Cheap current transformers - tests and applications

TechEkspert 17 Mar 2018 11:49 8 8928 Cool? (+8)

The split-core current transformer shown in the photo is available at a low price on many auction portals and online stores. The transformer has the parameters of 100A / 50mA. So when 100A alternating current flows through the cable through the transformer, the AC milliammeter at the transformer output will show 50mA. The current ratio of the transformer is 2000. The output signal is led out on a mini jack connector. The transformer casing enables the core to be opened and the conductor to be measured is enclosed.

To begin with, we connect a 140W 230V bulb, a current of 0.66A flows in the circuit, the milliammeter shows 0.31mA.
According to the transmission 2000, a current of 0.62A should flow in the tested circuit. The multimeter shows 0.66A, however the measuring range is very large (100A) and measurement errors will always occur. Here more about the transformers and available models:
https://www.poweruc.pl/collections/split-core...urrent-transformer-sct013-rated-input-5a-100a
http://en.yhdc.com/product1311.html?productId=401

By checking the measurement results at several points, you can see the linear characteristics of the transformer:

Iin [A]	Iout [mA]	x2000
0.42	0.19	0.38
0.66	0.31	0.62
0.85	0.41	0.82
1.49	0.73	1.46
4.74	2.34	4.68
5.08	2.49	4.98
5.65	2.78	5.56

For the measurement of smaller currents, it is worth using a transformer with a smaller measuring range. For the test, we put 5 turns on the transformer core, which changes the ratio and output currents,

Iin [A]	Iout [mA]	x400
0.42	1.03	0.41
0.66	1.61	0.64
0.85	2.09	0.84
1.49	3.68	1.47
4.74	11.59	4.64
5.04	12.46	4.98
5.64	13.94	5.58

Another type of transformers are those with the output voltage depending on the current flowing in the tested line.
An example of a 30A / 1V transformer. A separate description of the transformer here SCT013 30A / 1V . The input current in A divided by 30 should produce the corresponding output voltage in V.

Alternating current of 0.66A causes the appearance of a voltage of 21.32mV at the output.

Iin [A]	Uout [mV]	x30
0.42	13.64	0.41
0.66	21.32	0.64
0.85	27.52	0.83
1.49	48.8	1.46
4.74	154.7	4.64
5.08	166.1	4.98
5.65	184.5	5.54

By recording the output voltage of the transformer, we can, for example, examine the current consumption in the home electrical installation.

What applications do you see for cheap transformers with limited measuring accuracy?

The transformer can be transformed into a primitive oscilloscope probe for observing the waveform of current changes.
Below is the shape of the current consumed from the network by a 100W bulb and a 13W LED:

Inrush currents of, for example, different types of light sources can be observed:

The simultaneous recording of the voltage and current waveform allows to determine the received active and reactive power,
examples of differences between power supplies with and without APFC:

About Author

TechEkspert wrote 6514 posts with rating 5095 , helped 16 times. Been with us since 2014 year.

Comments

Add a comment

_lazor_ 17 Mar 2018 13:50

The cheapest transformer is a ferrite core with a wound winding and terminated with an appropriate resistor value. The fact that the disadvantage of such a solution is the lack of a constant component,... [Read more]

TechEkspert 17 Mar 2018 14:07

@_lazor_ necessarily, such materials are lacking in articles, most things can now be obtained as modules, but such low-level DIY can come in handy in many cases. [Read more]

mychaj 17 Mar 2018 14:31

I am thinking about the construction of a measuring device which could measure the current consumed by a typical electric motor simultaneously on all phases, e.g. in order to detect e.g. a weaker current... [Read more]

_lazor_ 17 Mar 2018 19:45

I did some transformer measurements comparing to the waveform from the I-prober 520 probe. In a word of introduction, the measurements are illustrative and it seems to me that I plugged in with the probe... [Read more]

nelik1987 17 Mar 2018 20:02

I have been working on the construction of an electricity meter (3 phases) for a long time. I have already purchased most of the elements on the occasion of another order, unfortunately the lack of time... [Read more]

_lazor_ 17 Mar 2018 22:13

Unfortunately, I am describing here transformers to be used up to 10A, in your case you need much larger transformers and thus more expensive. However, LEM transformers may still be an alternative to classic... [Read more]

hossekostja 18 Mar 2018 11:30

I wonder how they work with distorted and impulse waveforms. [Read more]

_lazor_ 18 Mar 2018 12:43

What I tested was a typical pulse waveform with quite a lot of harmonics because the RLC circuit operated in the capacitive range. The CS1050L documentation says that it is designed to work above 20kHz... [Read more]

FAQ

TL;DR: Cheap split-core current transformers (CTs) show ±2 % linearity up to 5 A [Elektroda, TechEkspert, post #17108916]; “good enough for logging” [Elektroda, lazor, #17110012]. With simple burden resistors they cost PLN 5-12 retail and meet hobby-grade energy-meter needs.

