V7-45 electrometer made in Belarus (welcome in attoamps world)

[bsw_m]

I bought this tool and am waiting for delivery.

This is a very rare device. According to my calculations, only about 600-1000 of these devices were produced. And it's great luck that I was able to find this device. Even if not complete.

I may be wrong, but I think this is the most sensitive instrument on Earth that can be put on a table. And which does not require vacuum, low temperatures and other things for operation.
Datasheet says, that this instrument can measure 1E-17A, with resolution 1E-18. Tests show that this device is really capable of measuring currents 1E-18A, and have sensitivity in 3E-19A...1E-19A range.

This instrument have first production date is 1986 year. Design of this instrument was carried out by MNIPI (Minsk Research Institute of Instrument Construction).
Before this instrument MNIPI also have other old instrument that can measure 1E-17A: V7-29, that have start production in 1972 year.

Some information about architecture of this instrument.
The current measurement in this instrument is based on current integration. Amplifier in current integrator is chopper stabilized.
All other info: restoration, photo, tests and others is coming soon. After delivery this instrument.

Historical:
This device became possible thanks to the work of such a person as Konstantin Sergeevich Shuklin. The electrometric unit and DRK-6 were developed by him. The appearance of the DRK-6 was preceded by almost 30 years of work.
The chief designer of this device is Glubokiy Iosif Andreevich.

[Vgkid]

Looking forward to this.

[TimFox]

I do not read Russian, but from the technical article I assume that this uses the same basic idea as the American classic Cary vibrating-reed electrometers.  According to a Varian advertisement from 1971  https://pubs.acs.org/doi/pdf/10.1021/ac60304a744   it could "detect" currents down to 10 aA.

[TiN]

Not quite, it's using PMT for current integration (or rather charge, with conversion). So it does require vacuum and have a tube inside

[TimFox]

In the article, is that a photo-multiplier (with photocathode) or an electron-multiplier in the envelope with the vibrating structure?  The Cary had an external head for the vibrating reed, but not connected to an external vacuum chamber.  Besides construction quality, these devices need a low-noise amplifier after the reed, but can exploit a phase-sensitive detector at a frequency above the 1/f noise corner of the amplifier.

[NWerner]

electron multiplier.

[MadTux]

Interesting

But then it probably can only measure current at a certain voltage drop, since it needs a certain voltage potential to strip electrons off the cathode in electron multiplier tube?
Not like Keithley with transimpedance amplifier that has essentially zero voltage drop.
Also, how good is the linearity of a electron multiplier tube? A few % at best, I guess?
Vibrating-reed electrometers are the coolest, IMO ;-)

[Bud]

What are the practical applications for this sort of test gear?

[MadTux]

Anything with ions/electrons, like measure beam current in mass spectrometer, electron microscope....
Also nice to select low leakage JFETs for repairs, leakage and material resistance in general...
Probably also chemistry, pH elctrodes and other electrochemical stuff works best with no current...

[TimFox]

Electron multipliers are useful for detection of various charged particles, including electrons and ions.  The linearity is good at low levels, until the increased output current disturbs the voltage distribution along the channel or in the discrete voltage divider network.

[exe]

Waiting for tear down

[RoGeorge]

Now i start clean it with ultra pure Methyl-alcohol, high pressure pure nitrogen and soap

- To clean small items against residual charge leaks, for example a very low current IC, does the order matters?  Whater with soap first, then alcohol, or alcohol followed by water with soap cleaning, which order do you use, please?

- Another question, can a normal pressured air can (gas duster) be used instead of high pressure nitrogen?

[bsw_m]

Simplified work principle the amplifier in external integrating unit:
Amplifier use chopper topology: as input modulator work DRK-6 (dynamic resonance capacitor model 6, or a more familiar term in English vibrating resonance capacitor), after the AC amplifier, demodulator on jfet and DC amplifier.
For highest range this amplifier work as classical TIA, and use in their feedback 100MOhm microwire precise resistor.
On other ranges amplifier work as integrator with capacitors in their feedback. This solution improved the thermal stability of the amplifier and eliminated Josephson noise.
At all this amplifier have low noise (not more 3aA p-p, typically 1aA p-p), low offset - typically 3aA, and have "rock-solid" temperature and long-therm stability.

[MadTux]

Now i go to my radioactive storage to get hundred atto-amps test current from it... stray tuned!

Like a beta source with 600 decays/s. Or alpha with 150 decays/s 

[guenthert]

[..]
Electrometric external block (vibrating capacitor ДРК-6, chopper electrometric amplifier, control of vibrating capacitor resonance circuits, precision voltage divider, electrometric relays)

[..]

    Those sapphire insulator within the space restricted triax connectors were nice (thanks for sharing), but if I understand correctly the magic happens in those external blocks.  Is there a schematic of those available on-line?

[bsw_m]

but if I understand correctly the magic happens in those external blocks.  Is there a schematic of those available on-line?

You right! Schematic is not available on-line, but some posts ago I describe the principles of work.

[NWerner]

At https://ampnuts.com/ you mention the possibility of MEMS vibrating caps.

When I first learned about vibrating capacitor the same thought occured to me.
As the technology for MEMS vibrating capacitor should be fairly mature since >10years
I wonder why we dont see any MEMS based electrometers. There are some scientific
papers - but thats it. A vibrating caps electrometer has the potential for
  - lower input current
  - robust input protection
  - reasonable cost

Where is the catch?
Does anyone has any insight about what might be the obstacles to the realization of such an instrument?

[Kleinstein]

One tricky part is the surface potential: The capacitor will show some offset voltage that can depend on the temperature an gas composition. So while there is some input chopping action, there can still be some offset-drift.

I am not so sure that MEMS would really help here. Mems tend to use electrostatic drive and this can give some direct coupling. There is also a possible insulation problem if all in in a single piece of silicon.
I don't see a large advantage from small size on the mechanical side, if the resonator part is made right. Going higher in the frequency and smaller in size also has disadvantages. So the classical cm size capacitor has it's advantages.
Modern MEMS parts tend to be quite small and high in resonance frequency.

[NWerner]

Often MEMS accelerometers are encapsulated in vacuum to minimize damping while in resonance.  SiO2 as an insulator is pretty good but could prove a weakness when compared to ptfe or sapphire especially with those tiny dimensions??

[branadic]

Where do you see zepto amps?

1 pA (pico) = 1e-12 A
1 fA (femto) = 1e-15 A
1 aA (atto) = 1e-18 A
1 zA (zepto) = 1e-21 A
1 yA (yocto) = 1e-24 A

So you look at 100 aA resolution, not 100 zA.

-branadic-

[branadic]

Ups, my fault.

-branadic-

[Kleinstein]

With the capacitor charge method there has to be a integrator reset from time to time. This could cause trouble of some kind. Some cosmic ray or radioactive decay could also cause spikes in the current. With just 2 spikes it is a little hard to tell.

The method of getting the derivative of the integrator voltage would be sensitive to some DNL errors of the ADC.

It looks a little like the readings are averaged over quite some time. So the fast side (e.f. be low some 10 or maybe even 100 seconds) of the Allan variance may just reflect the averaging. It does not make much sense to measure below aA in 0.1 second. Errors in the form of 2 large spikes is not that suitable for using the Allan variance.

[bsw_m]

The method of getting the derivative of the integrator voltage would be sensitive to some DNL errors of the ADC.

After integrator this instrument have differentiator and ADC after it. So, ADC DNL is not big trouble. But differentiator must have good linearity.

[TimFox]

The isolated spikes:  could they be decay alphas from materials or impurities (e.g. boron)?

[Kleinstein]

Are those step like smaller jumps (e.g. 0.8 hours in the 1st log) the errors from the auto zero ?
I would expect the Auto zero and also the integrator reset to use mechanical contacts of some kind. There is a chance to have some contact problems, especially if not used much for a long time. Even with mechanical switches there could be some kind of charge injection when opening the switch. Details one when and where the contact opens can make a difference and thus cause some instability.

Impurities / radioactive decay could be an issue, but why boron ? It is usually heavy isotopes that show alpha decay, like impurities in the solder or maybe radon gas.