electrometer front end for dmm

I want to make measurements on pyroelectric crystals but I think the input impedance of a typical dmm is too low (10 Megaohm or so). Any suggestions on a simple electrometer adapter for a dmm? I can't afford a new electrome ter so I was hoping I could make a simple circuit to add to a dmm for these measurements.

Scot

Reply to
bonesisalive
Loading thread data ...

I guess you could make a JEFT front end?

you can also use a JFET input type op-amp

you need to look around but I think you can obtain up around 1G ohm input Z and I am sure you can get more. P.S.

they make active fet probes. Look for "high impedance active probe" circuits, I just found a couple...

Jamie

Reply to
Maynard A. Philbrook Jr.

Use LMC6482 as a voltage follower. $2.08 Digi-Key, Newark, etc

Note 13: Guaranteed limits are dictated by tester limitations and not device performance. Actual performance is reflected in the typical value.

formatting link

Might add a 10k from output to ground to provide a load for the output stage.

Reply to
Steve Wilson

co, 1.5 MHz GBW

Beat me to it. The LMC662 is another choice (cheap), and the LMP7721 (3pA, expensive).

formatting link

Some articles for our OP:

formatting link
uff-anyhow
formatting link
Cheers, James Arthur

Reply to
dagmargoodboat

I've used an LPC661, which turned out to have an amazing 170aA of bias current. Maybe I just got lucky, because that's *much* better than the datasheet says it is.

Jeroen Belleman

Reply to
jeroen Belleman

Apparently the bias current comes from the input protection circuit. Paul Grohe mentioned sonewhere they made some devices without input protection and could not measure the input current. They could not use the devices however. They blew the inputs whenever they tried to handle them.

Paul has some interesting posts on how to wire the devices for lowest leakage and how to clean them. You have to sift through the crud to find them, but they are worth reading. Sorry for the wrap.

formatting link
$20grohe%22$20lmc660

Here's the tinyurl version

formatting link

Reply to
Steve Wilson

You could get lucky and have the Vcc and Vss ESD diode leakages cancel.

I guess you could trim the supplies to *make* them cancel.

This is the pA parts tester that I built:

formatting link

formatting link

formatting link

Its leakage was probably about 10 fA.

A simple 2N7000 source follower should be pretty good. That will stay halfway on for hours with the gate floating, so leakage must be pretty low.

formatting link

--
John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
 Click to see the full signature
Reply to
John Larkin

A long time ago, someone posted a reason why this won't work. I have been trying to find the post but have had no luck. I forget the explanation, but it should be easy to test.

Simply set it up to measure the input bias current, then adjust the supply voltages to see if the bias current can be switched to the opposite polarity. If so, it has to go through zero somewhere.

Reply to
Steve Wilson

Actually, if the bias current could be trimmed to zero by adjusting the supply voltages, the required supply voltages for the specified bias current whould be noted in the datasheet. And if the diode leakages acted like resistances, the bias current would change as the input voltage moved through the operating range.

There should also be a plot somewhere of the input bias current vs VCC and VEE, and another one showing the bias current vs input voltage.

The LMC6482 and LMC660 specification for input bias current do not mention the supply voltage, and there are no plots showing changes in the bias current as a function of input voltage or supply voltage. The only plots involving bias current show the increase with temperature.

So clearly the input bias current leakage does not act like resistances, and it is not possible to trim the bias current to zero by changing the supply voltage.

This raises the question of what creates the input bias current, and where does it go?

Reply to
Steve Wilson

empco, 1.5 MHz GBW

ue.

3pA, expensive).

re-stuff-anyhow

Very impressive. I remember a Bob Pease article somewhere about just that, that the ib was lower, caused by the protection diodes, and spec set by testing throughput.

The LMC6001 was, I think, tested for lower ib.

I goofed on the LMP7721--ib=3fA, not pA (but what's three orders of magni tude between friends?).

Cheers, James Arthur

Reply to
dagmargoodboat

Clearly not possible, because the data sheet doesn't specify it?

--
John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
 Click to see the full signature
Reply to
John Larkin

No, because there is nothing on the datasheet to indicate the input voltage or VCC where the input bias current is measured.

Easy to prove. Go ahead and measure it.

Reply to
Steve Wilson

If it were possible to set the input bias current to zero by adjusting the suppply voltages, they would certainly make a very big deal of it and show how to do it. It would be under the title "Adjusting The Input Bias Current To Zero", or something similar. There would be plots showing the input bias current vs time and temperature.

SED would have a ball posting articles on low current measurements. There would be a race to see who could get down to 1 electron per second.

Keithley would go nuts. They would have to redesign all their femtoamp current meters and rewrite all their low current measurement articles, such as "Counting Electrons: How to measure currents in the attoampere range",

formatting link

Reply to
Steve Wilson

...

There are other possible behaviors; if the output is somehow thermally coupled to the leakage on the inputs, it could have positive feedback. So, you could trim the leakage near zero, only to have it snap from one rail to the other. There might be no stable equilibrium.

A dual FET used as preamplifier, with a standard op amp behind it, was the old solution. If duals weren't at-or-near unavailable, it'd still work.

Reply to
whit3rd

Think about an eprom floating gate. It must leak well under an electron per hour.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

The leakage currents will be in the fA range, so the diode power dissipations will be in the fW range. That will raise the diode temperatures by nanokelvins. Not much chance of thermal runaway.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

It must be possible. At some positive input voltage, the upper ESD diode conducts. At some negative voltage, the lower ESD diode conducts. There is obviously a zero current point somewhere between. Question is, where?

Of course, nobody would guarantee this for a $1 opamp, and certainly wouldn't test for it.

One easy test would be to make an opamp follower with the + input floating. Turn it on and see where the output settles.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

Nah,if Bob Pease designed it, he would arrange the thermal coupling to give negative feedback. He discussed thermal coupling in one of his articles.

Just for reference,

formatting link
has a complete set of Bob Pease articles including comments. The site is in Italian and I can't find the page that refers to the pdfs. However the pdf urls are here. They are around 15mB each:

formatting link
formatting link
formatting link
formatting link
formatting link
formatting link
formatting link
formatting link
formatting link

I haven't had a chance to go through the articles and find the one on thermal coupling.

Reply to
Steve Wilson

whit3rd is referring to the thermal coupling from the output stage back to the input circuits. The LMC660 can supply 40mA with 15V Vcc. If it delivered 40 mA to a 187 ohm load, it would dissipate 300 milliwatts in the output stage.

(7.5^2) / 187 = 0.300

That would definitely cause the chip to get hot.

Reply to
Steve Wilson

Well, if you want low leakage, don't load the opamp hard. Don't load it at all.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.