Leakage inductance value is realistic?

treez · 16th September 2012, 07:53 PM

Hello,

I want to modify this 198W flyback (vin = 90-270VAC) to power a 190W Class D amplifier........(with rails of +/-40V)

http://www.fairchildsemi.com/ds/FS/FSCQ1565RT.pdf
(page 39)

Do you think that their leakage inductance value on page 40 of just 0.5% of primary inductance is realistic for a single sandwich wound transformer?

Terry Given · 17th September 2012, 12:23 AM

Treez,
10uH/210uH = 4.76% which on the surface seems pretty high. note that the transformer windings are interleaved (is that what you mean by single sandwich wound).

its not necessarily meaningful to look at % leakage in a flyback, as leakage is uncoupled inductance and doesnt involve the core, whereas magnetising inductance involves the core. so as core gap increases (eg higher xfmr power density) then Lmag decreases but L_leak stays the same, so the % leakage gets worse. that being said I always do look at % leakage.....

and while 5% leakage sounds extraordinarily high, it can be difficult in practice to do better than this. dont forget that for a "real" leakage inductance measurement place the xfmr in the target PCB and short out the secondary diodes & caps. this makes it a lot worse, but the loop inductance on the secondary side adds to the xfmr secondary leakage, and then reflects back to the primary.

I have a design with Lm = 430uH, L_leak = 3uH, n = 9. so 12nH of loop inductance in the secondary adds 1uH to the primary leakage. and when the xfmr manufacturer measures leakage by "shorting" the secondary with a clip lead, measured leakage is more like 8uH

treez · 17th September 2012, 05:29 AM

Thanks Terry,

Woop my maths was out by a factor of 10........that really is amazing that somebody would attempt a 198W flyback with 5% leakage as Fairchilsemi have done........that leakage inductance must ring like heck with the 1nF drain capacitor at switch off......theres very little there to dissipate this ringing aswell.

I am now wondering of Fairchildsemi are being real with this 198W offline flyback design?

Terry Given · 17th September 2012, 07:51 AM

no worries.

ignoring the FET & xfmr capacitance, it will ring at 1.6MHz. I dont know the other capacitances, but its fair to say the ringing frequency will be 1 ~ 1.5MHz. I was just about to wonder what the primary winding Fr ratio is at this frequency, when I noticed they are using Litz wire (20 strands 0.08mm diameter). skin depth at 1.5MHz is .05mm, at 1MHz its 0.066mm. the transformer drawing doesnt show the number of layers, but it looks like each half winding fits on a single layer. But when ringing there are still 2 layers (no energy transfer to secondary so interleaving doesnt allow this to be halved), so Fr_ring will still be fairly high.

this means the primary winding will suck up (damp) some of the ringing. If they didnt use Litz the transformer would probably cook. and the drain voltage will be a hairy mess, thats for sure. E fields galore!

ultimately as long as they can suck the heat out of the transformer then who cares. but unless you really know what you are doing, this is a brave approach.

oh, and I should add: in a flyback, %leakage = %input power stored in leakage inductance, that the primary winding has to deal with (or the drain clamp if any - but RCD or Zener drain clamps always dissipate m*P_leakage where m > 1 (m=3 is a good guess)

16th September 2012, 07:53 PM	#1
treez Banned Join Date: Sep 2012	Leakage inductance value is realistic? Hello, I want to modify this 198W flyback (vin = 90-270VAC) to power a 190W Class D amplifier........(with rails of +/-40V) http://www.fairchildsemi.com/ds/FS/FSCQ1565RT.pdf (page 39) Do you think that their leakage inductance value on page 40 of just 0.5% of primary inductance is realistic for a single sandwich wound transformer?

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
!!DC Leakage!!	amadork9	Solid State	4	25th December 2010 06:34 PM
is this glue or leakage?	bobbyrae	Parts	6	2nd July 2010 12:07 PM
How to measure leakage inductance	percy	Power Supplies	17	29th August 2008 09:57 PM
Electrolyte Leakage?	TrevorK	Parts	3	6th July 2006 03:44 PM
Designing transformer leakage inductance	DaBit	Power Supplies	3	17th January 2006 08:08 PM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

17th September 2012, 12:23 AM	#2
Terry Given diyAudio Member Join Date: Mar 2008	Treez, 10uH/210uH = 4.76% which on the surface seems pretty high. note that the transformer windings are interleaved (is that what you mean by single sandwich wound). its not necessarily meaningful to look at % leakage in a flyback, as leakage is uncoupled inductance and doesnt involve the core, whereas magnetising inductance involves the core. so as core gap increases (eg higher xfmr power density) then Lmag decreases but L_leak stays the same, so the % leakage gets worse. that being said I always do look at % leakage..... and while 5% leakage sounds extraordinarily high, it can be difficult in practice to do better than this. dont forget that for a "real" leakage inductance measurement place the xfmr in the target PCB and short out the secondary diodes & caps. this makes it a lot worse, but the loop inductance on the secondary side adds to the xfmr secondary leakage, and then reflects back to the primary. I have a design with Lm = 430uH, L_leak = 3uH, n = 9. so 12nH of loop inductance in the secondary adds 1uH to the primary leakage. and when the xfmr manufacturer measures leakage by "shorting" the secondary with a clip lead, measured leakage is more like 8uH

17th September 2012, 05:29 AM	#3
treez Banned Join Date: Sep 2012	Thanks Terry, Woop my maths was out by a factor of 10........that really is amazing that somebody would attempt a 198W flyback with 5% leakage as Fairchilsemi have done........that leakage inductance must ring like heck with the 1nF drain capacitor at switch off......theres very little there to dissipate this ringing aswell. I am now wondering of Fairchildsemi are being real with this 198W offline flyback design?

17th September 2012, 07:51 AM	#4
Terry Given diyAudio Member Join Date: Mar 2008	no worries. ignoring the FET & xfmr capacitance, it will ring at 1.6MHz. I dont know the other capacitances, but its fair to say the ringing frequency will be 1 ~ 1.5MHz. I was just about to wonder what the primary winding Fr ratio is at this frequency, when I noticed they are using Litz wire (20 strands 0.08mm diameter). skin depth at 1.5MHz is .05mm, at 1MHz its 0.066mm. the transformer drawing doesnt show the number of layers, but it looks like each half winding fits on a single layer. But when ringing there are still 2 layers (no energy transfer to secondary so interleaving doesnt allow this to be halved), so Fr_ring will still be fairly high. this means the primary winding will suck up (damp) some of the ringing. If they didnt use Litz the transformer would probably cook. and the drain voltage will be a hairy mess, thats for sure. E fields galore! ultimately as long as they can suck the heat out of the transformer then who cares. but unless you really know what you are doing, this is a brave approach. oh, and I should add: in a flyback, %leakage = %input power stored in leakage inductance, that the primary winding has to deal with (or the drain clamp if any - but RCD or Zener drain clamps always dissipate m*P_leakage where m > 1 (m=3 is a good guess)