electrometer front end for dmm

In science, one accepts observations. In math, one does not (definitions and axioms OK; logic OK; yardsticks, NO).

So, science has always to be open to the possibility that the next (observation, experiment, test) will cause a theory to be invalidated. Math does not (though Godel looked hard at proof-of-consistency issues, and the next PROOF might cause a kind of invalidation).

Excellent and tactful answer. You don't really expect him to understand it, do you?

My understanding about things are obviously far above the plane of yours, that much is evident!

Would you like to try again with your childish comprehension responses? It better serves to publicize your inability to understand reality.

They name streets after you "One way" and if that is the way you think? You will come to a dead end!

Jamie

So we don't fully understand simple electronics because we still use these silly "constants found by trial en error", like e? I guess we don't understand anything about solid geometry because we still use that guesswork Pi thing.

In Jamie's defense, one of the metrics used by quantum field theory mavens is the number of undetermined parameters required to make the standard model fit experiment. Smaller is definitely better, because of course the more you understand, the fewer free parameters you need.

These aren't mathematical constants like pi or e, they're things like the mass of the electron, Planck's constant, and the fine structure constant. My rather out-of-date recollection is that there are about eight of them at the moment.

Cheers

Phil Hobbs

OTOH, without an understanding of the mechanism, a fit-to-the-curve is only so predictive. One of the many problems with the AGW model religion.

OK, but at some level, the "mass of an electron" is simply a conversion factor to fit it to our arbitrary units.

No, PI is a constant that has real meaning.. that is obvious.

Lets not be ridicules.. We are talking about constants that have no direct relations but are used to force an out come that matches a real live situation instead of using something like PI for example, which would make since.

There are many examples, and I'll give one. A very commonly used coil formula to calculate L.. L = 0.012*N^*u*A/l

Ok, so we have this, "u" being a permeability value which is ok, you need that of course, but then you have 0.012? This only tells me that something is unknown and fudge! Maybe a correct "u" table should be created to you don't need that magic number?

another one. L = u*N^*A/i that's easy and there is no magic numbers put in place, but then you have this. L = u*N^*A/I+0.45*D The first one does not care about the diameter but the second does and inserts a magic number.

I've seen Boltzmann and Planck constants being use in area's where It really has no need, yet be corrected with a magic constant down the line.

I used coils for a simple example for the simple minded.

Jamie

Thanks Phil, you understand exactly where I am coming from!

Jamie

q_e (er,

used

subject

the

up

and

researchers

of

And now for some really annoying news they are finding out that the fine structure constant isn't all that constant. We have done experiments that have noticeably bent it. (~5%)

?-)

But these measurements still are waiting for independent approval. As they say: "Strong claims need strong arguments."

More validation of the existence of Finagle's variable constant.

The LMP7721 has the nice benefit of low offset voltage, which can make guards work much better and means that lower feedback resistor values can sometimes be used. The disadvantages seem to be the restricted power supply voltage, and that it is not available in DIP package.

There is an interesting thread here (with no ultimate explanation) about performance variation with different date codes on some varieties of these low leakage opamps:

Chris

I've been trying to find your post for some time, and finally it shows up.

The bias current is the difference between two diode leakages. It is possible they are closely matched.

How did you measure the 170aA?

Did you use the capacitor at the input technique? If so, what kind of capacitor?

If you connected it to gnd, did you also try connecting it to VCC to see if cap leakage may have contributed to the measurement?

I've been working on an offset circuit to bring the guard offset to zero. So far nothing usable.

Thanks very much for the link. I added it to the list of Paul Grohe's posts.

Tempco, 1.5 MHz GBW

alue.

ut

(3pA, expensive).

mpere-stuff-anyhow

hat,

agnitude

That was an interesting discussion. Especially Paul's offhand comment about the LMP7221 using bootstrapped ESD diodes to lower the leakage, which injects current noise into the input. TANSTAAFL.

John posted a favorite AD op amp the other day that used a hybrid chopping scheme to null Vio. Patented, but something like it might adapted to null ib too...maybe. (Time for dinner!)

Cheers, James Arthur

It was the integrating pre-amplifier of a small ionization chamber. The integration capacitor was 10pF. It drifted linearly at a rate of about 17uV/s. It took four days to run into the rails. ;-)

The input node was wired in the air and I avoided fingering the capacitor and the amplifier, but I did not try to clean off the flux residues. I was surprised it worked as well as it did.

Jeroen Belleman

Amazing! Thanks

I found a bunch of Youtube videos that might be interesting. The first two are standard LTspice analysis, and the Noise Simulation might be helpful:

LTspice: AC Analysis

LTspice IV: Noise Simulations

The rest are 45 minute to 1 hr and feature Bob Pease and Paul Grohe, although Paul doesn't get to say much with Bob around:)

This is a partial list. I could not stand to watch them for more than 5 minutes, so I didn't try to get all of them. I noticed that Bob picked some very wierd circuits to analyze, when it would have been better to start with a simple circuit and get the basics down, then try more complicated versions. But he seemed to want to show how brilliant he was and make everyone else feel stupid.

He also got tangled up in knots on a simple bridge circuit. Apparently he never heard of a Kelvin connection. Anyway, FWIF here are some links:

Low-Noise Amplifier Design and Analysis

What's All This Femtoampere Stuff, Anyhow?

Precision Current Source

Milliohms, Micro-ohms, Nano-ohms...

Whats All This Scope Probe Stuff, Anyhow?

Whats All This High-Noise Amplifier Stuff, Anyhow?

Whats All This Current Source Stuff, Anyhow?

Current Sources

nput impedance of a typical dmm is too low (10 Megaohm or so). Any suggest ions on a simple electrometer adapter for a dmm? I can't afford a new elec trometer so I was hoping I could make a simple circuit to add to a dmm for these measurements.

Grohe

of

ectrometer IC from TI. See if the crystal you are using has enough capacita nce to hold a signal on the input of the fA leakage IC.

) But I

And if

ickle

this way.

as I

on top

ux.JPG

Pease would run this through the auto-dishwasher. Said it worked very well. jb