High value resistors, gigaohm to teraohm.

[Dmytry]

Is there some good way to make 10⁹ to 10¹² ohm range resistors? (buying them is expensive).
I've been experimenting with pencil traces on paper, that is understandably unstable, going to try with pencil traces on the ceramics then cover it with glue. I was wondering if there's some well known way to do it, from the good old days.

I need those resistors for sensitive ionisation chamber project (measuring currents of several fA. 1000 cubic centimetre air ionisation chamber has current of approximately 1E-8 A at dose rate 1 Sv/h , i.e. 1fA at 0.1 mkSv/h .). As an alternative I was thinking to use reed switch relay to discharge.

 

[Averagesupernova]

If the actual resistance is not critical and just needs to be a bleeder then you may find that the reverse leakage of a diode is enough to do it. I suspect you want something stable though with a known spec'd resistance.

 

[Dmytry]

If the actual resistance is not critical and just needs to be a bleeder then you may find that the reverse leakage of a diode is enough to do it. I suspect you want something stable though with a known spec'd resistance.

I can measure it no problem, just need it to be linear and not have a lot of stray capacitance and not to work as voltage source (liquid in capillary tube wouldn't work well).
I may end up building some altogether different design, using reed switch to discharge the input into op-amp periodically. The reason I want to go with 'normal' amplifier is that this way i can perhaps detect individual tracks.

 

[Averagesupernova]

I guess I would suggest multiple mustiple high value resistors in series. Any contamination can cause significant leakage.

 

[es1]

Interesting problem. This website recommends using an integrator which would remove the need for the resistor.
http://www.tmworld.com/article/319701-Femtoamp_fA_measurements.php

You do need a opamp with less than a fA input bias current though. But I guess you needed that anyway to pick the voltage off the resistor.

I've heard of amps with 10s of fA input bias for light sensing circuits. Not sure about less than 1 fA. Maybe you can do a manual trim on a 10fA amp or something...

 

[Dmytry]

Interesting problem. This website recommends using an integrator which would remove the need for the resistor.
http://www.tmworld.com/article/319701-Femtoamp_fA_measurements.php

There's some problems with math in that article... e.g. 1fA , 100 seconds, 100uF cap, you will get 1 nanovolt, not 100 microvolt. should use 1 nF and 100 seconds to get 100 microvolts, of course.

You do need a opamp with less than a fA input bias current though. But I guess you needed that anyway to pick the voltage off the resistor.

I've heard of amps with 10s of fA input bias for light sensing circuits. Not sure about less than 1 fA. Maybe you can do a manual trim on a 10fA amp or something...

I found the less than 1fA opamps... LMC660 , LMC 6044 , and the best one (made for ionisation chambers) is LMP7721 .

My other idea is to integrate on the internal capacitance of the ionisation chamber, then discharge through opamp using a reed switch, getting a voltage pulse. Voltage pulse would be a lot easier to amplify, I can remove low frequencies. And this way i can use high bias current op amp.

Also this guy built real good chambers using just a FET:
http://www.techlib.com/science/ion.html

 

[es1]

Just out of curiosity, what's the advantage of using the internal capacitance of the chamber over just a cap?

That is an awesome link. Lot's of good ideas there. The guy's a really good circuit designer. I would definitely copy if I were to build a chamber. :)

 

[Dmytry]

Just out of curiosity, what's the advantage of using the internal capacitance of the chamber over just a cap?

That is an awesome link. Lot's of good ideas there. The guy's a really good circuit designer. I would definitely copy if I were to build a chamber. :)

Well, faster voltage rise. Capacitances are small... if I have current of 1fA, then 1pF takes 1000 seconds to charge to 1v .
He says tho he get 40fA at background which is interesting. Maybe he got a lot of radon.