Difference amplifier for off-line current shunt

[LukeW]

- Off-line 240VAC
- Current sensing at up to 15-20A max with a shunt resistor of say 5-10 milliohms.

- No Hall effect sensors or current transformers - too bulky, too expensive. Best if the solution is compact and cheap. Galvanic isolation from the mains line is not needed.

- At the moment, the current shunt is in series with the neutral line to the load, and the 5V SMPS output is relative to the mains neutral - that is, the ground from the SMPS is connected to the neutral on the SMPS input side, connected to the neutral of the mains supply. But keep in mind that the local DC ground is a self-contained thing - it's a floating, isolated system.

So, effectively, one side of the current shunt is ground and the other side of the current shunt is very close to ground - say 100mV RMS, if it is 10A RMS across 10 milliohms.

Anyway, any recommendations for a differential amplifier that might work well in this application, with a single-ended supply?
Is the AD8210 likely to be a reasonable choice, or are there any cheaper candidates?

[NiHaoMike]

Just use any plain opamp with AC coupling. Offset voltage would then be irrelevant.

[LukeW]

As a bit of a follow-up here, I have sketched up a couple of candidate circuits.

I'm not sure which is best, in terms of low cost, small size and component count, and a reasonable degree of precision.

I'm measuring current in a 240VAC 50Hz 10A system, with the grid neutral referenced to the local ground in the analog side of the system (before ADC and optical isolation), with the shunt resistors in the neutral line.

The LMP8601 has the smallest component count, but the 8-SOIC package is relatively large. The single-opamp approach is probably the cheapest, especially if a cheap opamp, eg. MCP6001 or something cheap and slow like that is used. But it adds a few extra cheap jellybeans such as resistors.

The ADC is referenced to 1024mV, and the signals are offset up by 512mV allowing the whole bidirectional sine wave to be captured. The gain figures are set up to clip hard to the 0/1024mV limits at about 16A of current, so there's a reasonable bit of headroom.

Any thoughts or comments? Which option do you think is best?

[LukeW]

Any takers?

[mij59]

Hi,

The INA170 is a high side current monitor, minimum common mode voltage is 2.7 V , so it will not function in your circuit.