The measured SFDR is 48dBc as compared to 84dBc expected from the datasheet due to harmonic spurs. The input sine wave frequency of 16MHz and the clock frequency of 80MHz were applied to ADS5562EVM board with added LVDStoCMOS converter on U15. The measured SFDR plot is attached.
Hi Vijay,
1. Please find the fft plot below with 0V differential input signal.
2. The input to differential signal is a 3.4Vpp sine wave at 16MHz. Please find the update plot below. As shown, a lot of spurs are present in the Nyquist band. These spurs need to be removed to match the datasheet plot.
3. SN74AVC16827 is added to U15 on the eval board which converts LVDS output to CMOS voltage. Please refer to the latest schematic of eval board for the details.
Regards,
Albert
Hi Albert,
I was able to replicate the datasheet level performance on my bench. See below plots:
1) 80 MHz clock and 15.5 MHz input
2) 80 Mhz clock and 16 MHz input.
I don't have a 16 MHz bandpass filter. Nearest center frequency bandpass filter I have is 15.5 MHz. So I used it for both the above plots. I used 2 SMA100A signal generators for clock and signal inputs.
I suspect that performance on your setup might be limited by quality of clock or input signals. I suggest you to try with bandpass filters.
Regards,
Vijay
Hi Albert,
This test was done with eval board on test bench. I used the TSW1400EVM for data capture.
Two SMA100A signal generators were used for clock and input tone generation. I used a TTE bandpass filter (part number: kc4t-15.5m-775k-50) for input tone. read more about filter specs at: 
The purpose of the filter is to suppress the harmonics from signal generator. I haven't used a filter on the clock input. As you see degradation of SNR with increasing input amplitude, cause for this can be noise on clock input.
I used the eval board in default settings. I have not added the LVDS to CMOS converter (U15). But as this in digital domain, it should not affect the performance. Input amplitude for 15.5 MHz frequency was about 15dBm. For the 16MHz case, the input amplitude from source was much higher as filter center frequency was 15.5 MHz.
Regards,
Vijay
