From: email@example.com (Tom Bruhns)
Subject: Re: Low leakage parts
Date: 4 Jan 2003 08:57:54 -0800
References: <firstname.lastname@example.org> <8npR9.16063$jM5.email@example.com>
NNTP-Posting-Date: 4 Jan 2003 16:57:54 GMT
Thanks, Spehro, for the observations in my absence over the past few
hours. :-) Yes, you are right. The circuit was left with no voltage
applied, and since the FET is ON rather hard, it's safe to assume the
drain and source are at very close to the same potential. The gate
just has a capacitor, as noted, between it and the source. No PCB
involved. During my absence, there weren't any other significant
voltage sources nearby...things were all turned off. I could have put
the thing in a Faraday cage, but really didn't expect it to still be
charged after that long anyway. Now that I know it can hold a charge
for quite a while, I may try the complementary experiment, with it
"off" and power applied to the drain circuit. The pin layout on the
TO-220 mosfet puts the drain in the middle and the gate off to one
side, so surface leakage should be detected easily. (Of course, the
same surface leakage should have been/has been discharging the
gate/cap for the past three weeks, too.)
Perhaps Phil lives in an environment with an abnormally high number of
ions floating around in the atmosphere, but the roughly 0.1fA of
leakage I'm seeing in 0.1uF polyprop caps at 10V suggests about 10^17
ohms effective shunt resistance, and that includes the air in my
office as well as surface leakage over the outside of the caps and the
bulk resistance of the dielectric in the caps. I expect they'd
discharge rather faster in bright sunlight.
On the subject of FR4 PCB leakage: I've seen bad leakage on uncleaned
boards, but modern materials and manufacturing techniques yield some
pretty low leakages at room temperature, even in moderatly high
I am NOT (!!!) saying that it's good engineering practice to RELY on
such low leakages, I'm just making observations that modern materials
and methods are resulting in some pretty impressive (to me)
performance. BTW, Bob Pease related some similar results to me in
another area. I think the careful process control is reaching into a
whole lot of different areas.
Spehro Pefhany wrote in message news:...
> On Sat, 4 Jan 2003 14:21:33 +1100, the renowned "Phil Allison"
> >"Spehro Pefhany" wrote in message
> >> On Sat, 4 Jan 2003 11:02:34 +1100, the renowned "Phil Allison"
> >> wrote:
> >> >
> >> > I suspect there was a DC leakage path HOLDING the 0.01 uF cap
> >> >up that was the real story here.
> >> He did say he removed power for the 21 days.
> > ** Does not prove there was *no* leakage path in his design.
> No, but no visible voltage source is a darn good start. Where do
> you think the DC could come from? AC sources are a dime a dozen, but
> mystery DC leakage is not so common. Electrolytic action, perhaps,
> if there was a lot of humidity. Regular electrolytics won't hold
> a charge for very long.
> > BTW I, for one, never take poster's words literally. Having done
> >electronics servicing for 30 years I know to disbelieve EVERYTHING I am
> >old - especially when it is only in writing.
> Yes, I've seen a *lot* of that stuff too and tend to discount the
> words until I get a measure of the person. But Tom Bruhns seems to
> know what he's doing.
> Best regards,
> Spehro Pefhany