Post Go back to editing

# How to sense current difference using a precision ADC?

Category: Hardware

Hello, we have a position detector that generates Ia, Ib, and Ignd current outputs, where the common mode of Ia and Ib is 120 uA and the difference ranges from -100 uA to +100 uA. We require a small form factor and low bill of materials solutions to convert the current differences to digital bits, ideally ranging from 16 to 20 bits, while also monitoring the common mode on a test point.

1. We are using OP27 as TIA amplifiers and would like to amplify 120 uA to Vref/2, could we connect the outputs of TIA amplifier directly to the RC SH circuit of AD4003? Is there a better alternative to OP27?

2. Does ADI offer any solutions for measuring and converting current differences (in the scale of uA) to high-precision bits, while also meeting our requirements for small size and low bill of materials?

In short, can we sense a differential current input using a fully differential amplifier and direct fed to AD4003?

Best regards,

Benyuan

Sum up the questions we had.
[edited by: liubenyuan at 4:44 AM (GMT -4) on 10 Apr 2023]
Parents
• Hi Liubenyuan:  The OP27 is a fantastic opamp, but I think the bias current is too high to support high resolution of 100uA signals.  20 bits is 1ppm, which compared to 100uA is 100pA.  For the purposes of precision, I would suggest doing the I-to-V conversions single ended, and then using inverting stages to achieve differential ADC drives.

But yes, if you can find a fully differential opamp with adequate precision, and low bias current, then you should be able to go directly TIA.  I showed a sample of such a circuit in DN473, here:  https://www.analog.com/media/en/reference-design-documentation/design-notes/dn473.pdf  The performance is not ideal on the un-fed-back output.

• Thank you, I will use ADA4610 to convert the Ia, Ib to Va, Vb.

The common mode of Ia and Ib is 120uA +- 10%, and the difference of Ia-Ib is in the range of [-100uA, +100uA]. I would like to use an differential ADC to sample the Va-Vb after two TIAs.

My question is, could I simplify the ADC buffer circuit by using two amplifiers directly, that amplify 120uA to 2.048V reference voltage?

Do I need a differential ADC driver (for example, LTC6363) that placed between Va, Vb and ADC (differential input)?