Posted on 23 July 2019

Compact High-Performance Current Transducers

Free Bodo's Power Magazines!




Insulated Highly Accurate Measurements from 1.5 to 50 ARMS

The Power Electronics market is in constant change and always on the lookout for new technologies to achieve better cost and performance. To enable applications with enhanced performance, current measurement is one of the key functions for global improvement in precision and efficiency without increasing the total cost with an easy adaptation to the topology of the total equipment.

By Dominik Schläfli and Stéphane Rollier, LEM SA


A major breakthrough was reached some years ago with the LTS / LTSR current transducers using the Closed Loop Hall effect technology coupled with a dedicated ASIC (Application Specific Integrated Circuit) specially designed for these products.

With the unpredictable temperature a key focus for further development was on the drift of the device over the total temperature range. LEM delivers the solution with the CAS / CASR / CKSR current transducers series covering nominal current measurements from 1.5 to 50 ARMS.

Although we were able to reduce the size even when nobody thought it could still be done, the insulation performances allows usage in standard industrial applications without particular mounting with a rated insulation voltage up to 1000 VRMS (Simple insulation according to EN 50178 standard with following parameters: OV 3, PD2).

The CAS / CASR / CKSR models have been specially designed to respond to the technology advances in drives and inverters in industrial environment requiring better performances in areas such as:
• Common mode influence
• Thermal drift (offset and gain)
• Accuracy (in the whole temperature range)
• Response time
• Insulation
• Size

Working principle

CAS / CASR / CKSR current transducers are closed-loop transducers. Their operating principle is that of a current transformer, equipped with a magnetic sensing element, which senses the flux density in the core. The output of the field sensing element is used as the error signal in a control loop driving a compensating current through the secondary winding of the transformer. At low frequencies, the control loop maintains the flux through the core near zero. As the frequency rises, an increasingly large fraction of the compensating current is due to the operation in transformer mode. The secondary current is therefore a good image of the primary current. In a voltage output transducer like the one discussed here, the compensating current is converted to a voltage through a precision resistor, and made available at the output of a buffer amplifier.

Closed Loop Fluxgate Technology used for the CAS-CASR-CKSR current transducers

Size reduction

The goal was to achieve, on the same effective footprint as LTS & LTSR family of current transducers, a new family of devices with better performance, lower current consumption, wider input range, and a mounted height above the host PCB not exceeding 16.5 mm. Compared with the previous state of the art for closed-loop transducers (LTS models), this represents a 33% reduction in mounting height.

Magnetic circuit

The performance improvements, notably the stability improvements over temperature, were obtained by using a magnetic sensing element based on the fluxgate principle. The fluxgate probe is composed of an elongated strip of soft magnetic amorphous material, surrounded by a winding. To couple it to the flux in the magnetic circuit, the air gap has to take a more complex form than with a Hall cell. It takes the form of an elongated cavity, which shields the fluxgate core from the direct field produced by the primary and secondary conductors. Due to their differing geometries, they produce different fields, and would each also produce a geometrically different field distribution in the fluxgate core. In closed-loop operation, an unshielded fluxgate probe would be maintained at an apparent zero flux operating point, but the flux density in the core could remain locally uncancelled, leading to premature saturation of portions of the core as input current rises, ensuing fall in probe sensitivity, and a much reduced transducer input range. Locating the probe in a cavity ensures it couples to the cavity walls, which are at much more uniform magnetic potential, whether the source of that potential is the primary or the secondary conductor, resulting in excellent local flux density cancellation in the fluxgate core, and wide input range. Only a small portion of the flux is coupled through the sensing element core, most of it flows through the air gap.

Flux density in magnetic core caused by primary current alone


The rectangular magnetic circuit presented the opportunity to accommodate 5 primary conductors instead of the 3 which can be realized when a ring core is used in the same form factor.

Primary conductor positioning

The requirement for a backward compatible pinout, coupled with the placement constraints on the magnetic circuit, meant that it would have only been possible to house 4 straight primary jumpers. With only 4 jumpers, it would not have been possible to provide a high current differential version with 2 jumpers in parallel per input and adequate separation. For this reason, a solution with bent primary jumpers was selected.

Flexibility to configure backwards compatible, increased separation or difference mode pinouts

The design of the housing and the jumpers was then optimized for late differentiation of the input conductor configuration. The fully assembled housing presents several holes and retention notches. Wires can be inserted into any combination of holes. The wires are then bent into the desired jumper shape, and clipped into the retention notches. The notches are offset with respect to the holes, allowing for versions with normal or increased separation between primaries. With this system, many versions can be devised to optimize separation, jumper resistance, differential sensing, and reference inputs.

Measurements of individual jumper sensitivity performed on samples show a maximum sensitivity mismatch of ca. 0.3 % for the right bent jumpers, and 0.2 % for the left bent jumpers. In difference measurements, typical common mode rejection ratios in excess of 50 dB can be expected.

Sensitivity mismatch, left-bent jumpers, relative to jumper 0 (central jumper), 5 samples and average

Fig. 5 shows the results of a null test in differential mode. The primary was looped in and out of the transducer (difference configuration, right bent jumpers, see Fig. 3). A current step of 25 A was applied. We compare the LTS 25, which had not been designed specifically for differential operation, to the CAS 25. The steady state error, relative to the applied common mode current, is ca 2% for the LTS , and less than 0.2 % for the CAS.

