starting I get the noise. As I already described when I use the ADUA1701MINZ everything is clear without any noise.

When I generate the sine wave (from your sample software.) the same noise is produced.

What I also tried, adding a 100nF cap to the output filter of the DAC output to reduce the bandwidth of the schematic, I hear a lot les noise!

The PLL option is green after manual read of the registers. So that’s not to problem.

Kind Regards

Hello JDWBE,

I would like for you to isolate this down even further.

Do not use the lab supply. Use the power supply that came with the eval board.

DO NOT connect your AMP Dev board. ONLY connect a passive headphone to the Output 1 on the eval board.

Connect the USBi to the eval board.

Then load in this program I will attach. It has an oscillator that will feed every output but I put in a volume control on the output of the oscillator not the output of the eval board. So when you turn down the oscillator the main signal flow will still be at full volume. So turn down the AUXADC0 pot to fully CCW (Anti-clockwise as they say in the UK)

Link compile download.

Now you should be able to turn up the volume pot and hear the 500Hz oscillator. Be careful because you can go to full volume with this.

Turn down the oscillator all the way off and see if you hear any noise.

Also, I would disconnect the USBi. Once you downloaded the program it will continue to run. So disconnecting the USBi will remove any ground loops/noise that might be introduced by your computer.

On my system I am using the Wall-wart power supply for the eval board and even with the USBi connected I here ZERO noise out of the eval board.

Note that plugging and unplugging the USBi will sometimes cause a glitch on the reset line on the board and resets it. I had it do it once while testing but it did not do it every time. I suppose you could program the EEPROM and have it selfboot. But, you should not have to go to that extreme.

I have seen switching lab power supplies cause issues with sensitive audio equipment but I am talking about class-A amplifiers with lots of gain. These eval boards should not be that sensitive.

Now add to it a computer and then the other amp eval board and there is lots of chances for ground loops to happen. This test should eliminate some things...

Dave T

ADAU1467Z IO routing and settings for Rev B eval board-Osc Vol.dspproj

This is a screenshot of the entire program