electrometer front end for dmm

My "solid state" 610C cost about that. Wonderful gadget.

I've done this with some CMOS switches to get to the fA level territory.

The fun thing is to automate trimming.

t impedance of a typical dmm is too low (10 Megaohm or so). Any suggestion s on a simple electrometer adapter for a dmm? I can't afford a new electro meter so I was hoping I could make a simple circuit to add to a dmm for the se measurements.

You need to learn A LOT MORE about pyroelectric crystal detectors before yo u even begin to waste time on an amplifier, which are simple and straightfo rward for the most part.

From the introductory literature:

"The spontaneous polarization of the detector material induces a net positi ve charge on one face of the crystal and a net negative charge on the other. T he amount of charge on the crystal faces is small and is fairly quickly neutra lized by deposition of ions from the surrounding atmosphere or by electrical conduct ion. Pyroelectric detectors, therefore, can only be used to detect modulated rad iation, and the modulation must be fast enough that the neutralization of the surfa ce charge can't keep up. These detectors can be operated in either the voltage mode or the current mode, and one mode or the other will be more appropriat e depending on the application. In the voltage mode, it is the voltage genera ted across the entire pyroelectric crystal that is detected. The incident radia tion is absorbed on the exposed or top surface of the crystal. The thermal energy i s conducted through the crystal to the back or bottom face bringing the entir e crystal to a roughly uniform temperature. The crystal is more or less unifo rmly polarized with net positive charge on one crystal face and net negative cha rge on the other. Clearly, voltage mode detection can only be used at modulation frequencies low enough so that the thermal energy has enough time to be conducted through the crystal before the incident radiation starts back the other way. Current mode detection is appropriate for high frequency modulation wh ere the temperature of only a relatively thin layer of the crystal at the expos ed surface is affected by the incident radiation. The modulation frequency is fast enough that the thermal energy is not conducted into the bulk of the crysta l. Since the absorbed energy is not distributed over the bulk of the crystal, the temperature excursions of the crystal material immediately adjacent to the exposed face are maximized as is the charge generated on the exposed face. In current mode operation, it is the current flowing on and off the electrode on the exposed face of the crystal that is detected. The temperature at the back s urface of the detector and the charge on that surface doesn't change at the high modulation frequencies used for current mode detection. In voltage mode operation, the amount of surface charge generated on a pyroelectric detector by the incident radiation is typically quite small. T he detector is basically a capacitor with a shunt or leakage resistance. A typ ical pyroelectric detector will have a capacitance on the order of 10 pF or more and a shunt resistance typically greater than 1012 ?. The responsivity of a typical pyroelectric detector operated in the voltage mode peaks for modulation frequencies somewhere in the 0.01-0.1 Hz range. It falls off at lower frequ encies due to discharging or neutralization of the surface charge and at higher frequencies due to the thermal mass of the crystal. A preamplifier, of cour se, necessarily provides an additional parallel shunt path for neutralization o f the surface charge, and the resistance of that path must be greater or at least not smaller than the shunt resistance of the detector itself if it is not to si gnificantly reduce the detector signal and degrade performance."

Right around that value, in very round figures. If you assume 10 fF capacitance per cell and a 3-V gate voltage shift for reliable forgetting, a 20-year lifetime implies

dq/dt < (3V)/(6E8 s)*1e-14 F* 6.2e18 (e-/C) = 3E-3 electrons/s, or about

Cheers

Phil Hobbs

Yikes, pretty good guessing.

I think a cool science project could be done with an eprom, specifically resolving single-electron leakage. Maybe.

Super-low-leakage opamps might be package dominated. The TO-can versions have way more bias current than the DIPs. Sadly, the V- pin is almost always next to the IN+ pin. I guess there's not enough market for a better pinout.

I should quantify the 2N7000 leakage. You can do some fun stuff with a 2N7000, a battery, a resistor, and an LED. There must be small-signal mosfets, without gate protection, that have really low leakages.

The SOT23 ones are better, and of course you can always use a dual--the right-hand section is much better isolated.

I still have some 3N163s in my drawer, complete with copperclad steel springs shorting their leads together. Might be fun to measure.

Cheers

Phil Hobbs

Oops, a typo :)

Assuming the absence of illumination, anyway. :)

Tim

Just like the olden days ;)

Typically an open circuit grid goes to -2 or -3V. Negative current is limited by cathode emission (the high-energy tail of thermally freed electrons -- thus, should be higher in thoriated tungsten types) and balanced by grid leakage (emissions from the grid wires being hot by proximity, and dissipation if forward biased).

Isn't there a similar thing with old fashioned JFETs, dependent on Vds?

Tim

"John Larkin" wrote in message news: snipped-for-privacy@4ax.com...

A related subject,

Tim

Hardly matters--diodes don't even come close to obeying the 'diode equation' (AoE has a graph somewhere--the exponential constant frequently differs from kT/e by a factor as large as 2).

What we laughingly call "high level injection" happens a lot sooner in a real diode than in a diode-connected transistor.

Cheers

Phil Hobbs

Dunno, never measured it. The pHEMT thing was for a client gig a couple of years back, for a research lab in the Far East. They had a biochemical application where they really wanted to measure 1 nA in a

I got pretty close--within a factor of three, which was better than I thought I could do, and learned a lot in the process. Since then I've used quite a few microwave transistors in lower-frequency applications, because they're magic. SiGe:C transistors are especially good--a BFP640 has an fT of almost 50 GHz, together with a practically infinite Early voltage, and a 1/f corner below a kilohertz. You can do a lot with that, if you're careful about the PCB layout.

Cheers

Phil Hobbs

So, with grounded source and positive drain, the gate goes negative? Sort of an electron venturi effect?

Yup. Odd as can be, but I have data.

Cheers

Phil Hobbs

Cool. Super-slow UV imager.

Some of the first CMOS imagers were actually DRAM chips, work done at JPL as I recall.

Oh, I believe you. It's very cool.

A Pease Puzzler was a bipolar NPN transistor: ground the base, pull up the emitter, and the collector goes negative. Bob explained it as photons generated by the b-e zenering and photoelectric generation in the c-b junction. But maybe he was guessing.

Should be able to prove that with a PMT tube.

Is there any info on what NPN device he was using?

tm

Sounds like capacitance internally to me, unless this is a constant output at the collector?

Jamie

Pease claimed negative DC at the collector.

Capacitance wouldn't make negative DC. Unless the b-e junction was making AC, as a zener relaxation oscillator, and the c-b junction was rectifying that.

That's the point--it is DC.

Cheers

Phil Hobbs