AC Current Measurement

[JKKDev]

Hey guys,

I've looked over the search results concerning the title but I can't find what I'm looking for. And as I feel it's a basic thing it should be asked in this section.

I'm building a computer-controlled dual-motor switch. Motors run on 240 VAC (expected current is around 1A however I haven't received them yet so I wasn't able to measure it. Let's assume it's under 5A). I will have a relay module with some logic to turn either of the two motors on (never both as they will be driving the same thing). I need to be able to monitor the current consumption of the input into the relay module (ac source -> current monitor -> relay module -> 2x motor).

This is my first time working with AC so excuse my ignorance. The solution that feels the most reasonable to me is a current transformer, however from what I've read it is used for high current loads (several hundreds of amps). The bonus I see in it is that it offers galvanic isolation.

The other common solution is a shunt resistor. Here I'm not sure how to go about designing the circuitry around it. From what I've read I should be doing high side measurement which means any component used will be exposed to the 240VAC potential (is this right?). Can I get some pointers on how to go about using shunt resistor in this setting and what kind of circuitry to build around it?

Any kind of insight into designing current measuring circuitry for 240 VAC inductive loads would be very welcome (links to guides, even books on the subject).

Thank you and have a nice day

[capt bullshot]

One can buy a CT rated for e.g. 5A:

https://www.mouser.de/ProductDetail/Triad-Magnetics/CST-1005?qs=U28YatCw3SIobGmak15BMA%3D%3D

It's not the cheapest solution, but by far the easiest and more or less intrinsically safest method. Safe in the way that you'd put an isolated wire through the CT, and your circuit is automatically isolated from mains.

[Kleinstein]

A current transformer is also working for smaller currents. So the 1 A range should not be a problem with a smaller core. If needed one could also use a few turns on the primary to increase the signal, but there should be no need at 1 A.

For a low accuracy requirement one can use some of the current compensated chokes (common mode chokes) as a CT with an additional wire through the center.

The isolation with a CT is definitely an advantage.

[ArthurDent]

There are lots of cheap current transformer/ammeter combinations on eBay if you want to experiment with them to see how they would work in your application. Do a search for: “current transformer ammeter” on eBay to see a list of auctions for these.

[JKKDev]

Thank you guys

I've decided that CT is definately the way to go.

The common mode choke intrigues me (for prototyping purposes not the final product). Would you connect the two windings together or use just one? I guess it depends on the wanted output current. Maybe even rewind it to achieve the exact wanted range. How would I go about sizing the burden resistor? I will research this on my own.. just a few questions that came to mind.

[ArthurDent]

Keep in mind that a current transformer must never be operated without the correct secondary load, or actually shorted. High voltages can be present if there isn't a load on a CT. Using a commercially made CT will make choosing the burden resistor simpler and you'll be sure the transformer is rated properly.

[JKKDev]

By that you mean never run it without correctly sized burden resistor right? I did a test with the above mentioned choke (turn ratio 30, max. current 10A (live wire),  burden resistor 15 ohm (sized to output 5V at max current)) and nothing worked :p

I guess I'll have to wait for a real CT (I chose another one on Mouser that's max. 3A but it has both primary and sens thru hole so you can run both on traces and there's no need for cables). After a bit of research on the type of motor the current numbers are all under 1.5A so as long as the motor doesn't stall this one should be fine (maybe a 3A fuse  wouldn't be such a bad idea either).

[capt bullshot]

By that you mean never run it without correctly sized burden resistor right? I did a test with the above mentioned choke (turn ratio 30, max. current 10A (live wire),  burden resistor 15 ohm (sized to output 5V at max current)) and nothing worked :p

For a 1:30 turns ratio CT on a typical common mode choke's core, one needs a near short circuit as a termination. This is because the core saturates quite quickly, and secondary voltage (across the burden resistor) must be well below a few 100mV for this to work. A typical 1000:1 core like the one from mouser should work maybe up to one or so volts over the burden resistor - a recommended value (and maximum too) should be stated in the datasheet. Don't expect 5V output from an CT.

Be careful when ordering a CT to look at the specified frequency range. CTs exist that are intended as current sensors in SMPS, they simply won't work at line frequency.

[JKKDev]

I'm aware of the frequency requirements as well the fact that I'll need an amplification stage after the CT so I'll be able to read the output with a microcontroller. Should I convert AC to DC and then amplify (AFAIK I'd need a negative supply to amplify AC and I don't need that for anything else) or is that a bad idea for some reason?

The CT I decided to go with for testing is this one: https://eu.mouser.com/ProductDetail/Triad-Magnetics/CSE185L?qs=HSwFsqQPTVMv4iimN5cHkA%3D%3D

On a side note: What the hey... you can file a patent on a couple of components (amplifier circuit in this case)? (source https://patents.google.com/patent/US8212570B1/en)

[Kleinstein]

For measuring the AC voltage with a µC, it depends.  It is usually better to first amplify the AC signal. If the µC has time and the ADC is fast enough one can sample the AC signal with the ADC and calculate the RMS value in SW. This works surprisingly good, even with an ADC of moderate resolution (e.g. 10 Bit).  Converting AC to DC with some extra analog HW is also possible and needs less time and SW at the µC, but more HW. For the AC/DC conversion there are precision rectifiers with 1 or 2 OPs.

For the CT one can use 2 anti parallel diodes for protection, that even with over-current there would be no excessive voltage.
For a small current (e.g. 1:1000 CT) one could use a trans-impedance amplifier for the current signal and this way minimize the burden seen by the CT. This way even the common mode choke may work (for small current).

[rstofer]

As mentioned above, be careful with the secondary - Ohm's Law is still a LAW, not a suggestion like some speed limit.

E = I * R

Suppose the CT wants to run 1 Amp through the secondary (I = 1) and R is nearly an open circuit, say 1M Ohm.  Well, according to Ohm's Law, E will go to 1 MV.  Sure, I picked outrageous numbers but if R is very high (open circuit), so is E.  It's Ohm's LAW.

Things go sideways at very high voltages!

[David Hess]

Amplifying AC has the advantage that AC coupling can be used to remove DC offset.

The current transformer can be used with a higher value load resistor for better sensitivity if the current is only part of the maximum current.  Current transformers are rated in volt-seconds meaning that they will support up to a specific voltage output at any current level once the secondary resistance is accounted for because it adds to the output voltage also.