Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
A POWER TRANSFORMER ASSEMBLY
Document Type and Number:
WIPO Patent Application WO/2014/035251
Kind Code:
A1
Abstract:
It is disclosed a power in which the core includes a first core element (11) of a first permeability and a first maximum remanence, stacked on a second core element (12) of a second, lower permeability and a second lower maximum remanence. This allows the start or inrush current of the transformer to be controlled independent of the idle or magnetizing current and vice versa.

Inventors:
NORDGAARD OLA O (NO)
Application Number:
PCT/NO2013/050128
Publication Date:
March 06, 2014
Filing Date:
August 06, 2013
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
NORATEL AS (NO)
International Classes:
H01F30/16; H01F3/14
Foreign References:
JP2006140243A2006-06-01
US20080157914A12008-07-03
US20030210123A12003-11-13
US20040178877A12004-09-16
FR2223800A11974-10-25
GB1015044A1965-12-31
JP2006140243A2006-06-01
Attorney, Agent or Firm:
OSLO PATENTKONTOR AS (Oslo, NO)
Download PDF:
Claims:
11

C l a i m s

A power transformer assembly including at least one primary winding for inducing a magnetic field in at least one core, and at least one secondary winding, said at least one core including first and second core elements, c h a r a c t e r i z e d i n that the first core element has a first sectional area and a first permeability and a first maximum remanence, the first core element being adapted to provide, together with a defined contribution from the said second core element, based on the relation between the permeability and the sectional areas of the said two core elements, a predefined maximum magnetizing current when excited within an unsaturated region of induction,

and the second core element has a second sectional area and a second permeability and a second maximum remanence, the second core element being adapted, with a defined contribution from the said first core element, based on the relation between the maximum remanence levels and the sectional areas of the said two core elements, to provide a predefined maximum inrush current in the at least one primary winding. 2. A power transformer assembly according to claim 1, wherein the assembly includes a first transformer (15) with a first primary winding, a first secondary winding and a first core (16) forming said first core element, a second transformer (17) with a second primary winding, a second secondary winding and a second core ( 18) forming said second core element, the first primary winding being connected in series with the second primary winding, the first and second transformers (15, 17) being designed to induce magnetic fields in their respective cores (16, 18) with identical field strengths. 3. A power transformer assembly according to claim 2, wherein the first

primary winding and the second primary winding have identical number of turns.

4. A power transformer assembly according to claim 1, wherein the assembly includes one core, the core including a first core segment (11) forming said first core element and a second core segment (12) forming said second core element. 12

A power transformer assembly according to claim 1, wherein the core elements are stacked onto each other with common primary and secondary windings.

A power transformer assembly according to claim 1, 2, 4 or 5, wherein the permeabilities and corresponding maximum remanence levels are determined by air gaps in said core elements.

A power transformer assembly according to claim 4, wherein the core includes first cuts (21) forming air gaps through the full cross section and second cuts (20) forming air gaps covering parts of the core sectional area.

A power transformer assembly according to any of the claims 1 to 7, wherein the second permeability and second remanence are lower than the first permeability and first remanence.

9. A power transformer assembly according to any of the claims 1 to 7,

wherein the core elements are toroidal core elements.

Description:
A POWER TRANSFORMER ASSEMBLY Field of the Invention

The present invention relates to power transformers in general and in particular toroidal transformers. Background

Toroidal transformers have many beneficial properties compared to traditional bladed types, but suffer from one major drawback; a larger inrush- or start current. However, the inrush current is a general problem for all optimized power

transformers. When connecting a transformer to the mains we'll always get a train of inrush current pulses with a magnitude determined by the polarity and phase of the applied voltage in the moment of connection, relative to the remanence in the core and the impedance in the primary winding. The initial pulse in the pulse train being the largest with the succeeding pulses rapidly attenuating as the core is

establishing its normal magnetic relations.

The amplitude of this initial current pulse/inrush pulse will have its highest value when the core has a high remanence, which is one of the properties of a toroidal core, and the voltage is connected to the transformer in the zero transition point followed by the half cycle with a polarity generating a magnetic field in the core with the same magnetic polarity as the remanence. The core is then rapidly brought into saturation, leaving the current to be limited only by the resistance and leak inductance in the winding. This inrush current pulse can in this case be more than 50 times the transformer's rated current for designs with high remanence and low winding resistance/leak inductance, causing tripping of fuses and overloading of power sources, etc., etc.

It is possible to limit this inrush current by using a current-reducing resistor and a relay (contactor). The resistor is connected in series with the primary winding in the moment of connection, and shorted by the relay/connector after 50-100 msec.

