Electronics > Projects, Designs, and Technical Stuff
7kwt LLC 45 khz transformer design issues
axizep:
Hello dear engineering friends and Dave himself ;D
Iam designing 7kwt dc-dc converter that converts 750v DC into a 48V DC (140A) using LLC topology.working frequency is 45khz when at full load and up to 200khz when idle.
and the problem is with transformer design.i have tried 6 different transformer constructions and all of them heat up very fast and severe.
All other components of the converter are at 50C,which is very fine as for me. transformer gets hot up to 70-80C on the surface of the winding just in 10 minutes under full load.
My first transformer designs where so called integrated designs,where resonant inductor was incorporated in a transformer using leakage inductance between primary and secondary.that design showed rapid heat build up on the surface of the primary which was facing secondary (primary and secondary where separated in space,thus big leakage inductance present). then i have read about proximity effect and made a conclusion that i have to make separate external resonant inductor and transformer should have lowest value leakage inductance as possible,because if i got it right,leakage inductance makes eddy currents and thus heating nearest layers of the winding leaving all the rest layers cold.
All the transformer designs had air gap at the ferrite core (1mm). Main core is EE85 big peace of ferrite (860mm2 Ae).
So i have proceeded to a foil secondary winding and finer litz wire (600x0.1 instead of 0.46x32). secondary winding is center tapped (half bridge synchronous rectifier) 2+2 turns 0.2mm foil 53mm wide.10mm2 for a 140 and 50% duty cycle
and primary winding is 16 turns (8+8 in series) 0.1x600 litz wire,half of primary is outside of secondary and half of it is inside in it.
that kind of design i have looked up in the infineon evaluation board of a 3 kwt llc converter appnote ( https://www.infineon.com/dgdl/Infineon-ApplicationNote_Evaluationboard_3kW_dual_phase_LLC-AN-v01_00-EN.pdf?fileId=5546d462580663ef01582eb629b70118 ) you can check it on the 25 page.
infineon design has even more secondary turns with a foil (5+5), it also has air gaps on all three limbs of ferrite,it also has external resonant inductor. but somehow it work on a fantastic 100khz+ frequency,where all the skin effect and proximity effect are very limiting.
i have used spacer to pull inner layer of a primary on a bobbin away from a fringing air gap field. but that doesnt helped alot.transformer runs very hot,isolation start to boil after some working time.
iam so exhausted from all those failed experiements,that i gave up and cant go on without a help. i have been struggling with this project for 13 months by now,half of it with a transformer.
If there was some power electronics engineer among you that is familiar with a magnetics and transformer design i would really appreciate to get some help.
Thank you for your patience,there was lots of words here to read (and sorry for my english). ::)
P.S. here some photos of my beast and transformer itself
MagicSmoker:
Are you interleaving the primary and secondary? Even just splitting the primary into two halves with the secondary in between dramatically reduces proximity losses. Also, note that even just a few turns on a new layer counts as a complete layer with respect to proximity loss. It is often the case that dropping down 1-2 gauges in wire size to squeeze all the turns into fewer layers results in lower total losses, even though the DC resistance of the winding increased.
TimNJ:
I'm having a hard time determining...
1.) What you think the problem might be related to
2.) What you've tried so far, and what exactly the results have been
Also, if you have any spreadsheets or design work that shows how/why you selected the current the design, that might be helpful. Maybe a transformer drawing?
For now, just some thoughts.
1.) Peak flux density? Any way you are saturating it? Maybe need more turns and a bigger air gap? (I don't think saturating LLC transformers is usually much of an issue, but it is definitely possible)
2.) Did you do any calculations to determine if 0.2mm foil was optimal? (http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.916.5488&rep=rep1&type=pdf)
axizep:
--- Quote from: MagicSmoker on April 15, 2019, 10:04:09 pm ---Are you interleaving the primary and secondary? Even just splitting the primary into two halves with the secondary in between dramatically reduces proximity losses. Also, note that even just a few turns on a new layer counts as a complete layer with respect to proximity loss. It is often the case that dropping down 1-2 gauges in wire size to squeeze all the turns into fewer layers results in lower total losses, even though the DC resistance of the winding increased.
--- End quote ---
to be precise,what i have done is: i have tried 2 different methods (dont know which one could be called interleaving,but i guess one of them could be interleaving)
my primary turns number in both cases was 16 turns.
in case 1 i have winded 16 turns of litz wire on a bobbin,then i have wounded 2+2turns (all the secondary foil turns) and ontop of the secondary another 16 turns of primary. both inner 16 turns and outer 16 turn where paralelled otside the transformer.
in second case i have winded 8 turns (half of a primary) then a whole 2+2 turns of foil secondary then on top of a secondary the rest of a primary (last 8 turns).
axizep:
--- Quote from: TimNJ on April 15, 2019, 10:11:26 pm ---I'm having a hard time determining...
1.) What you think the problem might be related to
2.) What you've tried so far, and what exactly the results have been
Also, if you have any spreadsheets or design work that shows how/why you selected the current the design, that might be helpful. Maybe a transformer drawing?
For now, just some thoughts.
1.) Peak flux density? Any way you are saturating it? Maybe need more turns and a bigger air gap? (I don't think saturating LLC transformers is usually much of an issue, but it is definitely possible)
2.) Did you do any calculations to determine if 0.2mm foil was optimal? (http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.916.5488&rep=rep1&type=pdf)
--- End quote ---
1)peak flux density is around 150mTl far away from saturation. ferrite is only slightly warm.
2)about 0.2 foil thickness- i have tried 0,3mm also that didnt helped much. i wanted to try 0.2 as i thought it impedance could be smaller than thicker one because of proximity effect. but it didnt show any effect.
a similiar transformer drawing could be seen on the infineon appnote of a 3kwt converter on a 25 page.
i have tried many designs,but none of them where cold,so iam asking if somebody knows how to design it right. basically i need turns ratio=8 and center tapped secondary (for a half bridge rectifier).
sorry your link says No document with DOI "10.1.1.916.5488"
Navigation
[0] Message Index
[#] Next page
Go to full version