TSW40RF82EVM: NCO noise dependence

Hi Marcel,

Although it is shifting in time, you are referring to the large "glitch"? The NCO is being used on both the ADC and DAC? Are you seeing this noise when the output of the DAC is connected to a spectrum analyzer or oscope?

Can you try making the NCO frequency as close to coherent to the sampling/FPGA clock as possible? The ADC has a 16bit NCO and the DAC, 48 bit, so we are limited by the ADC's NCO for frequency resolution (about 45kHz increments).

For example, 65536 samples are captured, and the clock is 2949.12 MHz and NCO is 200 MHz. Try setting the ADC/DAC clock to 2949.12 MHz and the NCO to 200.115 MHz.

Does this serve to change/reduce the noise in the captured waveform?

Best Regards,

Dan

Hi Dan,

In the description of the problem I indicated round values of 200, 250 and 300MHz, but I took into account that not all NCO values are allowed. In fact I used 199.98, 250.02 and 299.97MHz.
The output of the DACs is connected to the inputs of the ADCs.
Analyzing the data for 299.98MHz I find that this noise appears from the frequency of 46MHz, goes through a maximum at 50MHz and then decreases. The frequency of this noise is 48.48MHz

It tells me that this frequency depends on the NCO

Thank you for your help

Best regards

Marcel

Hi Marcel,

What I'm suggesting, in regard to the NCO, has to do with coherency. If you could please set the NCOs (DAC and ADC: example 200.11 MHz as seen in my last post) to a frequency that is coherent with the sampling clock (2949.12 MSPS at 65536 samples). Does this change the nature of the noise?

Does this noise that you described (at 48.8 MHz) appear on an oscilloscope or spectrum analyzer?

If you can share your configuration and pattern file, I can try to reproduce this issue.

Best Regards,

Dan

hello Dan
Sorry for the delay I was not in the office the past 2 weeks.
I will resume testing tonight.. In my past attempts I took into account the allowed frequencies. I have no oscillo allowing me to control the signal emitted.