This solution works well in most instances, but is not accepted in certain

applications as it introduces an additional possible source of error. This is the case for systems where high reliability is required: E.g. medical equipment, offshore applications, etc. For systems with UPS backup power, which is often the case for the above mentioned applications, there are also stringent requirements as to the maximum values of both inrush- and idle currents, which is a design challenge. For such applications, many suppliers traditionally choose to use oversized El transformers. In this way, it is possible to force oneself inside the maximum limits for the inrush value defined by the standards but there will be no harmonization with the idle current value requirement. An ideal solution would of course be if the inrush current and idle current could be controlled independently in the design, to the desired values, with a minimum increase in the core size. This is however not possible in traditional transformer design.

From JP 2006140243 A there is known a toroidal transformer with an air gap in the core for reducing the inrush current. The air gap covers 80% of the core area, while 20% is left un-gapped in order to hold the laminations together. This is the only purpose of the un-gapped portion, and as shown by the hysteresis curve in Fig.4b, this portion has no influence on the magnetic properties of the core, or the inventor not being aware of its magnetic influence.

Summary of the I nvention

It is an object of the present invention to provide a solution to the above mentioned need.

This is obtained by the present invention in a solution as defined in the appended claims.

The design parameters for controlling transformer inrush current and idle current are different. For a given core size/winding combination the requirement for a low inrush current is mainly a low remanence. The requirement for the idle current is a minimum value for the permeability. The basic idea is therefore to use a core consisting of two tailor-made stacked elements where one element is designed based on the idle current requirements and the other one basically on the inrush current requirements. This is for the "idle" core done by introducing an air gap to reduce the remanence to a level where the necessary permeability needed to meet the idle current requirement still is maintained. (The introduction of an air gap will reduce the remanence as well as the permeability.) In the "inrush" core the same basic principle is used, but as the permeability requirements always are much lower to meet the inrush specifications, a wider air gap can be used, for a larger reduction of the remanence.

By stacking the above core elements and creating a predefined total sectional area and predefined difference between the mutual core sectional areas, transformers may be designed with a modified/combined hysteresis curve leading to a

start/inrush current and idle/magnetizing current according to desire. By selecting suitable core materials, cross sections and air gaps in each of the core elements, the requirements for inrush- and idle current can be satisfied with a minimal total core cross section (minimum total core size). Depending on the requirements, the total cross section of the transformer will normally be 40-60% larger than for a similar optimized power transformer where no provisions for the reduction of the inrush current is made, but compared with the use of an oversized El-solution for inrush current limitation it is a significant achievement.

Brief Description of the Drawings The invention will now be described in detail, with reference to the appended drawings, in which:

Fig. 1 is an illustration of a typical inrush sequence in a transformer (prior art),

Fig.2 is a diagram showing the flux density in a transformer core (prior art),

Fig.3 is a diagram showing the flux density for two different operating voltages (prior art),

Fig.4 shows how the operating voltage may be reduced to avoid the inrush current,

Fig. 5 shows the flux density (hysteresis) diagram for a core with air gap (prior art),

Fig. 6a-e illustrates several different embodiments of the invention, Fig. 7 is the flux density (hysteresis) diagram of the inventive transformer, and

Fig.8 shows the contribution of the inrush core element to the control of the idle current. Detailed Description

Fig. 1 shows the applied AC voltage (2) and the current (1) in the primary winding of a toroidal power transformer when power is applied (no load connected) in the zero transition point. There is a current pulse in each positive half cycle. The first peak is large as power is applied, but the pulses rapidly sound out. However, this large first pulse poses a problem, particularly in large power transformers.

Fig.2 shows the magnetic flux density as a function of the magnetic field in a transformer with operating voltage applied. The flux density follows the well-known hysteresis diagram (3). When the applied voltage is removed, as it passes through zero, the core will be at the peak of its flux density. Even if the magnetic field then falls to zero, the flux will not go to zero, but fall to a rest value B R , i.e. the core will keep a residual magnetism. This residual magnetism is called remanence and will vary with the type of transformer according to the properties of the core and its construction. Fig.3 shows the hysteresis curve for two different operating voltages. The angle and height of the hysteresis diagram determine the size of the magnetizing current (idle current). In standard mains power transformer designs, the operating voltage, core sectional area and number of turns in the primary winding are adapted so that the induction swings between ± ~ 1.5 Tesla (T). This is shown as hysteresis curve (4), while the corresponding operating voltage is shown as (4') and the resulting idle current as (4"). If the voltage (and induction) is increased, the idle current will rapidly increase as the core goes into saturation. Curve (5) shows the situation when the induction has been increased to 1.75 T and the transformer has become partly saturated. When the core material becomes saturated, the magnetic effect of the core disappears and only the resistive and inductive effects of the winding alone decide the size of the idle current. Thus, there will be a sudden and substantial increase in the current drawn, as shown at (5"). At an induction higher than 2.0 T, the core is regarded as deeply saturated. In the illustration, curve 5" is shown as a nice sine wave, but in reality the shape of the curve will become heavily distorted as the transformer goes into saturation.

