Archive of the former Yahoo!Groups mailing list: MOTM

previous by date index next by date
previous in topic topic list next in topic

Subject: Re: [motm] Cascading S&H

From: "Tony Allgood" <oakley@...>
Date: 2000-04-16

Hi all,

The Roland 'The Synthesizer' book states that three S/H can be cascaded
using a common clock. Because of this I decided to do a four stage S/H
at some time in the future, but I also came upon the problem of the
aquisition time of each s/h. In other words, once the first S/H is
triggered to aquire, the next S/H in the cascade is also sampling the
new incoming signal. No good for cascaded patterns at all. However, you
could bodge this by creating delays in the triggering chain. This can
easily be achieved by simple RC circuits, the last S/H being set to
trigger before the previous one. There is no need to have a second
series of 'buffer' S/H if the delay is small enough, and the aquisition
time is short enough. The BBD has to because of the sheer number of
stages involved. With four stages, you could get away with this delay
method.

Did that make sense?

BUT it could be that the natural delay in clocking a signal through the
S/H will create enough time for the next S/H to reach the correct value
anyway. Certainly the Moog Modular page at that university would suggest
this.

Regards,

Tony Allgood Penrith, Cumbria, UK

http://www.techrepairs.freeserve.co.uk/projects.htm