ADS131A04: Current sensing using high precision analog-to-digital converter

Hi Ron,

Welcome to our e2e Forum! The ADS131A04 would have the speed and simultaneous sampling that you are looking for. Your 6mV input is very low, even with the max gain of 16 in the ADS131A04, you won't be using much of the input range. Is the 6mV level for the 9000A? Would that mean the 2000A level is ~1.3mV? What sort of shunt are you using?

Hello Tom,

thank you very much for your answer. We already purchased the ADS131A04 to give it a try, but we still have problems to get an accurate measurement. This means that there seems to be an offset, which is not constant.

Yes, without looking at the temperatures, one can say that 2000A ~ 1mV. Additionally we have other measurement points, where this relationship is even "worse".

Our "shunt" is the busbar itself. It is a conductor with a rectangular crosssection of 160mm x 90mm and a length of  and we're measuring the voltage drop over a defined distance of 200mm. Unfortunately this distance cannot be increased. We just thought, we could digitalize our measurement signal with such a high resolution analog-to-digital converter without amplifying it before. But maybe this is the wrong way doing it.

Best regards,

Ron

Hi Ron,

Take a look at our AMC1301, it's an isolation amplifier with +/-250mV input range.  Not sure what voltage levels you are looking at or if you need isolation (I suspect you might), but I wanted to bring that device to your attention because we are going to be releasing a +/-50mV version of the device later this year which might help with your application.  The output is a differential voltage which you can apply directly to the input of an ADC such as the ADS101A04.

Hi Ron,

Are you making any progress here? Another device you might look into is the DRV425 - we have an application note describing the use of this device in (essentially) a differential configuration for use in bus bars. You can find more detail here: www.ti.com/.../technicaldocuments

Hello Tom,

thank you very much for your two answers. I've been on vacation, that's why this response is a bit late. I am sorry for that.

I had a look at your AMC1301 isolation amplifier. Your assumption that we might need an isolation amplifier is right, because
in the worst case we can expect 3,5kV to ground. But since we're using only battery powered measurement equipment for prototyping,
there is no need for an isolation amplifier at the moment. Furthermore AMC1301 has a voltage offset of 200 µV =0.2 mV which
would be higher than our desired resolution of at least 0,1mV. But I it should be possible to "subtract out" this high offset, right?
Maybe using a two stage amplifier configuration would be helpful, containing a high precision amplifier followed by an isolation
amplifier.

I also had a look at your DRV425. But if I understand this correctly one needs to drill a hole in the middle of the busbar. Unfortunately
it's not possible for us, because our busbars are changed every 30 days. Additionally we have neighbour busbars, that might influence
the magnetic field that is measured by the sensor configuration.

Because we have multiple measurement sites, we also thought about  using a multiplexer, so that only one amplifier-adc-combination is needed for multiple measurement sites instead of using individual amplifiers for each measurement site. Is there any high precision multiplexer (for a test, 4 input channels would be sufficient) in your range of products?

Thank you so much for your efforts!

Best regards,

Ron

Hi Ron,

Yes, the offset from the AMC1301 could be calibrated. Skin effect from the AC component and (I suspect) some self heating of the bus bars themselves might be the reason behind your initial offset and instability. If it's alright with you, I'll send you a note to your my.ti registered e-mail account and we can take this off line for the moment.

Hello Tom,

thank you very much for your answer and efforts. Until now, I didn't receive any note from you via my my.ti registered e-mail account.

But I had one more look at your portfolio and I found this one: www.ti.com/.../ina233.pdf

For me this looks like a very nice solution.

What Do you think about that?

Thank you very much.

Best regards,

Ron

Hi Ron - I sent you a note. I think the INA233 would be a good option to try. Let us know how you make out.