clock feedthrough vs kickback noise

[pansoo kim]

kickback noise

In adc, I have knew that we can use mos switch at input stage to prevent kickback noise from latched comparator to preamp stage.

But in reality when using mos switch, the clock feedthrough from the mos switch is more significant than kickback noise from comparator.

How about you?

 

[sunking]

kick back noise adc

use bottom sampling and try

 

[flushrat]

preamp is to isolate kickback noise

 

[sankudey]

R u working with MOS only or BiCMOS?
The way I saw is to isolate is to aploy level shifter atleast at two levels...1. when you are coupling from pre-amp to latch...use one level shifter..this helps....and reduce/protect upto some extent the pre-amp from the unwanted forced signal distortion.....2. at the input of pre-amp...use another level shifter....this pretocet the very input of the comparator from the above distortion........

If your accuracy requirement is not too high (>=4/5mV)...you also could go for two-stage pre-amp...which itself very rugged...but definitely power and area is sacrificed.....
....one more good topology to prevent kickback noise (may not be too 'to the point' for u) is the re-configuration of the latch where from the kick back noise got generated....the references are...1. Johns Martin book (only Idea)....2.JSSC..Jensen....16GHz HBT comp....Sep-2003