Measurement of Leakage Inductance

Easy peasy said:

air gap does not increase Llk, this is a known physics result.

Click to expand...

Allow me to explain. The leakage is indirectly increased with make up winding gaps to restore the B field caused by adding small gap, not directly from gapped core itself.


> Pulse Tx design for 10kW<Pdesired< 100kW are very different to low power pulse Tx's

Naturally. and a pulse transformer with a flat response over 3 f decades is rare

SunnySkyguy said:

The leakage is indirectly increased with make up winding gaps to restore the B field caused by adding small gap, not directly from gapped core itself.

Click to expand...

Respectfully - the above is not an explanation .... are you saying if you add more turns to the pri and sec the leakage goes up ?

Hi all thanks for the reply

The load is Resistive load 60 Ohms.

Test Pulse: 13us On time (actual operation will use 10us or less)
Test Frequency : 267Hz (250Hz will be used finally)
The Transformer specs are
Primary : 3 Turns
Primary Wire: 6mm2 *2 Teflon Coated Wire
Secondary: 24 Turns
Secondary Wire: 4mm2 Teflon coated wire

The core used is
Toroidal Core: Inner Dia: 8cm, Outer Dia -10 cm , height 3 cm (UMNT-A1008030, details on 2nd page of attached document)
I have tried interleaved and simple winding, and I am getting results of leakage measured lowest through interleaved winding even though the ringing is still high.

The core details are attached.

The IGBTs are connected through copper bars with width of 25mm and thickness of typically 2mm. The length of bus bars is around 30cm. As mentioned in earlier posts by SunnySkyguy and Easy Peasy, lot factors affect parasitics.

However, even though optimization may be done, getting the required parasitic during fabrication of transformer will be difficult due to the limited engineering facility in-house.

So my question is
a. whether the measurements made through LCR are valid, I am seeing options of Ls-Rs, Ls-Rdc, Ls-D, Ls-Q but no Z measurement
b. Which factor can greatly reduce the leakage of the Transformer (Transformer shape (toroidal, U-Core, winding)
c. Any winding structure and drawing for same which can reduce leakage in toroidal/U core
d. The value of frequency used for Transformer power capability is 267Hz or 50kHz in the area product formula for selection of transformer.


My requirement for rise time is less than 1-2us which I am observing through PSIM simulation is achievable by leakage less than 1uH.

Please let me know if any other information is required for making comments

How long are the wires to the load - and their separation ... ?

We usually measure transformer parameters in Ls-Rs setting. I wonder, why you don't measure leakage inductance at the secondary side?

Looking at the waveforms in post #6, we see much stronger oscillations on the secondary side. Considering expectable secondary leakage inductance values, this contradicts the claimed resistive 60 ohms load. Either the resistor is strongly inductive or there is huge wiring inductance.

Hi,

* photos of the complete test arrangement and
* exact part number (or link to datasheet) of the used load resistor could be helpful.

Klaus

The wiring length is about 15-20 cm. Considering the length of wire from IGBT modules to output terminals where the load resistance is connected through 5cm wire on both sides.
Will secondary leakage be different from the turns ratio * primary leakage?.
Is toroidal core suitable or which is other suitable shape.. before optimizing.

Hi

sabu31 said:

The IGBTs are connected through copper bars with width of 25mm and thickness of typically 2mm.

Click to expand...

sabu31 said:

The wiring length is about 15-20 cm.

Click to expand...

for stray inductance the width of a copper bar (cross section) is not much of interest.
A 25x2mm bar will have about the same induactance as a 0.5mm diameter wire.

***
Also the length is no good measure for inductance.

more important is the distance to the return path, or in other words: the enclosd area.
--> a 50cm twisted (signal with it´s return path) wire may have much less inductance than a 10 cm wire wide apart from it´s return path.

**
Many power resistors a "wire wound". They act as a good inductor at higher frequencies. Thus they are not useful for HF measurements.

Klaus

Hi all,

I am attaching the load data, without using a transformer. The load is connected directly to the IGBT Brigdge. I haven't used twisted wire though in both cases (without and with transformer). The resistors are non-inductive type.

I wonder if using Litz wire, suitable for skin depth and overall diameter to reduce wire inductance will help from say 0.8nH/mm down to <0.1 nH/mm

This might not all be relevant but useful to review how low leakage L and high coupling capacitance are related while impedance is sqrt (L/C) with low series R degrading damping all conflicting tradeoffs.

