Cap with least leakage

[11-90-an]

Thinking of a sample and hold circuit that holds for say, 5 mins.... however, I think that the cap will leak significantly and the output would be inconsistent... any way to combat this?

[ElectricDruid]

Lots of stuff on this very topic in this thread:

http://synth-diy.org/pipermail/synth-diy/2005-January/121002.html

[antonis]

Thinking of a sample and hold circuit that holds for say, 5 mins....

Same but for 20 mins, say..
(can't recall the date for peak voltage detector of custom made Ni-Cd battery charger together with temp monitoring..)

[Rob Strand]

PCB leakage is also a problem.   See stuff on guard rings.

[11-90-an]

Lots of stuff on this very topic in this thread:

http://synth-diy.org/pipermail/synth-diy/2005-January/121002.html

So if done properly, i could hold down a signal for about 20 mins.... hmmm....
Some people stated 2 S/H in series, the first with small cap value, and the other with big cap value. However, what would be the switching system? Both would be on at the same time? Alternating?

I was thinking of 2 S/H in parallel, and each output will be fed into a 100 ohm resistor and averaged with a summing amplifier...

However, each of the components have to be matched really well, (especially the 100ohm resistors)

What do you guys think?

[Rob Strand]

Here's a few articles I rememeber.

Several days,
https://www.electronicdesign.com/technologies/analog/article/21801224/whats-all-this-capacitor-leakage-stuff-anyhow

Guard rings etc,
https://www.ti.com/lit/an/snoa664/snoa664.pdf