SigmaStudio ADAU1701 multiple White Noise Sources correlation

Hi Gordon,

You'll need to manually write a unique seed value to each one of the sources if you want them to be uncorrelated.

You can do this using the capture window to figure out the addresses of each seed and then using the Read/Write window to manually write a unique seed value: https://wiki.analog.com/resources/tools-software/sigmastudio/developmentenvironmnent/hardware_windows/registerreadwritewindow

Hi Gordon:

If you use all noise source with the same seed, the signals are going to be correlated. Have you tried writing a different seed to each one of the addresse of the white noises being used. You can do so by using the sequencer. The white noise with seed was coded in such a way for it to be able to repeat it's output. I wonder if it would be better of you, in your schematic to use the simple White Noise that doesn't contain a seed. If I remember correctly, this White noise uses a random number generated from the PC as a kick.

Please let us know your results.

Miguel

Hi Brett and Miguel,

Thanks for your prompt reply.

We do use this technique in our products where the DSP program is loaded from the microprocessor, and then writing seeds works. But I have not been successful when controlling the DSP from SigmaStudio directly.

I did set up 4 separate noise sources with seed.

When the switch on the noise sources is toggled, I see 2 instructions written to the dsp, the first is a seed number, the second is a 1 or 0 depending on the prior state:

Block Write

Time: 17:14:36 - 69ms

Cell Name: W Noise1

Param Name: WhiteNAlgNew21seed

Param Address: 4

Param Value: 0.619525074958801

Bytes: 4

Param Data:

0x00, 0x4F, 0x4C, 0x99

IC: IC 1

Block Write

Time: 17:14:36 - 78ms

Cell Name: W Noise1

Param Name: WhiteNAlgNew21enable

Param Address: 3

Param Value: 0

Bytes: 4

Param Data:

0x00, 0x00, 0x00, 0x00

IC: IC 1

Each toggle shows a new seed number sent and each time it's different.

I just did try writing a new seed to the register address captured in the window by using the register read/write window. This doesn't seem to affect the noise source. I can turn the source on and off and send new seed numbers with the read/write window, or also by copying the output string into the sequence window and sending the commands that way.

Every once in while toggling the noise source on and off affects the correlation between two of the sources, but not all.

I don't see that the noise source responds to the new seed number as expected ?

Regards,

Gordon Cook

Hi,

I just inserted 4 non-seeded white noise generators, and they will come up correlated when the program is compiled and downloaded.

If I then write a new seed number to the address that's one more than the address for the on/off toggle location the noise becomes uncorrelated.

So it appears that writing a new seed to the non-seeded WN generator works to restart the noise generator, but this doesn't work as expected with the seeded WN generator.
Does this make sense?

Regards,

Gordon Cook

This question has been assumed as answered either offline via email or with a multi-part answer. This question has now been closed out. If you have an inquiry related to this topic please post a new question in the applicable product forum.

Thank you,
EZ Admin