LMC6001: How to amplify pA current

Hi Kevin,

I would use a TIA (transimpedance amplifier) for this. This is just an inverting amplifier with a high feedback resistance:

en.wikipedia.org/.../Transimpedance_amplifier

What is the source of the measuring current?

Kai

Hi Kevin,

Here are a few more resources on how to design transimpedance amplifiers.

Transimpedance Amplifier Cookbook Circuit

What op amp bandwidth do I need? Transimpedance Amplifiers

1MHz, Single-Supply, Photodiode Amplifier Reference Design

For low level current measurements you must make sure that the input bias current of the op amp is less than the current you are trying to measure, otherwise the input bias current of the op amp will dominate the measurement causing high error. I recommend watching out TI Precision Lab videos on Offset Voltage and Input Bias Current.

You may also need to add a second non-inverting gain stage to the output of the TIA to get the output swing you need.

Thank you,

Tim Claycomb

Kevin,

The circuit as you have drawn it has two stages of open loop amplifier gain - will never work as the offset voltages will push both into the supply rail. you will need to close the loop with a first stage feedback R to the inverting input as high as possible that will still meet the desired BW. Then 2nd stage voltage gain to finish target gain. Will need to know the source capacitance to even get started.

Hi Michael,

I have updated the schematic as attachment. Could you please help review it?

Thanks.

Kevin Xiong

Please help review my updated schematic. Thanks.

Still not right,

1. what is the estimated current level you are trying to sense,
2. what is its capacitance + any cabling capacitance
3. what is desired bandwidth

Hi Kevin,

I don't see a TIA...

Kai

What's the TIA? Thanks.

Hi Kevin,

we recommended you to use a TIA. You will probably need such a scheme:

The left OPAmp is wired as a TIA (transimpedance amplifer). "Transimpedance" just means that the measuring current is converted into a voltage.

Kai

Hi Michael,

1. what is the estimated current level you are trying to sense,
The current level is fA ~nA

2. what is its capacitance + any cabling capacitance
Designing, no specific value

3. what is desired bandwidth
The BW is 100Hz

So could you please give me a specific reference schematic? I'd appreciate your help first!

Thanks.
Kevin Xiong

Do you need DC coupled here?

Do you have a specification sheet on what you are using to generate the current signal?

I will probably be looking to use the LMP7721, testing max 3fA at room temp input bias current.

here you go Kevin, something to get started with,

1. LMP7721 1st stage - it has a brand new 2019 TINA model that internally appears to have 15pF Ccm + 5pF Cdiff - that will be the source capacitance for now,

2. As you increase the feedback R to get the transimpedance gain - you push the required feedback C down  to difficult low levels - I went for 2Mohm but then what is called a T network in the feedback to get another gain of 11 in the first stage transimpedance gain

3. This is still quite high BW by itself, so I added a gain of 10 200Hz Active filter to control the noise power bandwidth - Used the OPA376 there,

4. there is an Output DC offset issue here - if you can be AC coupled add a blocking cap into the first filter R

5. Input referred current noise is quite low, but hopefully you can average anyway

Again, just something to get started with, sure there are plenty of modifications you will want to make.

LMP7721 Zt stage plus gain of 10 post filter.TSC

Also, I know R values in the active filter stage will seem a little high - no impact on noise with all that gain in the first stage - my MFB design flow limits the maximum C values to those that available at low cost in C0G dielectric - those get more expensive as you get over 33nF or so. So yes using uF kind of values in the 200Hz LP filter would push the R values way down, but your filter shape will have a lot more tempco moving to X7R vs C0G dielectrics. May not matter in your application.