1 mA Accurate Current Clamp that is wide.

[horse_battery_staple_guy]

Would anyone recommend a 1mA accurate current clamp that is larger than 60mmx60mm?

I have one machine that due to the way it works, cannot be completely isolated from mains, as it has tubing and hoses that are connected to other parts of the building which it needs to operate.

I've verified that there's stray currents coming from the pipes and hoses by measuring voltage drops along them and disconnecting them. I also know the currents come from those external sources and not the machine by using only ground and disconnecting live/neutral.

I can't actually modify the external sources. So I'm stuck with those stray currents. What I thought about was measuring all leakages with timestamps and compensating to find out the real leakage current of the machine.

I know they introduce about ~20 to 50mA to the system. This value changes when the machine operates because the hoses and tubes change in temperature due to the fluids they're carrying. (I measured changing voltage drops with relation to temperature)

How could I measure 1mA to 1A with a decent accuracy ±1mA? It does not have to be stable across days because I can disconnect the sensors to zero them during the tests.

I was looking at some popular current clamps in the forum from UNI-T but those don't have the ability to be connected to a PC to log data. I also saw the UT251 which has the accuracy I need but its not big enough for the hoses and tubes. The UT253A is wide enough, but as far as I'm concerned, the error is too big for my use case.

I also thought of just monitoring the voltage drops with a datalogger or custom circuit but... the resistance changing makes it complicated for me right now.

[Terry Bites]

Fix the leakage current problem.
Even if the two machine areas have separate grounds, it should be still be possible to earth bond the two ends of the system.
In fact, its probably mandatory.

[horse_battery_staple_guy]

I wish I could, but I don't have access to those parts.

I did raise the issue with maintenance, but they told me it might take a while, so I'll be stuck with these conditions meanwhile  .

On the other hand, I was able to get the leakage current down to 2~5mA by bonding the hoses and tubes that come into the machine to earth. Hopefully I can still do something about it while maintenance fixes their end.

Some of the pumps and compressors seem to be improperly grounded.

[TimFox]

Back before 1980, an engineer in my university building was looking for a ground fault in the building's power distribution.
The distribution in question was a set of four rectangular bus bars (3 phases plus neutral) in an enclosed vertical shaft with access panels.
He made a home-brew current transformer, starting with laminae cannibalized from power transformers that had enough free space to encompass the four busses, carefully insulating the outside of the iron to avoid contact with the copper bars.
The tricky bit was to grind down the two contact areas on each of the "C" and "I" pieces to ensure intimate contact for the magnetic path.
Having built this unwieldy object, it was straightforward to calibrate it by passing a known AC current through the opening in a wire.
He then used it to look for a net current flow in the four busses, which represented a parasitic current to ground (avoiding the busses).
If I remember correctly, the problem turned out to be a lamp over a drafting table that had a ground fault.

[coromonadalix]

60mm x 60mm  for 1ma current sense ?   sorry   but your read values will be wothless,  you need to be close as possible to the wire .. not a pipe or a shielded wire ?

you play with an needle in a .....


from what you're saying,  bring a master electrician  and find your leakages, verify all your grounds, socket phases, line / neutral    etc ....

and not having any kind of access  is a no go,  you make one

and sadly  you are mysterious of your equipment(s) ... that doesn't help anyone here, and you may have some static issues, or charge(s) accumulation(s)

[horse_battery_staple_guy]

60mm x 60mm  for 1ma current sense ?   sorry   but your read values will be wothless,  you need to be close as possible to the wire .. not a pipe or a shielded wire ?

you play with an needle in a .....


from what you're saying,  bring a master electrician  and find your leakages, verify all your grounds, socket phases, line / neutral    etc ....

and not having any kind of access  is a no go,  you make one

and sadly  you are mysterious of your equipment(s) ... that doesn't help anyone here, and you may have some static issues, or charge(s) accumulation(s)

Is it impossible to measure small currents with those kind of dimensions? How come?

I actually know the source of the leakage currents now though. I asked for permission to at least turn off some of the breakers for well known device, machinery, etc. Found it was the extractors of the entire area and the lights. 40 to 20mA in the extractors and 2mA in the ceiling lights. I have no access to either of those, but now its much easier for maintenance to fix that.

I honestly figured the machine itself was not relevant. The most important part is that it uses a GFCI, is made of metal and is pressurized and connected to extractors. I didn't really want to go into specifics.
But essentially, the extractors are there to .. extract fumes from the machine, and the hoses are pressurized with water, and those hoses are braided internally. (They're 6000 psi rated, and they happen to be conductive because the crimps bite into the hoses). It's also connected to tubing carrying pressurized air (100 psi) but I managed to find it is not carrying any leakage current. I know i'm being silly with being really vague about it.

[jonpaul]

study or FEM model the conductor DUT or busbar fields.

Add the CT core.

Center the CT on the conductor

Noe deflect the CT a few mm from center.

Issue will be obvious...fringing flux and leakage flux will make the CT reading very inaccurate.

Simple test;

Run a fixed resistve load eg heater or incandescent lamp on a fixed mains V, eg 1 kW @ 120V should be 8.33 A

Take electricians clamp on ammeter, observe the current on a 10..20A range.

Now move the conductor in the CT window.

Easily 10 % variations

Jon