Hello, I want to use the AD9913 DDS for super nyquist operation. I don't need very high signal levels or particularly great phase noise. Maybe a signal down as low as -40 or -50 dBm would be ok. Given that, is there any problem trying to extract the 58th image? The Fref would be 100 MHz, and assuming no multiplier, I would like to be able to target ~2940 MHz. Thank you.
There is no fundamental problem with extracting a high order image other than the reduced signal level and subsequent reduction in the signal to noise ratio, both of which you seem willing to accept.
Thanks Kenny. I was wondering what the actual power levels might be roughly. If I assume the sinx(x) function, I expect about -40 dB down from max, but I'm not sure how realistic that is to expect from the package. If you have some idea, or can try it out on an eval unit, that would help me make a case. Thanks again.
I don't currently have access to an Eval. Bd., so I am unable to do a quick check. Beside, as you note, the sinc response is only part of the story as there are the package and PCB parasitics to consider, as well. So, even with setting up an Eval. Bd., there would be losses associated with the PCB and external components that were not given design consideration for >2 GHz operation.