I am currently using the AD9957 on a custom board and I am seeing a similar issue to a discussion on the TxEnable/PDCLK metastability issue (causing occasional spectral inversion). Is there any fix to this problem other than synchronizing PDCLK to SYSCLK/4N with a flip flop as shown in the other post on this problem? Also, I have tried using the auto power down mode of the 9957 to avoid this issue but it seems that in auto power down mode PDCLK is output regardless of the value of TxEnable. Is there a specific setup for using this auto power down mode and even if it does work, would this solve the synchronicity problem? I would also suggest that this issue be added to the 9957 data sheet (it is currently outline in the 9857 datasheet and it seems as those both exhibit this behavior.) Thank you to anyone who has any insight!