Archived Forum

Best insight into the SCT-013-XXX is here -

http://openenergymonitor.org/emon/buildingblocks/report-yhdc-sct-013-000-current-transformer

I belive the turns ratio is 2000:1 and I think the only other difference through the range is the presence/value of a burden resistor or protection diodes.

I believe richmc is correct, but it's unproven. The value of the internal burden resistor (where fitted, i.e. all except the -000) appears to be selected on test to give the rated output.

I also have a suspicion that different models come from different factories or sub-contractors, since the cable screen is sometimes connected and sometimes not.

as stated, it is unconfirmed, but if the number of turns for the 100 amp and the 30amp are the same, then the only real difference is the addition of the internal burden resistor and the diodes.

assuming that is the case, then we would expect that an Arduino would see a full scale, 100 amp reading at it's 5 volts on the ADC pin.

that would also mean that a 30 amp reading would only be 30% of that ?

based on the simple fact that if you are only ever going to see a 20 amp load, then the ADC of the arduino will be completely un-effect and there is no 'loss' of accuracy and no 'un-used' range of the ADC.  it is just that a 20 amp load will only ever rise to a certain voltage.

my question is then if we can do anything to get a more accurate reading in the 0-20a range.

a suggestion is to add a zero-dropout-circuit. so that currents less than 0.1, the software just reads 0.  more peace of mind than anything else.

A little context to your post would be helpful - this thread is over 2½ years old!

We believe all the SCT-013-xxx range are the same device. The only difference is the value of the internal burden resistor, all except for the -000 that is, which has 8 V (from memory) protection diodes instead of an internal burden, and requires you to provide an external burden. Therefore, if you require a 0 - 20 A input range, your best choice is to use the -000 variant and calculate your external burden resistor value to give you a peak-peak voltage across the burden that is few percent less (*) than the maximum input voltage range on a 20 A primary current.

* to allow for component tolerances - 5% should be enough, but if you anticipate a 'peaky' current wave, you should allow a lot more unless you can calculate or measure the crest factor and use that when calculating the peak-peak value.

It wou be nice if the burden resistance could be controlled in software, even if it's just two or three ranges. The software could sample at the lowest range, and step up to the next range if needed.

Or, if redesigning the board, use a dedicated IC like the Cirrus Logic CS5463, which should in theory have better resolution.

I know this has been an old thread, but interest is still high.  I wish that someone would confirm with the factory as to the turns and also the machines they use so as to get an accurate count.  the machines that to that these days are so simple they should be dead on.

 I have photos of my -030 that is open.

uploaded 2 pics in case that helps.

mine has 2 resistors. a 000 and a 62

it also has diodes 7.5 each

The bare wires from the transformer hover over the board, protected by varnish, but look like they could cause problems for some.

Normally, the turns ratio is adjusted in manufacture according to the properties of the core etc so that the errors over the working current range are minimised. Therefore, nobody can "confirm with the factory as to the turns" because it is the factory's prerogative to change it in order to meet the published specification - and just in case you didn't know, current transformers are specified by the current ratio in actual units, and not by the turns ratio. That should of course be obvious from my first sentence.

Therefore, the SCT-013-000 is a 100:50 mA transformer, not a 2000:1 (which would mean it would handle a 2 kA primary current and give 1 A of secondary current flowing in the burden).

In terms of the "voltage output" types, it is normal for the manufacturer to adjust the value of the burden resistor in manufacture again so as to obtain the specified output over the required range. That's the reason for the multiple pads visible in your photo.