How to design an amp to sense weak current (1uA~2uA)?

[jordan76]

how to measure 1ua current

Hi

For week current (several uA), how to design a sense amp to sense it for low voltage(1.6V)? what about below 1uA range?

Any input is welcome. Thanks!

regards,
jordan76

 

[v_naren]

since your input is a current signal and it is so very low
I think the amplfier you need will be a current to voltage amplfier type..I mean a transresistance amplifier.

So you can use a common gate or a common base amplifier to present a very low input impedance and at the same time obtain a high transresistance gain.

Then following the Common gate or common Base stage you can use a voltage amplifier

or u can also use an opamp like 741 to implement the current to voltage amplifier...in this case you need to find out the highest frequency occuring in your current signal?

also I am using you are talking about a AC signal right?

 

[tucura]

To sense current you need a current reference comparable to
the current you want to measure. If > 1uA any current source
can be used. < 1uA leakage complicates matters. You can amplify
the current but you have to bias the sources of the mirror pair
to reduce leakage. As a current to voltage converter a capacitor
is best (no leakage) but requires a reset cycle.

 

[jordan76]

Hi

First of all, thanks for your reply!

v_naren,

If I use the common gate amp to obtain the current to voltage conversion, maybe the time needed is too huge since the current signal frequency could be around 10MHZ~20MHZ. How do you think about it?

tucura,

If I use a reference current source, how could I design such a weak accurate reference current source considering the process,voltage and temperature variation?

Any comments or opinions? Or even some detailed circuit structures?

regards,
jordan76

 

[v_naren]

are u planning to design this whole thing on chip or in discrete form?
If it is going to be on chip which process are you going to use?

I dont think it is a big problem to amplify 20MHz!!!

The current mirror method of amplification is also very good....but what about the voltage across your 1uA surce will it always stay above 0? and there are other problems too.....please tell if u are going to this discrete or on chip?

 

[jordan76]

v_naren,

Sure,it is on chip. And the process is 0.25um technology.

Also could you suggest some detailed circuit structure or reference?

Thanks!

regards,
jordan76

 

[v_naren]

ok..here are the problems....is your current source floating type?
Can you connect one of your terminals to ground or a + voltage?

is your current source really a current source??

I mean what is the resistance of your weak current source?

 

[jordan76]

v_naren,

Please allow me make it clearer:
The weak current(around 1uA) is generated by a weak NMOS whose gate is VCC,source is GND. Also the drain has large capacitance due to the multiple dummy load and the long metal wire.That is why I worry about the speed issue of common gate amp as I-V converter.

BTW,we can precharge the drain node to VCC if needed.

regards,
jordan76

 

[hanjiemy]

I think these transistor work@subthreshold region.

but i am not sure.

 

[ynhe]

Hi
Maybe TI's opa380 is suitable for your application.

1.GAIN BANDWIDTH: 90MHz
2.> 1MHz TRANSIMPEDANCE BANDWIDTH
3.INPUT CURRENT RANGE: 10nA to 1mA
4.SUPPLY RANGE: 2.7V to 5.5V
5.auto-zero topology


ynhe

 

[jordan76]

hanjiemy,
Thanks.

ynhe,
What I need to design is an IC,not a PCB board. Thanks anyway.

regards,
jordan76