Page 1 of 1

SCT-013-030

Posted: Wed Apr 28, 2021 4:46 pm
by TimWardCam
Well, I think that's what I've bought, what it actually says on it is "SCT013 30A/1V".

So, I think this is a "voltage output" version, which I connect straight to an ADC chip without having to worry about adding my own load resistor, as I read is necessary for some other models of SCT-013, is that correct?

The thing that confuses me is that the data sheets I've found say that the internal load resistor is 62Ω, whereas according to my multi-meter what I've actually got is around 35Ω. Maybe I should ignore this discrepancy, and just believe that with 30A input you get 1V output?

Would I be right in thinking that that's RMS in both cases?

So to connect to the ADC I want a voltage divider to shift the voltage up a bit so it's always above zero, and then I measure the instantaneous voltage, take lots of measurements fast, and hope I capture the peak, and then do some sums to remove the bias and convert peak to RMS?

I'll be measuring mains current at 50Hz up to 10A.

Thanks for any advice from anyone who has used one of these.

Re: SCT-013-030

Posted: Wed Apr 28, 2021 7:13 pm
by jayben
Your test meter is probably showing the combined resistance of the transformer coil and the load resistor.

Since the transformer output is isolated & floating, you can connect one side to a potential divider that divides the supply voltage by 2, and the other side to your ADC input.

I don't think the 'is it RMS' question is meaningful; the transformer just defines a relationship between the instantaneous current in, and the instantaneous voltage out. So if you use an RMS meter to measure the mains current, and an RMS meter to measure the output voltage, they should match the given ratio.

Re: SCT-013-030

Posted: Sat May 08, 2021 9:42 am
by TimWardCam
OK, so it does work pretty much like that.

In fact for me it worked a lot better to use the differential capability of the MCP3008 and compare the two sides of the SCT-013-030 against each other - this reduced the spurious current measured when there was actually none at all by a factor of 150. One really shouldn't use a digital supply voltage (or indeed any supply voltage) as an analogue reference voltage!!! - and I decided that hooking up one more wire to use the MCP3008 in differential mode was less trouble than trying to produce a conditioned analogue reference voltage.

And then the power I was calculating from the voltage measurement was out by a factor of around 1.75, when comparing the measured current against the rating plate of a 2kw fan heater and the data sheets I've found. I've no idea what that was about - could be something in my RMS arithmetic, but I don't think so - but putting in the fudge factor is good enough for my purposes. (I'm calculating RMS from the samples, not trying to spot the peak.)

Using
import adafruit_mcp3xxx.mcp3008 as MCP
from adafruit_mcp3xxx.analog_in import AnalogIn
I'm getting around 1,200 samples per second, which is around 12 per half cycle. Which again is good enough for my purposes - repeated measurements over one second come out consistent enough for me. I'm vaguely aware that there are alternative libraries that are claimed to go lots faster, but as what I've got is good enough I don't need to worry about doing any better.

Re: SCT-013-030

Posted: Sun May 09, 2021 6:54 pm
by TimWardCam
And, amusingly, when I move the kit from the lab (my office) to the field (the greenhouse) when there's no load the differential reading is exactly zero every time. This was so unexpected that I took the kit apart again trying to find the soldering error!

Eventually I came to the conclusion that the chip is really quite accurate, and it's just that the garden is a much less magnetically noisy environment than the house.