Null test for an applied step of common mode current, and 0 differential current

The CKSR has 4 primary conductors instead of 3 as in the CAS and CASR models. It is possible to measure down to 1.5 ARMS nominal (using a CKSR 6-NP) with a 4 turn layout) . With this layout, the totalized current passing through the transducer aperture is still 6 ARMS (its designed nominal current).

The CKSR primary pinout is different from the CAS and CASR models, achieving higher creepage and clearance distances (8.2 mm, internal distances). This can be of interest for applications requiring higher working voltages.

Using EN 50178 standard as a reference for industrial applications, the possible reinforced rated insulation voltage is of 600 VRMS (Conditions of use: CTI: 600 V (group I), Overvoltage category: III, Pollution Degree: 2).

Immunity to transient common mode voltage

Capacitive coupling between the primary conductor and the compensation coil causes common mode currents to flow through the transducer electronics when the primary to secondary voltage varies (e.g. during switching), usually causing an unwanted signal to appear on the output.

While the amplitude of the error pulse appearing during the transient itself is of minor concern (in most applications, common mode transient appear as a consequence of intentional switching, and the transducer output can be ignored during a known short interval), a slow recovery due to ringing can be more problematic.

State of the art transducers use shielding to divert the current through the ground pin, preventing it from entering the transducer signal path. This shield is either realized as a metal foil, a metallized plastic enclosure, or, as in the present case, a final winding layer connected to ground on one end, the other end left floating.

Example of transducer response to a common mode voltage transient, if not equipped with an electrical shield

Error voltages for positive and negative transients are symmetrical, resulting in negligible DC error components, even if the rising and falling edges have different slopes. Despite the presence of the shield, the response to a current step is excellent, with very little overshoot (less than 5%), and fast settling time to the final value.

Transient common mode voltage response, transducer with shield winding

Electrical data

The CAS / CASR / CKSR current transducers series have been designed to work with a single + 5 V power supply to cover an input current range up to 150 A, with nominal currents ranging from 1.5 A to 50 A.

All three transducer families use the same precise, stable 2.5 V voltage reference to provide the output offset.

Key electrical parameters

Despite the reference stability, it is the dominant contributor to offset drift, and for utmost stability, the CASR and CKSR families provide a dedicated pin to access the reference voltage. It becomes possible to measure the difference between the output voltage and reference, eliminating the influence of both the initial uncertainty on the reference voltage and its variations over temperature and time.

The REF pin has two basic functional modes:

• The first mode is called “Ref out mode”. In this mode, for a primary current of 0 A, the output voltage is equal to the voltage at the REF pin. The 2.5 V internal precision reference voltage provided at the REF pin stays stable although the primary current changes and is used by the transducer as the reference point for bipolar measurements.
• The second mode is called “Ref in mode”. In this mode, you can apply an external voltage to the REF pin to override the internal voltage reference, the external voltage being then used by the transducer as the reference point for measurements. By using an external reference, it is easier to connect the transducer to devices such as an ADC.

In most applications, the output of the transducer is connected to an ADC whose output is processed by a DSP or a microcontroller.

The internal reference of these DSPs or ADCs can go down to 1.8 V.

In this application, if you have an internal reference in the DSP with external access, you can supply the transducer’s reference in with it.

Step response, transducer with shield winding, less than 5 percont overshoot

The minimum external voltage is 0 V and maximum 4 V. However, this mode defines different measuring ranges according to the level of the external voltage reference used (0 to 4 V) and according to the model used (6, 15, 25 or 50 A model) (Please refer to the relevant products data sheets).


We have shown that, despite the challenges inherent to a very compact design, it has been possible to refine the performance of closed loop current transducers in some key aspects:

• A really good accuracy especially in temperature (low initial offset, low thermal drifts for gain and offset)
• High flexibility for specific customer applications (wide choice of current ranges: 6, 15, 25 and 50 ARMS, each model is multi range, unique packaging)
• small current measurements as low as 1.5 ARMS while maintaining high performance
• high Creepage and Clearance distances with the CKSR models for higher insulation
• Reliable operation in rough environmental conditions such as high humidity combined with high temperature
• Recovery times after transient common mode voltage events better than 500 ns
• Primary common mode current rejection ratio typically 50 dB

These advantages are suited for high performance drives, inverters for renewable energy sources integrating a good control of the DC current injection in the grid, servo drives for wafers production or highly accurate robots and all kinds of low drift applications.



1) W. Teppan, “Transfer Functions of Current Sensors”, 5th Int. IMEKO, TC-4 Symposium on Electrical Measuring Instruments for Low and Medium Frequencies, Vienna, Austria, 8-10 April 2019.
2) Pavel Ripka, editor, “Magnetic Sensors and Magnetometers”, Artech House, Norwood, MA, USA 2001.
3) E. Favre, W. Teppan, “ State-of-the-art in current sensing technologies“, Symposium on Power Conversion and Intelligent Motion, PCIM, Nürnberg, Deutschland, 2003.
4) W. Teppan, “Magnetic Field Sensor and Electrical Current Sensor therewith”, PCT Application WO 2004/074860 A1.


VN:F [1.9.17_1161]
Rating: 0.0/6 (0 votes cast)

This post was written by:

- who has written 791 posts on PowerGuru - Power Electronics Information Portal.

Contact the author

Leave a Response

You must be logged in to post a comment.