From: Winfield Hill
Subject: Re: Low leakage parts
Date: 4 Jan 2003 13:56:19 -0800
Organization: Rowland Institute
References: <email@example.com> <firstname.lastname@example.org>
X-Newsreader: Direct Read News 4.20
Winfield Hill wrote...
> John Larkin wrote...
>> I did a post a while back: a 2N7000 with a 9-volt battery and an
>> LED in its drain, gate hanging open. This can be set on or off
>> and stay in that state for many days. You can also perch the
>> gate voltage just at the turnon point and get the LED to stay
>> on dimly, with no perceptable brightness change over many hours.
>> ... I figured that the gate leakage was in the 100 electron/second
>> range, maybe much less... attoamps.
> At low currents, such as 1 to 5mA, the transconductance of a
> 2n7000 approaches that of a transistor, where an 18mV change
> makes a 2x current difference. If we assume your FET stayed
> within 2x current over 5 hours, that's dV/dt < 1uV/s. Given
> a 40pF gate capacitance, we get I = C dV/dt = 4E-17A, or 40aA,
> which is 250 e-/sec.
OK, I've taken a few measurements and can use them to improve
the calculations. I measured a 2n7000 in the vicinity of 0.3
to 2mA, and found its transconductance to be about 30 to 40%
that of a transistor, with a 2x current change corresponding
to a 75mV change in gate voltage. Second, both Ciss and Crss
must be charged/discharged, which is more like 60pF total. So
a 2x change corresponds to a charge of q = C V = 4.5pC and if
it occurs over a 5 hour period, that corresponds to an average
gate current of 2.5 x 10^-16 or 0.25fA, which is 1560 e-/s.
> You don't mention the FET's gate-drain voltage, but assuming
> 2V, we get R > 5E16 or 50000000G, or 50M Gig-ohms, another
> interesting value. The high-quality plastics used by most
> semiconductor manufacturers must be impressive, especially
> considering that the gate leakage is likely in the gate oxide
> rather than the plastic.
Using the above values, revise this downward by about 6x to
8P (8 peta-ohms = 8E15), still assuming a 2V drop.
Also, I have just repeated John's experiment, a mini two-hour
version, largely confirming the above. But with a few rather
interesting results. :>)