The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Mixed-Signal Design >> How to implement S&H without hiden state?
https://designers-guide.org/forum/YaBB.pl?num=1064281661

Message started by toctory on Sep 22^nd, 2003, 6:47pm

Title: How to implement S&H without hiden state?
Post by toctory on Sep 22^nd, 2003, 6:47pm

I try to run the example in Mr. Kundert's paper(simulation sc filters with spectreRF), but I cant create a sample-and-hold block without hiden state.

I'v read the paper Hiden State In SpectreRF, and still got no idea about the S&H in pss analysis. Could any one help me about this?

Thx in advance.

Title: Re: How to implement S&H without hiden state?
Post by Andrew Beckett on Sep 22^nd, 2003, 10:08pm

What don't you understand? Ken's hidden states paper
describes a track-and-hold model which doesn't suffer
from hidden states - is that not what you want?

Regards,

Andrew.

Title: Re: How to implement S&H without hiden state?
Post by toctory on Sep 23^rd, 2003, 12:07am

Thanks for your earnest reply.

There is a Track & Hold Model in Mr. Kundert's paper, but what I need is a Sample & Hold block. Should I make a similiar model by use of the capacitor, switch and buffer to complete a S&H model? I have no idea about that.

I'm really interesting in how Mr. Kundert implements the S&H block which is used in the paper Simulating SC Filters with SpectreRF. In the paper, the refered file in List 1 is named "sh.va", but I cant find it in this website.

Thanks again.

Title: Re: How to implement S&H without hiden state?
Post by toctory on Sep 23^rd, 2003, 3:32am

Maybe I could use the Track-and-Hold block, and make the track aperture very very small (et. 1/1000 period). Would this method will bring any degradation to the simulation result as refer to the ideal Sample-and-Hold model?

Title: Re: How to implement S&H without hiden state?
Post by Andrew Beckett on Sep 23^rd, 2003, 4:02am

I spoke to Ken about this a couple of weeks ago, and he said that he'd used the models from the hidden state paper...

Now, in a real implementation, would it actually be a
true sample and hold (i.e. an infinitely narrow aperture)?
Or would there actually be some time for the storage element
(the capacitor) to charge up...

I'd expect most real sample and holds to be track and
holds (i.e. the capacitor is charged whilst the switch is closed,
and held when it is open).

I guess if you don't want the output to track the input
whilst the input is changing, you could always do
this with two track-and-holds, one after the other, with
the clocks inverted - that way the second tracks the
held voltage from the first. Something like that anyway...

Andrew.

Title: Re: How to implement S&H without hiden state?
Post by Ken Kundert on Sep 23^rd, 2003, 9:36am

The track-and-hold model can be found in http://www.designers-guide.com/Analysis/hs-models.va.
You can make a sample-and-hold but placing two track-and-holds in series. Or as you say, you can approximate a sample-and-hold with a track-and-hold by making the track time short.

-Ken