The remanence of a magnetic toroidal core without an air gap, working with an induction of ± 1 ,5T, may have a typical maximum remanence value of 0.9 to 1 ,4 T. (Let us assume the remanence is 0,9T in this example.) When power is applied to the transformer in the "unfavorable" zero transition point, the flux will start rising from 0.9 T instead of -1.5 T and quickly drive the core into saturation. The sum of the remanence (0,9T) and the induction represented by the operating voltage (2 * 1.5 T) constitute a theoretical value of 3.9 T, which is far past the saturation value. The current in the transformer will then, when the core saturates, only be limited by the resistance and leak inductance in the primary winding. This is what causes the inrush pulse.

In order to avoid this inrush pulse completely, the operating voltage must be considerably reduced. For cold rolled electrical steel we can assume a beginning saturation to start at around 1.6 T and with a remanence of 0.9 T, there will be only 0.7 T (±0.35 T) available up to this point. To stay within this magnetization span, the operating voltage must be reduced to almost 20 % of the original value (± 1.5 T). Then, we will exploit only 20% of the core's capacity. This is illustrated in Fig.4. (6"') shows the available value for the operating voltage spanning the range between the remanence point B R and the saturation value (7). The operating voltage (6') will then result in a hysteresis diagram (6) spanning only a small part of the core's capacity.

An alternative to reducing the operating voltage is of course to increase the sectional area of the core five times and keep the original operating voltage.

However, a five times increase in the sectional area of the core equates

theoretically to 25 times increase of the volume. In practice, the wire size must be increased to compensate for increased resistance in the wire due to the increased length, which means the core center window must be increased accordingly. This probably means a 30-50 times increase in volume and weight. The transformer becomes enormous. For cores with higher remanence the situation will be even worse.

The above example is based on values to avoid the inrush pulse completely. In practice, an inrush current of 5-10 times the normal nominal operating current is allowed. Thus, the size of the transformer will not increase as much as indicated, but nevertheless the possible efficiency/ capacity of the core is still in no way fully exploited.

If the permeability of the core material is reduced, e.g. by introducing an air gap, the shape of the hysteresis curve will be flattened (10), Fig. 5. (9) shows the curve for the uncut core. As shown by the figure, the remanence will also be appreciably reduced, which is a desirable feature. If the air gap reduces the maximum remanence from a value of 0.9 T to 0.1 T, the core will still be driven into deep saturation at start up, but the sum of the remanence and the magnetic flux caused by the operating voltage will correspond to a theoretical induction of 3.1 T, compared to 3.9 T in the un-gapped core. This means the operating voltage only has to be reduced to about half the original value to obtain a transformer without inrush current, or the core sectional area doubled instead. This is much better than for a core without air gap, which requires a five times increase in sectional area, or more.

A drawback with this air gap is a considerable increase in the idle current, which often is unacceptable. One way of reducing the idle current is to reduce the operating voltage, which again is problematic.

Fig. 6a illustrates the principle behind the present invention: A transformer (15) with a high permeability (16) core is connected in series with a transformer (17) with a low permeability core (18). The permeability values are determined by installing air gaps in the cores.

There are different ways of introducing the air gaps. They can be distributed as in powder cores or in bladed El cores. In toroidal electrical steel cores the air gaps are normally introduced by cutting the core and inserting a non-magnetic/low-magnetic material of required thickness in the gap.

The first core (16) is designed with a permeability adapted to maintain the necessary inductance in the primary winding to keep the magnetizing current close to, but not exceeding, the maximum value specified, and at the same time securing a lowest possible remanence.

The purpose of the second core (18) is to add to the cross section of the first core the necessary cross section to avoid the design to go into saturation, when connected in the "unfavorable" zero transition point of the voltage, before the instant value of the sine voltage across the winding(s) has been reduced to the necessary magnitude to keep the inrush current within the required value. As the voltage applied to the primary winding is following a sine wave function, this normally means after ~5/6TT of the first half period. Any remanence in this core (18) will add to the required cross section necessary, but as the purpose of this core is to control the inrush current it can be designed with substantially lower permeability, and correspondingly lower remanence than the first core (16). The air gap size in this core (18) is not critical, e We are talking about air gaps to bring the maximum remanence down to around 0,1T, which in practical terms means a reduction of around 90 %.

