Designer’s Guide Community

neoflash

Community Fellow

Offline

Mixed-Signal
Designer

Posts: 397

Noise of S/H in Sampling Phase
Jul 14^th, 2011, 10:58pm

This is the S/H stage in pipelined ADC. I have a question about this noise analysis.

After the sampling phase, all charges on the Cs will be transferred to Ch.

The question is that whether the switch (3) contribute noise to the result. I have read some article and saying that it doesn't. I'm confused and feel it should impact since the noise charge is stored on Cs bottom plate .

Hope people can suggest and provide reference, thanks!

Back to top

sh_noise.PNG

IP Logged

raja.cedt Senior Fellow Offline Posts: 1516 Germany	Re: Noise of S/H in Sampling Phase Reply #1 - Jul 15^th, 2011, 12:07am hello, where is switch 3 here? Up to my knowledge every switch wil contribute noise, so when you sample there is some noise charge will be accumulated on Cs, and when hold it will get deposited on Ch. This is roughly kt*c1/c2^2, in addition to this there kt/ch on holding cap. Thanks.
Back to top	IP Logged

HdrChopper

Community Fellow

Offline

Posts: 493

Re: Noise of S/H in Sampling Phase
Reply #2 - Jul 15^th, 2011, 5:37am

Hi,

Both switches 1 and 3 will contribute to sampled noise on Cs. Moreover you could consider both of them as a single switch, thus a single thermal noise source with an ideal resistor in series.
As a consequence the sampled noise power will be KT/Cs, which is independent of the Ron from switch 1 and switch 3 added.
Bottom line: both switches contribute but the noise power is not Ron dependent.

Hope this helps
Tosei

Back to top

Keep it simple

IP Logged

neoflash

Community Fellow

Offline

Mixed-Signal
Designer

Posts: 397

Re: Noise of S/H in Sampling Phase
Reply #3 - Jul 15^th, 2011, 9:24am

Switch 3 is the switch in the middle labeled with (3).

raja.cedt wrote on Jul 15^th, 2011, 12:07am:

hello,
where is switch 3 here?
Up to my knowledge every switch wil contribute noise, so when you sample there is some noise charge will be accumulated on Cs, and when hold it will get deposited on Ch. This is roughly kt*c1/c2^2, in addition to this there kt/ch on holding cap.

Thanks.

Back to top

IP Logged

neoflash Community Fellow Offline Mixed-Signal Designer Posts: 397	Re: Noise of S/H in Sampling Phase Reply #4 - Jul 15^th, 2011, 11:43am From simulation, the total noise appear at Ch in hold phase is 2*KT/C no matter Ron(1,2,3)'s value. Just another good example of KT/C theory. It is just that this complex network and sample/hold charge transfer make it look so blurred.
Back to top	IP Logged