# Circuit Design for Current Measurement at High Voltage

We're working on a power supply unit for ionic pumps. My supply voltage is 2kV to 3kV with current varying from 0.5mA to 1mA. We wanted to add a high side current measurement indication in these units-an analog voltage (like 5V for 0.5mA). First part of the circuit is sensing the current and in our research we came across the ACS product line, particularly speaking of ACS722KMATR-10AB datasheet attached here it offers Isolation for 4800 Vrms. For amplification we were thinking of using ADA4528, MAX4238 and AD8428. We want to know a few things:

A. Is the term Isolation Voltage in any way related to the Voltage Rating of the sense resistor of ACS722KMA? If not then what is the Voltage Rating of the sense resistor?

B. The Basic Isolation Working Voltage for ACS722KMA is stated 1097 Vrms in the datasheet, there is a little confusion if ACS722KMA be able to survive constant 2kVDC to 3kVDC ?

C. I came across another article, The schematic attached in this EEVBlog suggests another simpler method of measurement but unfortunately I wasn't able to find current sense resistors with High Voltage Rating.

How can I measure 1mA from a 2kV source?

• Possible duplicate of Current sensing of 3kV and higher Voltage Source May 12, 2017 at 11:44
• In part A, you mention the sense resistor--there is no sense resistor in this type of current sensor. It works by sensing the magnetic field that the current produces, not the voltage across a resistor. May 12, 2017 at 12:08
• You need to measure a few mA - the drift and errors from this device will dwarf your actual signal. May 12, 2017 at 12:13

For Ion pump current sensing, you don't need speed. So a voltage-to-frequency method is quite appropriate. A current-sense resistor (R1) easily converts current from the HV supply to a voltage. A microcontroller can convert an input voltage to an output (logic-level) frequency. Even LMC555 circuits can work, or 4046 voltage-controlled oscillator. From there, a high-standoff voltage opto-coupler translates logic-level frequency pulses down to ground.
The conversion from frequency back to voltage is easily done by a one-shot monostable, followed by an RC low-pass filter. Or if digital readout is desired, a simple frequency counter can work too, using a microcontroller again.
As suggested by Jeroen3, the optic-coupler can also carry serial UART-type data down to a receiving microcontroller, instead of frequency-proportional pulses. This would provide faster updates on the read-out display.

simulate this circuit – Schematic created using CircuitLab

Is the term Isolation Voltage in any way related to the Voltage Rating of the sense resistor of ACS722KMA?

A. Yes. The isolation barrier is inside the chip between the conductor and the hall effect sensor.
The voltage rating of the current channel is of no concern. Inside is 0.85 milli ohms. There will be no significant voltage.

The Basic Isolation Working Voltage for ACS722KMA is stated 1097 Vrms in the datasheet, there is a little confusion if ACS722KMA be able to survive constant 2kVDC to 3kVDC

B. It is designed for continuous operation at 1097 Vrms, and can survive shorts bursts of higher voltages up to 4800 Vrms. (respecting the V/us limits)
You're probably going to have a hard time finding 3kV SiO2 isolator chips.
Optocouplers are the only way.

I wasn't able to find current sense resistors with High Voltage Rating.

C. You don't need a high voltage rating for a shunt resistor. The 3 kV is common mode voltage for the shunt resistor. You do however need to pay attention to the inductance of the resistor, since that might temporarily raise the voltage.

If anyone would suggest any other component or technique

D. If it's only for indication you could perhaps use a digital method? A little micro sampling the shunt and sending UART frames over 1 optocoupler.
The more complicated way is to build a differential amplifier that allows for such high common mode. The concept as used in high voltage common mode probes.