ADAU1761 Analog performance
Thread Summary
The user encountered a spurious 'birdy' signal (~-80dBFS) at 5.5KHz on the ADC line input of the ADAU1761, which varied with time and could be modulated by touching the input pin. The issue was resolved by correctly configuring the test equipment, and isolation between the analog inputs and outputs was confirmed to be over 90dB when using single-ended inputs and outputs.AI Generated Content
Hello,
I am implementing a new design using the ADAU1761 and currently have a couple of performance issues that I need to resolve. Both of the issues below are present on the development board ADAU1X61EBZ and my own hardware.
1. The audio from the ADC (line input) has a low level (~-80dBFS) spurious 'birdy' that varies in frequency with time but usually settles around 5.5KHz. It's frequency can be changed / modulated by touching a scope probe on the line input pin of the 1761. I have been unable to supress this spurious response.
The spurious 'birdy' is only present when the line input is connected to the ADC in the mixer config.
2. What is the isolation between the analog inputs and the analog outputs? I can't see this specified on the datasheet anywhere. To test this I am driving the line input with 1VRMS sinewave at 1KHz, and measuring the output with the input mixer paths disabled. A far as I can tell the isolation is less than 50dB which seems rather poor.
Any help greatly appreciated.
Ian.
Audio Signal Processors
SigmaDSP Audio Processors
adau1761
SigmaDSP
Ian,
I have not been able to re-create either of the issues you describe here. For the case with the spurious response, are you seeing this on the analog or digital outputs? The FFTs of the output signals that I looked at with the input signal disabled (but the input signal paths enabled) shows the output signals at below -100 dBFS for all frequencies.
I also ran your ADC/DAC isolation test and got much better performance than you mention. On the evaluation board, I saw the input-to-output isolation to be >80 dB with a full-scale input on the pins, and the input signal paths disabled.
It would be helpful if you could send any further information you have about these issues to sigmadsp@analog.com. It might be useful for us to see the SigmaStudio project you are using for your tests, as well as any plots of the responses that you are seeing.
Hello Jeradl,
Thanks for your reply. The spurious response is present on both the analog and digital outputs but only when the analog line input path is enabled. I have just re-measured the eval board again and whilst I can still hear the spurious response it's much closer to the noise than I measured earlier.
The isolation issue may be the way that we were making the measurement. As the final application requires single ended in and out, we have our test equipment configured that way. Single ended to single ended I still see around 45dB isolation. However, with single ended in to differential out I see the best part of 80dB.
For the final application neither of these issues are likely to cause problems in the real world.
Ian.
Ian,
My measurements in which I got >80 dB isolation between the inputs and outputs were all done with single-ended inputs and outputs. Specifically which inputs and which outputs are you using to make your measurements? It might be best to double-check that we're both looking at the same signal paths.
Hi JeradL,
This turns out to be a measurement issue to do with the configuration of our test equipment. The BNC connector o/p from our generator was configured as differential +ve on inner, -ve on outer, rather than +ve on inner, ground on outer. Having set this up correctly I see over 90dB of isolation.
Thanks for your help.
Ian.
Ian,
We're glad that this has been worked out. Good luck with the rest of your product development.
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
Please log in to reply.