Great Difference Bias Current AD8629/AD8629S

Hello there analog community,

 

We are developing  payloads for space missions and we have some important doubts regarding part AD8629/AD8629S

which we need to be resolved to ensure the functionality of our designs.

 

The main doubt is related with the space model AD8629S in which the Bias Current seems to be much bigger than those in the commercial model.

 

In the commercial datasheet the bias currents are 30pA (typical) to 100pA (max) and shows a graph (Input Bias Current vs Input Common-Mode Voltage for several temperatures) which imply that the bias current could be 10pA for Vcm = 0.5V (e.g) and 25ºC (and much less for even lower temperatures! which is good for us).

The space part datasheet (http://www.analog.com/media/en/technical-documentation/data-sheets/AD8629S.pdf ) tells bias current limits are -500pA (min) to 500pA (max) but it doesn’t say which are the typical values or at least

the variation of this bias current VS Temperature (similar to those in the commercial datasheet).

 

So, why is such a change between bias current values from space part AD8629S to commercial part AD8629? What are the typical values for this parameter in the space model?

Some Op Amps whose commercial part has pA as Ibias says to not use a compensating resistance in a transimpedance configuration because the bias current is already compensated internally (so adding this R will do things worse). Can we use a compensating resistance to cancel the effect of Ibias in AD8629S?

 

I hope you can shed some light on this... Thank you very much in advance.

JavierM

  • Thank you very much Harryh and sorry for the delay in my answer.

    You are right, it is better to not rely on the typicals and I'll pay extra attention to this in the future.

    Anyway there weren't so many operational amplifiers that meet our requirements (Single Supply, Rail to Rail, RadHard, Low Bias ~pA, maybe dual).The day we look for them the AD8629 appears as an option and still is, of course, we have to do some modifications to improve the sensor signals and maybe do some characterisation to delimit the initial bias current.

    Thanks for the info about the autozero parts, good to know. Regarding your question, I assume that compensating resistance in combination with the 8pF input capacitance will create a pole and make the system unstable, right? In the case of a transimpedance amplifier (which is actually the configuration in our design) this can be solved with a feedback capacitor, but, as you said before, the bias cannot be compensated in this case.

    Thanks again, and let me know if I have miss something