Problems in a comparator with the sampled signal as inputs

[nxing]

problems with comparator

Hello everyone,

I am designing a comparator for the Pipelined ADC application and I have a problem with my comparator: When I use the continuously signal (since wave) as inputs for comparators, it works perfectly. however, when I use the sampled signal as inputs, it can not give the correct results? Also, I found that the outputs of the compartor are affected by the value of sampling capacitor. Does anybody have any suggestion.

Thanks!

P.S. The comaprator that I am using is a dynamic comparator with a build-in threshold. It appears in Allan's book of Figure 8.5-6.

 

[maxwellequ]

Re: problems in Comparator

I don't have Allan's book, so I don't know exactly what is your comparator (maybe you can post it...).

Anyway it does not make sense that you don't have the correct operation with a sampled signal. Probably there is something wrong with your testbench. Is the input common-mode voltage in place?

What happens: you comparator does not regenerate at all or you simply get wrong results? If you get wrong results, this is with which differential input voltage (i.e. if you put a large differential input voltage you still get wrong results ?).

It seams that you have a switched capacitor input network. Are you waiting the time necessary for charge distribution before making the comparator regenerate ?

If you give more details it will be easier to help you.

 

[nxing]

Re: problems in Comparator

Hi Max,

Thanks for your reply. Actually, I found the answer. It's because one of the regenerating transistors is in the deep saturation region. Also there are other problems from my sampling circuits.

Regards,

 

[mady79]

Re: problems in Comparator

Have you looked at the offset of the comparator ? .Probably increase the channel length of the 4 i/p transistor's to see some changes .