Transformer leakage inductance

[maicael]

Hello
Is there a way to determine the leakage inductance if a transformer especially if one does not know who wound it.
If I have one before me how would I do it.is it a diref measurement with our inductance meter?
I read some notes but they were too wordy.

 

[KeepItSimpleStupid]

I did a quick search and came up with:

https://en.wikipedia.org/wiki/Leakage_inductance

Looks pretty straight-forward under the heading: Leakage factor and inductance

except for measuring mutual inductance:

https://www.daycounter.com/LabBook/Mutual-Inductance.phtml

So your left with measuring the resistance of the windings which you may be able to get from your inductance meter, otherwise use a 4 terminal method. Measure the current through and the voltage across.

 

[ronsimpson]

If you are just comparing two transformers to see which is coupled better this is what I do.
Short out one of the windings. (use copper braid like solder wick, do a good job of shorting, Low R and Low L)
Measure the inductance of the other winding.
When the "transformer" is shorted out you are left with leakage inductance.

 

[dr pepper]

The above is just what I do, you can do it both ways too, if you use alc bridge the measurement is reasonably close.

 

[maicael]

Ok.
Thanks I get it now.
I also want to know how is it that these sine wave inverter that make use of the leakage inductance of the transformer and just a capacitor at the output.
These trafos are driven by microcontrollers so I wan to better understand the inductance action and how it is that the high frequency pulsing some how allows the leakage inductance to take effect and smoothen the output?
Basically what am asking is there some from if harmonic elimination after the trafo has been pulsed allowing for the inductance to work with the cap?
Is the uController responsible for the harmonic elimination

 

[Les Jones]

Using the leakage inductance will be the same as having an inductor in series with the output. As the reactance of an inductor is proportional to frequency it will filter out the harmonics more than the fundamental. If you then add a capacitor in series with the inductor which resonates at the fundamental frequency the series tuned circuit formed will have a very low impedance at the fundamental frequency.

Les.

 

[maicael]

Using the leakage inductance will be the same as having an inductor in series with the output. As the reactance of an inductor is proportional to frequency it will filter out the harmonics more than the fundamental. If you then add a capacitor in series with the inductor which resonates at the fundamental frequency the series tuned circuit formed will have a very low impedance at the fundamental frequency.

Les.

Thanks but what I want to actually know is if the dead time which I know avoid shoot through in hbridges is also responsible for allowing the leakage inductance to work well as opposed to no dead time in ordinary square wave inverters. What I mean is after the positive cycle pulsing and the dead time is effect, is this when the inductance/capacitor effect takes over judging from what I read about how inductors operate. Hope my question is clear?

 

[ronsimpson]

There are too many type of inverters.
I think one type you talk about uses a 60hz transformer and just sends square waves through. No dead time. +110 volts or -110 volts.
There is another type that sends out 160 volts to make the peak right (also -160V) but also has a "0 volts" time so the average (RMS) works out to 110V. Is this the second type you are talking about?
There is another type that used a 100khz transformer. It uses high frequencies. If it had a 200 volt supply then 50% on/50% off will (after the filter) give you 100 volts. (25/75% will give you 50 volts) Here the 25 in "on time" and the 75% is dead time) By changing the on/dead time you can make a good sign wave. Usually has a micro controller. Is this the one you are talking about?

 

[maicael]

There are too many type of inverters.
I think one type you talk about uses a 60hz transformer and just sends square waves through. No dead time. +110 volts or -110 volts.
There is another type that sends out 160 volts to make the peak right (also -160V) but also has a "0 volts" time so the average (RMS) works out to 110V. Is this the second type you are talking about?
There is another type that used a 100khz transformer. It uses high frequencies. If it had a 200 volt supply then 50% on/50% off will (after the filter) give you 100 volts. (25/75% will give you 50 volts) Here the 25 in "on time" and the 75% is dead time) By changing the on/dead time you can make a good sign wave. Usually has a micro controller. Is this the one you are talking about?

Yes am talking about the one that has the "0 volts" time. Its usually a 50/60hz trafo pulsed at maybe 6-8khz then with a cap at the output filters it.
It seems that the "0 volts" allows for harmonics to be cleared or something like that allowing the inductance to take effect I think?

 

[ronsimpson]

There is one type of inverter where;
The output is 160 volts for two periods of time then 0V for one, -160 for two and 0V for 1.
(160, 160, 0, -160, -160, 0) The peak is 160V and the average is 107. Close the peak and average of a sign wave. Then with some LC filtering to remove the high frequencies harmonics you get close to a sign wave.

 

[dr pepper]

Thats a good trick les, I'm surprised flybacks not commented on that.