The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: Mike Monett
X-Mailer: Mozilla 2.02 (Win16; I)
Subject: Re: Low leakage parts
References: <email@example.com> <3E17D633.378A@sneakemail.com> <3E184B38.588B@sneakemail.com> <firstname.lastname@example.org> <3E186B3C.5BA0@sneakemail.com> <email@example.com> <3E18805E.4AA@sneakemail.com> <firstname.lastname@example.org>
Date: Tue, 07 Jan 2003 16:00:40 -0500
NNTP-Posting-Date: Tue, 07 Jan 2003 16:00:09 EST
Organization: Bell Sympatico
Tom Bruhns wrote:
> Mike Monett wrote in message
>> Have you thought about the LMC662: 2fA typ, 1.3uV/C tempco? This
>> would raise the minimum current that can be measured, but the
>> drift with temperature is attractive for long-term meaurements.
> An interesting experiment is to plot the input bias current versus
> common mode voltage. You may find that it passes through zero at
> some voltage. It did with the old CA3130 op amps. If that happens,
> and it's stable enough, you can get to exceedingly low input bias
> current. It may even be possible to bootstrap the power supplies
> to keep it low over a moderate range of input voltages.
That would be a nice trick!
> What was the best one could do with "electrometer" vacuum tubes?
It must have been pretty good. I recall a class at MIT in 1967 where
Prof. Searle mentioned measuring the diode IV curve down to
attoamps. While everyone sat stunned at the concept, another
professor laughed and said this was less than one electron per
second. So they were in the right ballpark. I assume they used
vacuum tubes for the measurement.
> What are the current best practices for low input bias current
The Keithley model 6430 claims 10aA resolution on the 1pA range
using a 5 second RC filter. The voltage burden is < 1mV.
Their graph shows the noise at 200aA p-p.
> I assume there are ways to get well below 1fA, but with what sort
> of bandwidth? (How few electrons can we reliably sense
> electrically?) 1fA*1sec at 1pF is only 1mV, but even 0.01fA for a
> long enough time (an hour, or a day) on a 1pF cap becomes easy to
An exciting technology is the single electron transistor. NIST
claims a sensitivity of 10-5 e in a 1 Hz bandwidth:
Using an RF-SET, Schoelkop claims a bandwidth of 100MHz:
C. P. Heij shows an inverter operating at 25 mK, and why room
temperature logic devices will be difficult to make. He shows a
curve of gain vs temperature up to 200mK:
These are cryogenic devices, but many companies are working on room
temperature versions. Matsumoto shows clear Coulomb staircase curves
It will be a while until we can run down to Radio Shack and buy one,
but they seem quite promising. Just think about the RFI/EMI
> If you connected that LMC662C as a follower, leaving the NI input
> floating, it would eventually drift to the zero bias current point
> (if such point is within the input common mode range) but it would
> take a while to get there, at probably less than 1mV/second slew
> (2fA and probably a bit more than 2pF input capacitance). ...
> Maybe some of the lurkers are doing work in the very-low-current
> area and can offer some insights. For me, it's just an area of
> casual interest.
> Cheers, Tom
That is a great idea! I hope someone tries it and posts the results.
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup