Precision DC amplifier

[Valejan]

I need very precision PGA amplifier for 24bit ADC.
Input voltage 0-10V, gain 1x 10x 100x, input current < 10pA
I tried the AD8610 witch swichted resistor, but it would need something better
Can you advise?

I take a construction with multiple IC

 

[FvM]

Which amplifier parameters aren't satifying? In which regard do you expect better performance with a "multiple IC" design?

 

[Valejan]

Uncomfortable is
- +-13V power supply, I have +- 16V for other amplifier
- Input Bias Current datasheet say 10pA on 25C, I measure 16pA on 22C, hundreds of fA up to 1 pA would be better
input ofdset and noise id on the edge.

 

[schmitt trigger]

How are you measuring 16 pA?
Board (FR4) leakage itself can be higher unless special precautions are taken, like guard rings and/or teflon inserts.

Anyways, the OPA129 has an Ibb of 0.1 pA, and can operate to +/- 18V.

 

[FvM]

AD8610 is specified with typical Ib of 2 pA, maximum 10 pA. Your measurements suggest an unsuitable PCB layout or dirty surface. Very clearly, you won't achieve 10 pA without well designed guard traces. To preserve it despite of ambient humidity, protective coating is strongly suggested.

I also doubt that you'll get lower bias current without paying for it in noise and offset currency.

 

[Valejan]

How, simple 500M ohm resistor + 7and 1/2 nanovolt mener.
Measure can be also flows around 10fA.
Details such as here.
https://www.analog.com/static/imported-files/tutorials/MT-038.pdf
Measured in air without PCB.
On 16pA measured +-2pA takes, measured piece had such input current and is in line with datashet see
−40°C < TA < +85°C −250 +130 +250 pA
Truth can not say what it was exactly when measuring temperature, but the beginning of September to 53 latitude, solar day 20- 22C , humidity 40%

The final designe, while at the FR4 but plan protective ring into which the injected voltage prevent leakage current

 

[pinout]

In days gone by I used to use the Teledyne Philbrick 1701 or 1702, these were femtoamp bias current devices and truely amazing, but now only found in museums! Look them up.

The key to operating in these areas is as mentioned teflon insulators and also 'air gaps' (use of rigid flying wire) and importantly 'board cleanliness'.
Our procedure was deflux with commercial solvent, wash with distilled water (lab grade), clean with lab grade alcohol and dry with nitrogen gas jet. (probably repeated several times).
With this method we had electrometer circuits that when you unplugged the electrode, the analogue reading remained stable for hours, probably days!

In your pA region, you should be able to get excellent results with the guard rings and good cleaning steps. Get rid of flux residues, that is cruical or you may design a hygrometer!
Then I suggest cleaning with IPA and a clean brush a few times as a more normal workshop routine. Lots of people don't understand 'cleaning', disolving the dirt and grime and then leaving it to dry just spreads it around, you really do need to wash it away.

If you do venture into fA you can put a guard ring around the inputs but have the hole as a big slot, i.e. produce a large TPH with the leads floating in space and 'air wire' to the pins.

 

[carpenter]

Teledyne Philbrick 1701 is perfect piece incorrupt old hardware.

**broken link removed**