Forum Post: RE: AFE5809: Question on Differential termination resistor location for ADC clocks input

Hi Chen Thanks for quick reply. Quote "for your question #2, could you please let us know where you found: -----------But I also see some design as "Clock source--> AC caps--->100ohm --> AFE" .-----------" We use TI LMK00301 to provide LVDS or LVPECL clocks to AFE5809 ADC and CWx16. AC coupled, and hardware will be designed to be compactible for LVDS and LVPECL clocks. However, I read LMK00301 datasheet and found the situations are different for LVDS and LVPECL. Please see section 9.3.2 desceriptions and Figure 33 and 35 in LMK00301 DS. (1)For LVDS, it is clear, we will use the recommendation of AFE datasheet. 100ohm --> AC caps--->AFE, which is consistent with LMK00301 datasheet figure 33. For LVDS, it is written in section 9.3.2 that the output of LMK00301 need a resistor to construct DC bias. So AC caps have to be place between 100ohm and reverivers(AFE). (2)For LVPECL, AFE recommends AC caps--->100ohm --> AFE, however LMK00301 recommends the reversed placement. As you can see from Fgiure35, the Rpu and Rpd not only provide V_bias for RX inputs but also equivalent to Diff termination ≈100ohm(single end ~48ohm). As AFE5809 has internall DC bias on its clock inputs, so only the 100ohm ternimation resistor isrequired to between AC caps and receivers. Personally I think it is the same to place 100ohm between AC caps and AFE inputs, or the reversed placement. How do you think? Thanks.