On Sun, Dec 07, 2003 at 12:23:08AM +0000, Dave Dilatush wrote:
> If you're expecting a 4-5 minute charge time for C2 (I assume
> you're figuring on achieving this by having a very low duty cycle
> coming out of U1B), and C2 is only a 1uF capacitor, both C2 and
> diode D2 must be EXTREMELY low-leakage devices.

The component values I've got now are just for getting the basic circuit
working; I was planning on using a 100uF or so to get it up higher.  I
wasn't planning on waiting 5 minutes every time I tweaked things
slightly.  :-)

> For example, suppose D2 were to leak just one nanoampere.  In 5
> minutes (300 seconds), one nanoamp of leakage will drag the
> voltage on C2 down by:
>
>   V  =  I * T / C  = 1E-9 * 300 / 1E-6  = 300 millivolts.
>
> More than a couple of nanoamps leakage, and the voltage on C2
> will never get to where you want it.

That is a good point.  The gate input leakage is specified as <= 0.1uA,
so that could be significant.

> If D2 is an ordinary silicon switching diode, its leakage could
> easily be that much; also, if C2 is any kind of electrolytic
> capacitor (tantalum or aluminum) then it, too, could leak that
> much.

Both of those are true.

> Also, when using ordinary switching diodes with clear glass
> bodies, note that they're photosensitive: impinging light will
> make them leak like crazy.

That's not something which had occured to me either.

> One final note: wouldn't this be a perfect way to use up some of those
> old PIC16F84's, or some other ancient, dinky PIC laying around?  Why
> play around with this funky analog stuff.

If I were more interesting in getting it working well than
experimenting, then yes.

> Hope this helps a bit...

It's certainly been interesting reading, thanks!

Cheers,

Chris

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listserv@mitvma.mit.edu with SET PICList DIGEST in the body