Aux_ADC and Single Slew Ext Vol introduces noise.
I've been getting noise (hiss) in my project when (and only when) a test signal is applied. I've tracked the problem down to the single slew external volume control, more specifically when it is controled by an aux_ADC.
If it is controlled by a DC source, it is fine. If I change the slew rate from (the preferred) 3 up to 10, the noise is gone, but leaves a laggy feeling control which I'd like to avoid if possible.
I assume the problem is noise? With that in mind, if I low pass filter (2nd order, 10hz) the aux_ADC before it reaches the volume control the noise is gone, however this introduces a stuttering effect if the pot is moved quickly. Raising the filter reduces the stuttering but also reduces its effectiveness at stopping the noise. The ADC filter modes in DSP register modes 2 seem to make no difference at all.
I guess a cap at the ADC input is the ideal solution but that's easier said than done, I'm ideally looking for a software solution that doesn't cause the control to slow down, if possible.
Thanking you,
Alex.
Hello Alex,
If you're using a ADAU1701 or -1702, their Auxiliary ADC filter and hysteresis modes are mislabeled -- see:
ADAU1702 hysteresis settings
The scope waveforms at the bottom of this thread show that these Aux ADC modes effectively handle ordinary amounts of noise, making me wonder why they had no effect for you. It may be good to measure the noise out of the Aux ADC by trying a project such as this one:
The DC Block Filter strips away the DC, leaving the noise. If your hardware allows for "live" operation via USBi, you can measure the noise with the virtual level meter. In any event you can listen to or scope your DAC outputs to find both the level and character of the noise. With the ADAU1701MINIZ hardware, nothing is audible or shows on the meter except for "zipper" noise when the pot is rotated.
To simulate Aux ADC noise I added a white noise source to the Aux ADC output, then tried several software filtering methods -- including a six-pole Butterworth filter and a rather complex slew-rate limiting scheme. None worked truly well . I managed to attenuate the full-strength white noise by about 40 dB, but this translates to 1 part in 100, not good enough to prevent audible level variations. I would rather suggest that you load a test circuit like the one shown above, then experiment to see what in your hardware is causing this noise. Even though you'd rather not modify your hardware with a cap at the Aux ADC input, I'd suggest you scope this input and/or try tacking on a cap temporarily, to see if this is really the matter. There's other possible causes -- power supply ripple, unstable PLL (those components must be right at the -1701), etc. Strange noise problems I had in a prototype were caused by excessive ripple from a 3.3V switching regulator.
Best regards,
Bob
Please log in to reply.