INA303: Floating differential input

Hello Mario, 

Thank you for your post. 

The only way to successfully float the inputs to the device are to also float the supply. Even in this scenario, you need to measure the input voltage with respect to the GND pin of the device to ensure it is not violating abs max ratings. If violated, it will damage the device. 

Do you still have flexibility on choosing a device? I would suggest a device such as the INA206 data sheet, product information and support | TI.com or the INA310B data sheet, product information and support | TI.com.

I hope this helps, 

Joe

Hi Joe,

thank you for the reply.

First of all, I need to highlight the importance of using a bi-directional device. Your kind suggestions are both uni.

Secondly, I draw a little schematics of what I was considering to have in an application. It's drawn with mouse but I can redo it by hand if it is not clear. Just let me know.

What I'd like to have is not to connect the shunt with the device GND at all. As I understand, this works for differential amplifiers and I've done it in the past. Since I'm new to the INA set of devices, I'd like to find out if this works in that case also. There would be no connection with the shunt resistor ecept the differential input of the INA device.

What do you think about this?

Mario,

Thanks for your patience. Due to the U.S. holiday, we are currently out of the office. Joe will reply when we return on Monday, 12/1.

Ok, I'll wait for Joe ;)

40 V to 400 V Uni-directional Current/Voltage/Power Monitoring Reference Design

Could I use this in some manner?

Hello Mario,

If the load ground is not shared with the INA ground, then the measurement will be wrong. Also, in that case the voltage would be floating and could still damage the INA. 

Yes, you could use a method similar to that reference design. Here is another reference design that you may be interested in also: https://www.ti.com/tool/TIDA-01608. However, if you use a digital device, then we do have some that can support 72V or more: https://www.ti.com/product-category/amplifiers/current-sense/products.html#726max=76%3B120&-3=Digital%20power%20monitors%3BDigital%20power%20monitors%20with%20integrated%20shunt%20resistor&

Here is a list of analog out devices that are bidirectional and fit your voltage requirements: https://www.ti.com/product-category/amplifiers/current-sense/products.html#2192=Bidirectional&726max=76%3B120&-3=Analog%20current-sense%20amplifiers%3BAnalog%20current-sense%20amplifiers%20with%20integrated%20shunt%20resistor&

Our new integrated shunt devices are also bidirectional: https://www.ti.com/product-category/amplifiers/current-sense/analog-integrated-shunt-resistor/overview.html 

You may also be interested in our isolated amplifiers: https://www.ti.com/product-category/isolation/isolated-amplifiers/overview.html or our isolated ADCs: https://www.ti.com/product-category/isolation/isolated-adcs/overview.html 

Regards,

Mitch

Since my sub-circuit is isolated (several kV). What about referencing the shunt resistor (IN_N of INA303) to the same potential - like on this graph below:

Hello Mario,

That would work for the INA, but the output would be with respect to GND_ISO, so you may or may not have a problem measuring it or using it downstream.

Regards,

Mitch

Thx, this is good to know! :) Downstream I have an ADC (also referenced to GND_ISO) and a ISO652x isolator. So this part of the circuit is completely isolated from the rest of the board. I'll ask one more question and than we can call it a day. You are aware that my input signal is AC and you think that this solution with GND_ISO tie should work regarding the CM voltage levels?