Community
OpenEnergyMonitor

Community

SCT031QL 600A, 0.333V inaccurate reading

Hi
i used a current sensor of YHDC 600A 0.333V (voltage type ) that exist in ebay

with NodeMcu(ESP8266) ,when i measure 1 A the result is not accurate ,according to calibration equation (current constant =rated input /output voltage) the result is 1801.8 !

the equation is here (https://learn.openenergymonitor.org/electricity-monitoring/ctac/ct-and-ac-power-adaptor-installation-and-calibration-theory?redirected=true )

this is the result at 1801.8 calibration and the sensor is connected (3.7 A to 4.2 A) .

when the sensor is not connected the result is (1.7 A to 2 A).

when i change the calibration to the default (111.11) the result is (0.15 A to 0.18 A) .

You are using a 600 A c.t. to measure 1 A? With an ESP8266?

The ESP has a 10-bit ADC. That means that it can detect 512 different voltage levels either side of zero. You are trying to measure a current that is 1/600th of the maximum that the input can measure. I’m not surprised that it is not accurate. You can read more about this in ‘Learn’.

If you want to measure currents that are around 5 A, you should use a 5 A (or 10 A) c.t, not a 600 A one.

thx robert
I used 50A,100A before ,if you read the datasheet of this CT you will find that this ct measure current range from 0 to 600A , i use this ct to measure different generator current its range up to 600A , i measured the small current and the result not accurate because the this CT in practical sense the current from 5A TO 600A so i tray a 9A to measure then the result is very accurate at current constant 1801.8 .

Thank you for reminding me to read the data sheet. I did read it, but I also think I know and understand the phasor diagram that explains how a c.t. works, and what the influences of various factors are in relation to the accuracy. And that is before one considers the effect of the input circuit and the errors of the analogue to digital conversion process when the peak-to-peak value of the 1 A input is at best only 1.66 times the size of the least significant bit.

Hi Ahmed , did you come right with the 600A CT?

Normally, a c.t. has its accuracy specified over a limited range - usually given as percentages of full-load current. As I intimated above, at low currents the accuracy deteriorates due to the way the current transformer needs to extract power from the current it is measuring to work.

If you read the thread in its entirety, you’ll see that the OP was trying to use the c.t. to measure a current well below the point where I would expect it to be reasonably accurate. I would expect the lower limit of accuracy to be 1% FLC at best, and this is borne out by the OP’s comment that at 9 A (1.5% FLC) he had accurate readings.

Therefore, I think that if you use the c.t. to measure currents for which is was designed, and in the way it was designed to operate, you should have accurate results.

1 Like

You need to use a burden resistor on these big current transducer , These do not output Voltage , they output proportinonal Current.

The title of this thread as well as the fact the OP stated in his post that the CT he used is indeed
of the voltage type seems to be a point you missed, not to mention that it contradicts your statement
These do not output Voltage , they output proportinonal (sic) Current.

Any CT that puts out 333 mV already has a burden. That’s why its output is voltage vice current.

@Bill.Thomson

I wonder whether @j_p has any qualifications or experience in electrical engineering?
This post looks more like a spammer testing the water.

I actually am a certified Electrical Engineer in Canada. With 10 year+ experience.
Please forgive my no so good english as my native language is French.

Nevertheles , I stand My point,

1-
Coil Type current trasnducer ALWAYS need to have a burden resistor. Most Of the Time It is include in the coil packadge itself for simplicity of instalation. This is often the case with <100A CT­
But as they get biger , they are not include inside , as it alows you to fine tune the reading range you want to get

2-
I took the time to consult the official datasheet of the Sensor mentioned , Did you ?
http://en.yhdc.com/product/SCT031QL-382.html
There are 2 datasheets for the exact same model number
One is having the burden resistor inside , they other is not

Meaby MEABY you have the other type . Seller of these cheap CT are not always the sharpest to details proprely wath they sell

3-
Having a Super High reading of 1801.8 . IS the typical result of not having a burden resistor in place. The Votlage will spike up, because there is nothing to limit it

4-
Meaby the burden resitor inside is broken, burned, misconnected ?

Im just Tring to help here, Dont be rude please

Ok , I must admit I’ve read it all wrong , and too fast
The burden resistor is probably ok
It is a precision issue , as the first post you have said :


The ESP has a 10-bit ADC. That means that it can detect 512 different voltage levels either side of zero. You are trying to measure a current that is 1/600th of the maximum that the input can measure. I’m not surprised that it is not accurate. You can read more about this in ‘Learn’.


My mind got stuck at the 1801 reading , with I’ve misread to 1801 A , Sorry , my bad

You’ll be pleased to know that I’m a Chartered Electrical Engineer here in the UK. There’s a branch of my professional body, the IET, in Toronto. I’ve been a member (it was previously called the IEE) for over 50 years now.

Exactly. And users who buy parts without a full knowledge of what they are doing and without checking with us first, sadly run into problems like this fairly frequently. As you suggest, it’s possible that some are persuaded by “incomplete” advertising by unscrupulous traders.

With an input that’s looking for about 1.1 V rms (0 - 3.3 V d.c input range), a part of the O.P’s problem was exactly because he had a 0.333 V output c.t. Had he used a ‘proper’ c.t. and the correct value of burden resistor (bearing in mind the VA rating of the c.t.), he would probably have been able to use rather more of the ADC’s input range.

Although there is no guaranteed accuracy range of the c.t. quoted on the data sheet, and as it is not rated “revenue grade”, I would not expect it to maintain the claimed 1% much below about 15 A, and so the other, probably more significant part, is the current being measured was well below that, so there would be a large error due to the magnetising current required alone, even before the voltage reaches the ADC.

Sorry I accused you of being a spammer, but it’s a common technique that these pests/bots employ: Pick some words from the original post so that the automatic filters allow the post through.