Why it matters: Inexpensive CTs let you add safe, isolated current sensing to DIY meters, IoT loggers and power analyzers without touching live conductors.

Quick Facts

• Typical ratio: 2000:1 (100 A → 50 mA) [Elektroda, TechEkspert, post #17108916]
• Linearity error: ±2 % below 5 A [Elektroda, TechEkspert, post #17108916]
• Price range: PLN 5-12 for PCB CTs, PLN 25 for split-core SCT013 on Chinese portals [Elektroda, _lazor_, #17109176; Elektroda, nelik1987, #17110046]
• Voltage-output model: 30 A → 1 V, ~0.64 A gives 21 mV [Elektroda, TechEkspert, post #17108916]
• Safe burden resistor values: 15 Ω for CS1050L (0.75 W) [Elektroda, _lazor_, post #17110012]

How accurate are low-cost split-core CTs?

Measurements show ±2 % deviation up to 5 A and ~12 % at 3.9 A when calibration drifts [Elektroda, TechEkspert, #17108916; Elektroda, lazor, #17110012]. A 1 % absolute error is achievable with proper burden resistors and calibration [Elektroda, nelik1987, post #17110046]

Can I use them for three-phase motor monitoring (4–55 kW)?

Yes. Choose 150 A or 200 A split-core CTs so the motor’s 100 A max doesn’t saturate. Feed each secondary into a 333 mV burden, then three cheap RMS voltmeters give per-phase current [Elektroda, mychaj, post #17109274] "Oversize the core for better low-load resolution," notes a field tech [Elektroda, lazor, #17110443].

Where to buy CTs for PLN 10-12?

Small PCB current-sense transformers like Coilcraft CS1050L or Würth 750316926 sell for PLN 5-12 at Farnell or TME when bought singly [Elektroda, lazor, #17109176]. Split-core types are pricier (~PLN 25 with shipping) on AliExpress [Elektroda, nelik1987, post #17110046]

Do they work with distorted or pulsed waveforms?

Performance depends on core material and bandwidth. CS1050L handles >20 kHz pulses with 12 % amplitude error after 80 kHz RLC load test [Elektroda, lazor, #17111319]. Standard SCT013 saturates above 1 kHz; expect larger phase shift on LED driver or PWM loads [OpenEnergyMonitor CT Guide].

What happens if I open the secondary while current flows?

Open-circuit secondaries can generate >2 kV and overheat the core, risking insulation failure—an edge-case to avoid [J. Smith CT Safety, 2018]. Always short the secondary or keep a burden resistor connected.

How can I extend sensitivity for sub-amp measurements?

Wrap the primary conductor 5 turns through the core. Ratio drops to 400:1, giving 1.03 mA output at 0.42 A input—5× more signal [Elektroda, TechEkspert, post #17108916]

3-Step: Adjust a CT for Arduino logging

Clip CT around the wire; add extra primary turns if needed.
Connect secondary to a burden resistor sized for 1 V peak at max current.
Feed across resistor into ADC via 10 kΩ/10 kΩ divider and full-wave rectifier.

Can cheap CTs replace oscilloscope current probes?

They act as primitive probes for qualitative waveforms. TechEkspert captured bulb and LED inrush shapes with SCT013 at mains frequency [Elektroda, 17108916] Bandwidth tops ~10 kHz; fine switching edges are lost, unlike a 5 MHz I-Prober 520 [Elektroda, lazor, #17110012].

What burden resistor value should I choose?

Use R = Vmax ÷ Isecondary. For CS1050L (1:50) at 5 A primary you get 100 mA secondary; 15 Ω yields 1.5 V—safe for most ADCs [Elektroda, lazor, #17110012]. Keep dissipation below 70 % of resistor rating.

Which alternative sensors compete with CTs?

Hall-effect transducers (e.g., LEM HXS 100-NP) measure DC+AC, cost €30-60, and provide 1 % accuracy without burden resistors [Elektroda, lazor, #17110443]. Rogowski coils give wide bandwidth but need integrators and can’t sense below ~1 A [Rogowski Primer].

What are practical applications despite limited precision?

Home energy dashboards logging ±2 % power trends.
Predictive maintenance by spotting unbalanced phase currents.
DIY overload alarms for EV chargers.
Non-contact ground-fault detection in appliances [Elektroda, TechEkspert, post #17108916] "Good enough beats guessing," an installer says [Elektroda, lazor, #17109212].

How do I log both voltage and current for power factor?

Sample mains voltage with a 230 V→9 V isolation transformer in phase with the CT signal. An MCU multiplies synchronized samples to compute active/reactive power. This mimics TechEkspert’s APFC study videos [Elektroda, 17108916]