| Electronics > Projects, Designs, and Technical Stuff |
| 7kwt LLC 45 khz transformer design issues |
| << < (3/6) > >> |
| axizep:
heres detailed pictures of a transformer and diagrahm |
| T3sl4co1l:
Ok, so you have several turns of the secondary on top of itself? That shields the inner turns from getting the image currents from the primary. Instead, current crowds to the edges. I assume the copper cross section is adequate for the DC requirements? Try using four sections of 1 turn each. That would be below the first primary section and above the last, I guess. Though that still leaves two overlapping turns in the middle. Well, that'll help at least. You could add one more primary section to get four single-turn secondaries. With three primary sections, you won't need as many turns per section, so you might then also consider slightly thinner Litz cable, either wiring the sections in series-parallel (well, in 3 sections, a combination wouldn't work, they'd have to be all series, or all parallel, wouldn't they..), or using pairs of cable in parallel, laying flat (bifilar), basically to give a rectangular cross section to the cable so you maintain a high fill factor while reducing the number of turns in each section. Tim |
| axizep:
sorry for the photoes,because i didnt took a 8+8 turns of primary on camera (i have disassembled transformer right now for the next design to try) i can show how 16+16 primary looked (in parallel) but it might confuse you. |
| axizep:
--- Quote from: T3sl4co1l on April 16, 2019, 09:46:18 am ---Ok, so you have several turns of the secondary on top of itself? That shields the inner turns from getting the image currents from the primary. Instead, current crowds to the edges. I assume the copper cross section is adequate for the DC requirements? Try using four sections of 1 turn each. That would be below the first primary section and above the last, I guess. Though that still leaves two overlapping turns in the middle. Well, that'll help at least. You could add one more primary section to get four single-turn secondaries. With three primary sections, you won't need as many turns per section, so you might then also consider slightly thinner Litz cable, either wiring the sections in series-parallel (well, in 3 sections, a combination wouldn't work, they'd have to be all series, or all parallel, wouldn't they..), or using pairs of cable in parallel, laying flat (bifilar), basically to give a rectangular cross section to the cable so you maintain a high fill factor while reducing the number of turns in each section. Tim --- End quote --- Thank you Tim for your attention! Dc current density for my secondary is 7.5A/mm2 for primary around 5A/mm2 I have secondary 4 turns in total (copper foil strip 0.2 thickness and 53mm height. it has around 70cm length) near the center of that copper strip i have center tap for a syncronus rectifier. secondary coil consists of that 4 turns (2 turn-center tap and another 2 turn) and was allways like that. meanwhile i tried different primary configurations.all of them had 16 turns in total.so transformer=8. I really wonder how infineon design works with 5+5 turns secondary and 100khz + freq. I have doubts that i will fit so many sections as you propose.transformer window is very narrow and allmoust tightly packed by now.also i have to have good insulation because of 750v.so i lay down some insulation sheets of some kind of fibeglass tape. Tim have you tried similiar designs of transformer and similiar power levels? have you encountered those heating problems? what best solution did you found? Did you had problems with a fringing field from an air gap? |
| jbb:
--- Quote from: axizep on April 16, 2019, 07:32:55 am ---...about deviding into two small transformers,that is problematic,i need to keep design's weight and simplicity down. --- End quote --- I respect that. More parts often mean more problems. But you still might find it easier to finish the design with 2 smaller transformers. You'd need half the copper in the secondary windings which makes life easier, and the surface area to volume ratio is higher, which aids cooling. --- Quote from: T3sl4co1l on April 16, 2019, 09:46:18 am ---Ok, so you have several turns of the secondary on top of itself? That shields the inner turns from getting the image currents from the primary. Instead, current crowds to the edges. --- End quote --- Hmm. In general, yes. On the other hand, I think that axizep is using a centre tapped secondary to do half wave rectification. That would mean that on any given half cycle only 2 of the turns will have significant current flow. So not as bad for proximity effect as 4 turns... Fringing flux could still be a b*stard, though. One other thought: the centre termination of the secondary looks a bit messy. It looks like there might be big lumps of solder & Litz wire in the centre tap part. Is this a standard method (I've never done a foil winding...)? Does anyone have other recommendations for terminating foil windings? Ultimately it looks like unless someone with more experience (ahem, Blueskull?) can chip in, you're stuck guessing until you get some more insight into what exactly is generating the losses. |
| Navigation |
| Message Index |
| Next page |
| Previous page |