In the transformer assembly, the hysteresis diagrams of the core elements (22 and 23) are summed forming a new hysteresis diagram (25) with a peculiar shape, Fig. 7. The figure shows the hysteresis for the core in a situation of a beginning saturation. The hysteresis diagram has a central part (24) representing the normal operating range, where the curve is almost as steep as the hysteresis diagram for the first core (idle) element. The curve is relatively flat in its outer parts, (25), (26), which is dominated by the second core (inrush) element (23). The transformer will only enter the ranges (25), (26) at start-up. By adjusting the size of the air gap (permeability) of the first core element, we may easily control the idle current while securing a lowest possible remanence in this element, while the second core element with its lower remanence will secure a low inrush current with minimum added core cross section.

Functionally, as the two cores add their parameters they will also contribute to the main function of the other core, meaning the inrush core will contribute to control the idle current with a factor depending on the cross section and permeability relative to the "idle" core element. This is shown in Fig.8. where at nominal operational voltage the induction is only spanning a part of the core's capacity with a resultant curve as shown (28). Here it shows the inrush core's contribution (27) compared with the idle core (26).

Likewise, the idle core will contribute to the control of the inrush current with its saturation induction minus the remanence. Even though it is fully possible to implement an embodiment as shown with separate transformers, it is preferred to combine the cores into one transformer, i.e. with one set of windings on two cores. Such an embodiment of the invention is illustrated in Fig.6b: A first core element (11) (with higher permeability) is stacked with a second core element (12) (with substantially lower permeability). The windings 14 are wound onto the stack of the core elements (11), (12). The figure illustrates that the permeability of the elements is lowered by the inclusion of air gaps (13) with different sizes. To completely avoid the inrush current, keeping the no of turns in the winding unchanged, the total sectional seize of the combined core must be increased to about twice the size, reducing the working induction to about half the value (± 0.75T), compared to the working induction of a core in a standard toroidal mains power transformer made of cold rolled electrical steel (± 1 ,5T). Normally, some inrush current will be accepted (allowing an inrush flux density of about above 2,0T) meaning the total core combined sectional area (the no of turns in the winding unchanged) only need to be increased between 1 ,3-1 ,4 times compared to the core sectional area of a standard toroidal mains power transformer with no provisions made to control the inrush current. The benefit of the inventive transformer is that the idle current and the inrush current may be controlled independently.

Fig. 6c illustrates a third embodiment of the invention: Using a single toroidal core with partial cuts through the core to obtain air gaps affecting only a part of the cross section. In the core (19) there has been made a cut (20) providing an air gap. The cut goes only partly down into the core and thus will only affect the upper part of the core. Several cuts may be made, of different size and number, Fig.6d. This opens for an option to use a single core with cross section equal to the cross sections of the combined cores, with one or more full cut air gaps (21) sized to control the idle current and one or more larger partial cut gaps (20) to control the inrush current (Fig. 6e).

This solution with partial cuts can also be used in bladed transformers (El, Ul) where the distributed air gaps caused by the blading already creates "idle current air gaps" and it is just to add partial cuts for the control of the inrush current. Determining the cross section (size) of the core elements:

A calculation sequence to determine the design parameters for a transformer based on this idea can be as follows:

1) The cross section of the "idle current" core element of the transformer is calculated in the usual way, as for a standard toroidal power transformer, based on the specified primary voltage, secondary voltage(s) and load (and any specifications as to physical size). In addition, values for the magnetic length of the core, the number of turns in the windings as well as wire diameter and the resistance of the windings are found. ) Based on the specified idle current (l 0 ), the necessary relative permeability (μ) of the core is computed from the equation: l 0 = (U * Ι)/(2*ΤΤ*ί*μ*μ 0 * n 2* A) where U is the applied voltage in Volt, I is the magnetic length of the core in meter, f is the frequency in Hz, μο is the permeability constant for vacuum (4*μ*10 "7 ), n is the number of turns in the primary winding and A is the core sectional area in m 2 . From a computational point of view, this equation is the applied voltage (U) divided with the inductive impedance of the winding (the rest of the equation). Normally, this impedance is much lower than the resistive component caused by eddy current losses, and thus we may ignore this last, resistive component. ) Based on the nominal permeability (μ η ) of the core lamination used, the size of the air gap necessary to reduce the permeability to the limiting value (μ) computed in 2) above, is calculated. An equation is used:

