Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

How to measure the leakage inductance

Status
Not open for further replies.

ishwaryasampath

Member level 1
Member level 1
Joined
Apr 25, 2013
Messages
37
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,286
Visit site
Activity points
1,506
we want to measure the leakage inductance of transformer @ 57khz but we have a scientific LCRQ bridge 6018 meter that have 100hz/1khz frequency only. so now how we measure the leakage inductance @ 57 khz. rpy me soon vry urgent:-(
 

Why do you mention the LCR bridge? Obviously it can't help with the measurement.

You can refer to generic impedance measurement methods, depending on available instruments and your creativity. You'll surely need a generator and preferably an oscilloscope. A resonance measurement by tuning a series capacitor is probably the most simple method.
 

we measured leakage inductance 52.9 uH @ 1khz and 115.6 uH @ 100hz.so now we need to measure leakage inductance at 57khz. if any formula is there to calculate that
 

leakage inductance Ls,
leakage inductance resistance XLs,
f frequency,
p 3.14,
XLs = 2 x p x f x Ls.
we dont know Ls then how to find XLs
please rpy me as soon as possible
 

As FvM mentioned, it can easily be done, although several calculations are required.

if you have a good sinewave generator, a 1% resistor and a scope, you can arrange a series R-L circuit.

First set the generator at 100 Hz and measure the voltage across the L portion of the circuit.
Now set the generator at 57 Khz, and repeat the measurement.

The resistor in series with the inductor creates a voltage divider with its reactance; and since you already know the value of L at 100 Hz, with a little algebra you can calculate what the reactance is at both 100 and 57000 Hz, and from reactance formula calculate the inductance.

This value will be off somewhat, because it will also include the winding resistance and therefore what you got is a complex impedance. To improve the accuracy of the results, you can also measure the resistance with a good multimeter, -if you null out the probes resistance- and then trigonometrically subtract the resistance from the impedance and obtain pure inductive reactance.
 

The real impedance part measured in a leak inductance measurement (shorted secondary) at 57 kHz is comprised of various components
- primary DC resistance
- primary skin and proximity effect resistance
- transformed secondary DC, skin and proximity resistance
- core losses

For this reason, it should be measured at the specific frequency rather than calculated. The complex impedance can be calculated from three scalar voltage measurements (generator, reference resistor, inductor) or alternatively two complex voltage measurements.

You can reduce the calculation effort my measuring a series resonance frequency, and calculating L from known C and f0, as already suggested.
 

we dont have a function generator thats the problem for us

- - - Updated - - -

we dont have a function generator thats the problem for us

- - - Updated - - -

what is the purpose of SER/PAR button in LCR meter. please refer the attachment pdf of our LCRQ meter.
 

Attachments

  • 105.pdf
    55.2 KB · Views: 287

we dont have a function generator thats the problem for us
I agree.
what is the purpose of SER/PAR button in LCR meter
I'm very sure it's explained in the instrument manual. But even the posted datasheet gives a clear hint by referring to parallel versus series equivalent circuit.
I think series equivalent circuit is appropriate.
 
In order to measure the leakage inductance conveniently, you need an instrument that can measure at your frequency of interest, instead of trying to measure at 1 kHz and apply a formula to get the result at 57 kHz.

The usual way to measure leakage inductance of a transformer is to short circuit one winding and measure inductance at the other winding. Here is a measurement made of a small ferrite cored transformer rated at about 300 watts. I used a Hioki impedance analyzer to make the measurement over a frequency range of 100 Hz to 1 MHz. This image shows the leakage inductance with the vertical scale ranging from 10 µH to 100 µH. You can see that the leakage inductance is nearly constant at about 55 µH up to a little more than 10 kHz where it gradually drops to about 40 µH.

The increase in apparent leakage inductance at the lowest frequency is due to a counter intuitive result of ohmic resistance in the shorted winding. I note that your measured leakage inductance at 100 Hz is substantially larger than at 1 kHz; this is due to this effect.

attachment.php


This image shows the real part (Rs) of the measured impedance (this is the AC resistance) over the same frequency range. Notice that Rs begins to rise beyond 10 kHz; this is due primarily to skin and proximity effect in the windings. Core loss effects are also included.

attachment.php


Here are both images overlaid. You can see that at the same time that increasing skin effect with increasing frequency causes Rs to increase, it also causes the leakage inductance to decrease.

attachment.php


I have measured numerous ferrite cored switcher transformers, and the behavior you see here is typical. If you were to take the leakage inductance you measure at 1 kHz (you got 52.9 µH) and divide by 2 (getting 26.45 µH), your actual leakage inductance at 57 kHz would probably be greater than 26.45 µH and less than 52.9 µH.
 

Attachments

  • Leakage.png
    Leakage.png
    5.6 KB · Views: 384
  • RealPart.png
    RealPart.png
    5.8 KB · Views: 344
  • Both.png
    Both.png
    6.8 KB · Views: 362
  • Like
Reactions: FvM

    FvM

    Points: 2
    Helpful Answer Positive Rating
An obvious conclusion from the impedance curves would be that the transformer isn't actually designed for operation above 20.. 25 kHz where it's copper is more ballast weight than conductor...
 

An obvious conclusion from the impedance curves would be that the transformer isn't actually designed for operation above 20.. 25 kHz where it's copper is more ballast weight than conductor...

Quite so. The transformer whose curves I posted isn't a particularly high performance transformer. It's wound with round wire and is not fully interleaved.

Given the relatively high leakage inductance the OP measured with his transformer I suspected it was probably wound with round wire and not well interleaved. I chose a transformer to illustrate what I thought would be characteristic of the OP's transformer.

Here are a couple of curves of leakage inductance and Rs for a much better transformer wound with copper foil and fully interleaved. Note the scale of the leakage inductance (green) image is now 1 µH to 10 µH.

attachment.php


attachment.php
 

Attachments

  • Leak2.png
    Leak2.png
    6 KB · Views: 321
  • Rs2.png
    Rs2.png
    5.5 KB · Views: 366
  • Like
Reactions: FvM

    FvM

    Points: 2
    Helpful Answer Positive Rating
Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top