measure high DC voltage and current

[cio74]

Hello,

I am wondering what could be a good method to measure large voltage and currents, say 1kV/300A DC power bar.

I think for voltage measurement a basic voltage divider made out of high voltage rated resistors will be fine (or few of low voltage in series), fallowed by an op amp and feed it to the ADC of a basic 3.3v MCU.

About measuring 300A, what would be a good implementation?

Thanks.

Cio

[Gyro]

There's an awful lot more to it than a few resistors and a micro when you get to 300A at 1kV, particularly DC. You are into instantly lethal voltage and arc-flash territory.

You need a properly constructed, protected and insulated voltage divider, if it flashes over very nasty things will happen. For the Current, an appropriate insulation rated DC current clamp on the bus bar is probably the way to go. A current shunt would pose many problems at such an elevated voltage.

I'm worried that, from the wording of your question and your posting history, you don't have the knowledge and experience to be doing this - please forgive me if I am wrong.

[cio74]

Of course I will not build anything like that on my own, clearly I have no such skills nor experience.

However, I am interested in how it can be done properly?

Thanks.

 

[tunk]

For current, maybe a DC clamp meter with bluetooth?

[Gyro]

Of course I will not build anything like that on my own, clearly I have no such skills nor experience.

However, I am interested in how it can be done properly?

Thanks.

Ah, Ok, I understand.

The biggest problem would probably actually be the voltage measurement - i.e. How would you get the voltage connection from the bus bar to your board, you can't just tack on a wire. In practice you would end up with an external purpose built (probably off the shelf) voltage divider, that would have suitably rugged bolted connection to the bus bar and ground / earth reference. Mechanical integrity and insulation are everything - losing the ground connection for instance could be catastrophic, as would an insulation breakdown. From the divider, you would then get a 'safe' voltage to go into your measurement gear. You wouldn't put a discrete resistor chain on your PCB.

Bus bar current measurement is quite a common thing to do, this means that there would be dedicated industrial current clamp devices off the shelf. DC measurement is less common than AC (simple current transformer), so would involve a Hall-effect sensor type. The main requirements are bus bar aperture, maximum current and insulation rating. You'd inevitably get some leakage current through the clamp's insulation too, that would need to be directed to ground.

It's really a matter of scale. A very low current 1kV supply isn't particularly dangerous. Measuring 300A at low voltage is more dangerous as it can cause thing to get hot and burn you. Put the two together (and add whatever the prospective fault current might be) and you get no second chances, horrific burns, explosions etc. Everything has to be 'industrial' (and expensive) in it's construction and ruggedness.

You would almost certainly still need galvanic isolation on the micro's data output too. 300A flowing through the load to ground could well cause the measurement ground to be at at a different potential to whatever you are communicating with - you definitely wouldn't want to create a ground loop through your test gear.

[Gibson486]

It was feasible until you mentioned 300Amps. With that current, you are better off with a clamp meter or something....

[coromonadalix]

current clamp meter with an high voltage probe

[CaptDon]

In locomotives we use VAMs to measure voltage which is a voltage divider insulated and tested to about 5 times the working voltage and LEMs which are like a dedicated clamp meter. They have three terminals +24, Return and signal output. They do not open up like a clamp but you have to pass your high current 'any voltage' cable through them. They have about a 1.5" square opening and since they don't 'unclamp' they are always accurate. The voltage limit is determined by the insulation of the wire you are passing through them. Locomotives run about 1400vdc for the link voltage and several thousand amps. Do the math, 4400 electrical horsepower at 1400vdc. The alternators have multiple series and parallel windings to get the voltage and current. RPM is usually 1050 max and with modifications to the voltage regulator circuit cards we can generate 1280vac 3 phase 60hz at 900rpm which can be cabled directly into many hospital and industrial substations.

[penfold]

In the context of those voltages and currents, its quite common to use a voltage transducer (see https://www.lem.com/en/voltage-transducers for examples and notes about how LEM brand are put together). As mentioned by other posters, there is a lot involved both electrically and mechanically to avoid flashover, tracking, arcing etc, a lot of those measures are apparent from the surface of the LEM voltage transducers' with things like the standoff distances and surface rippling to reduce surface tracking.

[cio74]

Thank you guys, much appreciated.