Other Parts Discussed in Thread: OPA564, INA181
TI,
I'm currently exploring use of the INA170 to measure the bidirectional current on the output of an Improved Howland Current Pump circuit that has a ground referenced load.
As you know, this means that the output voltage of the pump can swing both positive and negative (lets assume the OPA564 for this example, because it has a great example circuit in the datasheet on pg 25 Fig 51, but this use could be any HCP configuration with a sense resistor)
If the example in Fig 51 is used then:
V+ = +5V and
V- = -5V
That means that in order for the INA170 to function across the entire common-mode range of the HCP, then the GND pin of the INA170 should be connected to negative V- (-5V) rail.
I'd also expect that the Ros and RL to be referenced to the negative rail as well. Using the example in Figure 51, Rs = 0.050 ohm and Is = 1.5A .
Now, the issue I see is that based on the datasheet, then 1V < Vref < 4V above the reference ground based on the INA170 datasheet. I elected to use the Vdig reference from the OPA564 datasheet for convenience, assuming a Vdig that is 3.3V above the V- rail. This should be a valid reference voltage relative to the negative rail.
based on that, I would anticipate that this solution *should* function as normal with an Vout that is negative.
However, I need to still get back to a ground-referenced output. My current thought is to use another amp like a OPAx156 to shift the output again by subtracting out the value of the negative rail. i.e [ Vsense = Vout - (V-) ]. Is there is a better way to do this, by extending the range of the reference Vref of the INA170 or something to that effect?
I had been experimenting in TI-TINA to see what would result from feeding the Vout back into Vref or the Ros pins but this hast been at all fruitful.
Any assistance is appreciated.
Thanks.