FWIW
Planar types https://coefs.uncc.edu/mnoras/files/2013/03/Transformer-and-Inductor-Design-Handbook_Chapter_20.pdf
Microwave MW types https://www.researchgate.net/public...Magnetron_of_KSTAR_LHCD_Microwave_Application

sabu31 said:

Will secondary leakage be different from the turns ratio * primary leakage?

Click to expand...

Expectable value for symmetrical transformers is (turns ratio)² * primary leakage. Symmetry should be roughly achieved with interleaved windings.

Actual leakage of primary or secondary depends on which windings have one side's winding surface area not surrounded more by the other's. So it's not the mutual flux coupling but the distributed interwinding flux leaked by the exposed surfaces not facing each other. The leakage L/C controls impedance^2 and the LC product controls frequency^2. Examples of asymmetry would be secondary between concentric primary layers.

Improvements in special winding cases with sections that reduce capacitance in series .

I'm curious bout the copper Faraday shield as this is commonly used on HDD rotary servo motors which are essentially high speed pulse rotary angle motors only primary windings over rare earth curved magnets or a "radial servo solenoid" with feedback from the embedded offset bits between sectors.


The leakage inductance is measured by LCR meter (e.g. Hioki 3532) connected across the one winding with the other shorted near frequency of interest then subtract open loop inductance. ( ?)

If you can reduce the area of the copper loop from the driver to the Tx and the Tx sec to the load, either by twisting wires, or at least having the send and return very close together ( watch the insulation ) - this will reduce the effective leakage of the Tx.

As Klaus says - the the resistors are wirewound - they will have appreciable L - which will affect rise time of current ( if not voltage ).

A single pri and single sec is the simplest Tx construction - but also gives the highest leakage, using a toroid is beneficial in that you have more associated area between pri & sec which is what counts for lowering leakage ( this also raises the C between pri and sec - but this cannot be avoided in a simple construction )

Going to interleaved wdgs on the toroid would yield better results.

Thanks for the reply. I have come across some paper where they mentioned use of auxiliary windings to reduce rise time. Will that help. Could you suggest references for transformer construction particularly for pulse transformer

sabu31 said:

Thanks for the reply. I have come across some paper where they mentioned use of auxiliary windings to reduce rise time. Will that help. Could you suggest references for transformer construction particularly for pulse transformer

Click to expand...

Did you read my link on the 1.5MW pulse Xfmr design?

SunnySkyguy said:

Did you read my link on the 1.5MW pulse Xfmr design?

Click to expand...

Hi SunnySkyguy,
I have gone through the paper broadly. But its mentioned that its using PFN network along with a transformer. So is the fast rise time due to PFN or due to transformer.?

Spliithgint he turns between two limbs of the U core will have any benefit over putting all windings on one limb?

sabu31 said:

Hi SunnySkyguy,
I have gone through the paper broadly. But its mentioned that its using PFN network along with a transformer. So is the fast rise time due to PFN or due to transformer.?

Spliithgint he turns between two limbs of the U core will have any benefit over putting all windings on one limb?

Click to expand...

Respectfully - PFN's ( pulse forming networks ) while they do work as intended - can only slow the rise time ( and fall time of a pulse )

there are no networks that can speed up rise time - other than to have the lowest total series inductance in the system ....

You can have a system that shorts the output of a Tx and/or a PFN and then releases it under high current to generate a fast rise time of both V & I

It is possible to get 0.7us x 10 us with 1uH, 1uF LCLCL ladder but not energy efficient.

Thanks, Easy Peasy and SunnySkyguy for the reply. I have a query regarding among the available Toroidal, U Core, E Core which is suited for Lower leakage. I am getting both views that toroidal is good because of the closed magnetic path and E core is good because the number of more number of sides is enclosed in the core.

Also, I have seen that in Pulse Power Transformer by ETH Zurich, they are using winding in both the legs as compared to one leg of U core. Does that make difference if the core is wound on one leg instead of two if the number of layers are similar?

Design procedure for compact pulse transformers with rectangular pulse shape and fast rise times

Microseconds range pulse modulators based on solid state technology often utilize a pulse transformer, since it could offer an inherent current balancing for parallel connected switches and with the turns ratio the modulator design could be adapted to the available semiconductor switch...

ieeexplore.ieee.org

​