4

I have a 10mohm 1% resistor that I want to use to measure current. Since I need multiple ranges I want to use multiple amplifiers with different gains (INA212-Q1, INA213A-Q1, INA214-Q1) with the resistor and connect each output to a separate ADC pin on an MCU so that I can avoid any sort of analog switching on the shunt side which could distort readings or interfere with the load.

Multiple INA21x-Q1 with a single shunt resistor

simulate this circuit – Schematic created using CircuitLab

My questions are:

  1. Is it sane to put multiple amplifiers on a shunt like this? And if not, which alternate scheme should I be looking at instead?
  2. How much will this affect the readings and the load?
Ignacio Vazquez-Abrams
  • 48,282
  • 4
  • 73
  • 102

2 Answers2

1

My questions are:

Is it sane to put multiple amplifiers on a shunt like this?

Sure, it is perhaps a bit expensive compared to changing the gain of one amplifier, but I see nothing else wrong with it.

And if not, which alternate scheme should I be looking at instead?

Changing the gain of the instrumentation amplifier would likely be more cost-effective. There are some with selectable gain or you could switch external resistor networks.

How much will this affect the readings and the load?

There should not be any noticeable effect on the reading from operating several amplifiers as shown provided you stay within the rails on all of the amplifiers. The source is extremely low input impedance, so effects from bias current, input noise current, and offset current will be way below the noise floor.

Spehro Pefhany
  • 376,485
  • 21
  • 320
  • 842
1

This is a perfectly sensible approach. What does the input of an amplifier push back onto what it's measuring: -

  • Voltage noise - yes
  • DC offset voltage - no this is an internal amplifier error
  • Bias and offset currents - yes, but with 0.01 ohms of shunt this can be ignored

As far as I can see it's only input voltage noise that can "pollute" the other amplifers and if they are all about the same noise level, the noise increases from N to: -

\$\sqrt{N^2 + N^2 + N^2}\$ = \$N\sqrt3\$ - this means it's 4.77dB higher.

If you can live with this, no problem.

EDIT - The use of the INA214A-Q1 - I think this is probably not needed because it has a gain (100V/V) that is only twice that of the INA213A-Q1 (50V/V). I would just use the two amplifiers, namely the 212-Q1 (1000V/V) and the 213A-Q1 (50V/V).

Andy aka
  • 434,556
  • 28
  • 351
  • 777