AD9208 signal is not smooth and has large burrs

This is conversion error. Can you please repeat the test with a slower ADC clock? like 2GHz for example?

Hello, I encountered a new problem when analyzing the sampled spectrum.

The input signal is 100MHz, and the corresponding filter is added to the link.

When the 2.5G clock is used for sampling, the ENOB value of the test result analysis is about 8.8.

When sampling with 1.5G clock, ENOB value is about 9.6.

Is this also caused by  conversion error. In addition, when the amplitude of the input signal is greatly reduced at 2.5G sampling, the ENOB value can also reach 9.6

At present, I am confused whether it is the problem of the ad9208 or the problem of the board design. Another point is that if there is conversion error in the ad9208 performance, why is it not mentioned in the data manual. If I want to replace it with AD9689, does it have the same problem.

Thank you first and look forward to your reply

This is  not board design issue.  if you concerned the conversion error, you can lower down the sampling rate, around 2Gsps, it should be fine.  for the conversion error, @ same sampling rate ,  AD9208 have same conversion error performance with AD9689.  if your are looking for higher sampling rate with better conversion error , you can consider AD9213. 

This is  not board design issue.  if you concerned the conversion error, you can lower down the sampling rate, around 2Gsps, it should be fine.  for the conversion error, @ same sampling rate ,  AD9208 have same conversion error performance with AD9689.  if your are looking for higher sampling rate with better conversion error , you can consider AD9213.