1/μ η =Χ+((1/μ)*(1-Χ)) where X is the relative length of the air gap in relation to the total magnetic length of the core. Or said in another way: (X * 100) is the length of the air gap in percent of the magnetic length of the core. In order to obtain a fully correct value of the air gap, the value must be adjusted with an air gap constant dependent on the size of the core sectional area, the geometry of the core section and the size of the air gap, among other. ) From stated or measured maximum remanence in the un-cut core

lamination, the resulting remanence of the core with the air gap is computed. The change in the remanence is close to linear with the change in the permeability, which means e.g. that a reduction in the permeability of 80% will result in about 80% reduction in the remanence. ) From the computed value of the resistivity of the wire, the computed inductive impedance of the primary winding with μ = 1, and the maximum value of the inrush current, an approximate value of the maximum voltage corresponding to the specified inrush current limit may be computed. 6) The corresponding point on the voltage curve (on the first half period of the sine wave curve) is then determined. Normally this is at around 5/6TT, but experience should be consulted when determining this value.

7) When the voltage point is determined, the required additional core cross section needed in order to saturate the core at this point is computed, taken into account the remanence of the l 0 -core.

8) To fill the required cross section gap, an additional, new core (the inrush core) with air gap about five to ten times larger than the air gap of the l 0 - core is added. The size of this air gap is not critical but normally the aim is to bring the maximum remanence down to value around 0.1 T (a reduction of the remanence with a factor of about 90 %). Further reduction of the remanence will have limited effect on the inrush sequence and only reduce the inrush core's contribution to the control of the idle current.

9) Then a new calculation of the design based on the above parameters should be made to optimize the size of the transformer by taking into account the contribution of the inrush core to the idle current function and also to adjust the design for the increase in wire resistance due to a larger core cross section caused by the adding of the inrush core etc.

10) Finally, an adjustment of the induction after test of a prototype should be considered to optimize the function relative to the magnetizing current and the inrush requirement. Normally a working induction of this design will be around ± 1.0 Tesla.

We then have a design where idle- and inrush current values are

controlled/calculated independently to the desired values, with a minimum increase in the core size.

The invention has been explained including two core elements, which may be realized as two independent transformers connected in series, as two cores stacked together with common primary and secondary windings, or as one core with two core segments. However, in some cases it may be necessary to increase the number of core elements, in particular for tailoring the frequency range of the transformer assembly, which is dependent on the magnetic properties of the core material. C l a i m s

1. A power transformer assembly including at least one primary winding for inducing a magnetic field in at least one core, and at least one secondary winding, said at least one core including first and second core elements, c h a r a c t e r i z e d i n that the first core element has a first sectional area and a first permeability and a first maximum remanence, the first core element being adapted to provide, together with a defined contribution from the said second core element, based on the relation between the permeability and the sectional areas of the said two core elements, a predefined maximum magnetizing current when excited within an unsaturated region of induction,

and the second core element has a second sectional area and a second permeability and a second maximum remanence, the second core element being adapted, with a defined contribution from the said first core element, based on the relation between the maximum remanence levels and the sectional areas of the said two core elements, to provide a predefined maximum inrush current in the at least one primary winding.

2. A power transformer assembly according to claim 1 , wherein the assembly includes a first transformer (15) with a first primary winding, a first secondary winding and a first core (16) forming said first core element, a second transformer (17) with a second primary winding, a second secondary winding and a second core (18) forming said second core element, the first primary winding being connected in series with the second primary winding, the first and second transformers (15, 17) being designed to induce magnetic fields in their respective cores (16, 18) with identical field strengths.

3. A power transformer assembly according to claim 2, wherein the first

primary winding and the second primary winding have identical number of turns.

4. A power transformer assembly according to claim 1 , wherein the assembly includes one core, the core including a first core segment (11) forming said first core element and a second core segment (12) forming said second core element. A power transformer assembly according to claim 1 , wherein the core elements are stacked onto each other with common primary and secondary windings. A power transformer assembly according to claim 1 , 2, 4 or 5, wherein the permeabilities and corresponding maximum remanence levels are determined by air gaps in said core elements. A power transformer assembly according to claim 4, wherein the core includes first cuts (21) forming air gaps through the full cross section and second cuts (20) forming air gaps covering parts of the core sectional area. A power transformer assembly according to any of the claims 1 to 7, wherein the second permeability and second remanence are lower than the first permeability and first remanence. A power transformer assembly according to any of the claims 1 to 7, wherein the core elements are toroidal core elements.