PCM1804 High THD+N%

Hello ASB,

I can take a look at your schematic for you. Typical causes for increased THD are poor capcitors (Y5V is generally not recommended) or the voltage rating not being high enough of the capacitors. We usually recommend twice the voltage you expect the capcitor to experience. Jitter in crystals will also cause THD. Only something that adds nonlinearities to the system can add THD.

Also how are you measuring the THD+N? are you isolating the PCM1804 and using an AP?

Do you get a better result if you use the EVM?

I'll get back to you ASAP,

Nate

I am measuring THD+N% on post DAC stage.  I have isolated the THD+N% to PCM1804 by eliminating any other possibilities in signal chain.  No I have not tried this on an EVM.  I have PCM1808 for other sources on the board and for playback I have several PCM1781 they all share the same master clock, the THD+N% for all those devices are below .1%   

ASB,

I took at look at your schematic and it all looks great. The only thing I noticed is that you don't have any AC coupling cap between your inputs to help with common mode rejection. I wouldn't imagine this would remove that much THD though.

Could you tell me a little more about your application?

Also if you provide I2C scripts I can try to recreate this on the EVM to so if it's a software or hardware issue.

Thanks for the post!

Nate

Hello Nate,

This Chip can only be configured through hardware pins.  Please look at the schematic for my settings.  I will add the caps for common mode rejection.

Hello ASB,

Sorry for the confusion, I obviously got a little mixed up here. All of your hardware configurations look good. One time I saw a mismatch in types of commucation used (one was i2s and the other was left justified) but since you have it set to I2S on your board, and the other PCM parts on your board are working I doubt this is it.

How does everything sound? Is there any distinguishing artifact that you can hear? sometimes this helps point to something.

Please let me know if the cap doesn't help and I can continue to look into this,

Nate

Hello Nate, 

I can hear a low level of Hiss in the background.  

Hello ASB,

Theres a few additional things I would like to look into to further isolate the issue:

1) Can you provide a spectrum analyzer FFT of the input pin into the ADC? The THS op amp could have some ringing at high frequencies that would cause extra noise.

2) could you also get an FFT of the power and ground pins to see if there is any noise present.

3) an FFT of the hiss would be useful as well. Some hiss can be a specific frequency which gives us a lot of information, as opposed to an increased noise floor.

4) also it would be worth trying to cut Vcom going to the preamp to ensure there isn't any loading issues with that line. Then just test the ADC to see if anything changes.

5) It would also be worth testing on an EVM to see if you get better performance and comparing differences between the boards.

These tests would be very helpful to further narrow this down, so the results would be greatly appreciated,

Nate

Hello Nate,

I will look the data that you suggested.  

I need some clarification of THS4521 and PCM1804 signal chain that I am using.  From my schematic you may have noticed that I am using single supply to power THS4521 and using Vcom (+2.5V) as Bias for THS4521.  Do I need to remove DC-offset before I feed this biased output to PCM1804, do I need to ac-couple my out put from THS4521. ?

Hello ASB,

You do not need to remove DC-offset as the PCM1804 is DC biased to +2.5 V. Additionally, sometimes a capacitor is used to ac couple the output, however, this wouldn't cause the noise you are seeing.

Please let me know if you have any other questions,

Nate

Hello  Nate,

By adding  a .01uf cap across VIN+ and VIN- of PCM1804 I was able to improve the THD+N% performance of my system.