INA210: Current measurement with µC

Hi Simon,

I will try to get our system's engineers to provide some insights on this and will get back to you. I have no experience on power electronics either. For applications like yours I usually see the customers uses the AMC devices.

Hello Mayrim

Thanks for the quick reply.
I am looking forward to their answer

Simon

Hi Simon,

One of our systems engineer have power experience so I am trying to get a hold of him to provide you some help/support here. He is currently out of the country on a business trip so his support will be delayed. 

In the meantime, what is your plan for dealing with the high Vcm Voltage? 

Hello Mayrim

Thanks for following up my case! I am currently using the AMC1200 in my prototype but I would like to get rid of the need for the separate power supply.

I am looking forward to the input of your colleague.

Kind regards

Simon

Hi Simon,

While we wait, I'd like to recommend a few reference designs that address current sensing with high-voltage level translation requirements such as yours.

1) TIDA-00332 demonstrates the use of a current-output amplifier to level-shift a current measurement from 400V to GND level. Note that this is a non-isolated solution and some power is wasted in the zener diode and supply current limiting resistor, however, the advantage of this solution is its simplicity and low component count.

2) TIDA-00313 is an isolated solution that uses a digitial-output current sense device (essentially a high precision ADC) with isolated outputs to send data to the uC from a -48V supply domain. This method provides a high resolution measurement but is usually only good for measuring low bandwidth signals. In other words, if you're interested in the average (DC) current through the inductor then this may not be a bad option to consider. 

3) TIPD196 is also an isolated current sensing solution and it uses the DRV421, a fluxgate sensor which needs an external coil to regulate the total magnetic field at its input. This will give you higher bandwidth for measuring dynamic currents as well as high accuracy, but it is more complex to implement than a regular current shunt solution due to the number for variables that you will have to control to get high accuracy.

Please have a look and let us know if any of these approaches interest you so we can help you further.

Best Regards,

Harsha Munikoti

Hello Harsha

Is your colleague already back? I was wondering if it is possible to use it if I place it in series with the mosfet instead of in series with the inductor?

Thanks for your elaborate answer, the design for 400V was really helpful.

Kind regards

Simon

Hello Harsha?
Are you still there?

Hello Simon,

I am the engineer that Mayrim was referring to in earlier posts, I have some background in DC/DC converters and may be able to help.

I have a few questions to help me get up to speed.

How important is the AC current signal to your control algorithm and what is your target switching frequency?  

Also do you have a min and max duty cycle you are targeting? 

Do you have a simplified schematic of your converter and where you would like to measure the current that you can share?

You can sense the current by adding a series resistor with the FET.  Ideally the resistor would be placed where the current sense amplifier does not see high common mode voltage transitions, i.e. not on the switching side of the converter.  If you place the resistor opposite the switch node you will need to be aware that the drop on the sense resistor will reduce your effective gate drive voltage so minimizing the value of this resistor is important.  

If you do plan to sense the inductor current the best place to do this would be on the output side of the inductor to avoid the switching transitions I mentioned above.

Regards,

Dennis

Hello Dennis

In fact, it's mainly the DC signal that interests me. I want to measure the current at the moment when it is equal to it's DC value, so in the middle of the on-time. Target frequency of the converter is 100-500 kHz. Max duty cycle is about 80%.

The schematic is quite straightforward. It's a boost converter going from 20-50V input to 80-120V output and I would put the shunt resistor underneath the switch such that the common mode voltage is almost zero. If I would place it in series with the inductor, the CM would probably be too high.

Kind regards
Simon

Simon,

Another thing to watch out for is the slew rate and settling time restrictions for your current sense amplifier.

The INA180 has a slew rate limitation of 2V/us.  Depending on your peak currents and gain of the amplifier you select, the output will not be able to slew faster than 2V/us.  This could be a problem if the device gain or current sense resistor you select is too large.

The maximum input current slew is calculated by:  Device output slew rate / GAIN / RSENSE.

Assuming your gain is 20 and your RSENSE is 2 mOhm the maximum rate of the input signal is 50A/us.  Most likely your inductor will have a current slew rate less than this, but if you used a higher gain with a higher RSENSE you could be slew rate limited.

The INA180 will also take about ~2.5us to settle to a differential input step.  I am afraid this may be too slow if you are operating the boost converter in continuous current mode at 500kHz.  Discontinuous mode might be ok since you will not have to settle due to the initial current step.

Regards,

Dennis

Great Dennis, thanks for your elaborate answer :)

I now know what to look out for!

Kind regards
Simon