I see a spuriouse on the RF output of LTC2000Y-16 that I can't seem to get rid of.
The spuriouse that I'm having trouble with is related to the folded back 2nd harmonic. For example:
I'm generating a 933.888 MHz sine wave, using a sampling frequency of 2490.368 MHz. The 2nd order product will then be located at 1867.776 MHz and this is then folded back to 622.592 MHz. The folded 2nd overtone is attenuated by the transformers so it is lower than 70 dB relative to the sine wave. So far everything is perfectly fine. However, at exactly half the folded 2nd harmonic (622.592 / 2 = 311.296 MHz), I have a spurious that is only 63 dB lower than the sine wave.
63 dB is within my specification (> 60 dB) but I would like to have a little more margin. In the "Single-Tone Spectrum at fOUT = 990MHz, fDAC = 2.5Gsps" figure in the datasheet, this spurious is invisible.
The spurious (half the folded back 2nd harmonic) is always there. I used 933.888 MHz just as an example.
The spurioses exists (and have about the same level) both on my own board and when using Demo Circuit 2303A. I'm using the DAC's internal pattern generator to generate the sine wave. On my own board I use both the DAC's pattern generator and our own FPGA's sine wave generator.
So my questions are:
Clarence.Mayott, can you please take a look at this query?
I have now tried to reproduce my problem but this time without luck.
On my own PCB I still see half the folded 2nd harmonic but I have a slightly different transformer configuration on its output so it is not that strange.
I can also comment that the 311.296 MHz spurious I mentioned in my first post is the 3rd harmonic (which coincides with the half folded 2nd harmonic for that frequency).
However, it would be nice if you could help me explain why the 3rd order folded back harmonic is just 63 dB below the sine wave output (that is generated with the internal pattern generator). Do you get the same result on your evaluation board?
Is the spur exactly 622.592/2 or is it 316.112? It could also be a fs/2-fout spur. Either way at that frequency are likely to be around -60dB. There is a way using the registers to trade off IM3 for simple harmonic distortion. The dynamic linearization register might help to tune your distortion components to get the performance you want:
I have made some changes to the RF output and it looks promising. I'm now waiting for the new prototypes so I can do the final verification.
Great! Let me know if you need any other help.