| Electronics > Beginners |
| Core material selection for Half-Bridge converter |
| (1/2) > >> |
| 09d08:
Hello, I am trying to design a half-bridge converter to step up 12V to 310V. The switching frequency I chose is 80KHz and the half-bridge circuit I've put together using a breadboard and a SG3525 seems to work well. I have two toroidal cores available, one is made of J Material(5000µ) and the other one is made of #26 material (75µ). How can I determine which one is more suitable for my use case? |
| MagicSmoker:
--- Quote from: 09d08 on June 14, 2019, 12:05:57 pm ---...I have two toroidal cores available, one is made of J Material(5000µ) and the other one is made of #26 material (75µ). --- End quote --- Neither material is good for this application, unfortunately. Powdered iron (#26) will have very high losses at 80kHz with any significant AC flux swing and the distributed air gap gives the material too low a permeability for use in a half-bridge/forward-derived converter as well. J material is a ferrite intended for EMI filters and RF transformers. It is not even characterized for power applications, though of the two choices it is the least bad one. |
| T3sl4co1l:
The above is correct; just to add some information: For a forward converter, the transformer should have high permeability. This rules out #26 anyway. (#26 would possibly be usable for the accompanying filter choke---you do have a filter choke, right?) I've used W material before, at up to 0.3T at lower frequencies (20kHz), and didn't get excessive heating. At 80kHz I would expect 0.1T is still adequate. The W material shows mu'' dominating over 100kHz, which if the mu(freq) curve still holds at large signal levels, suggests the magnetizing current is mostly power loss. But because |mu| is still quite high (~10k), magnetizing current is low, and so is power dissipation. The situation will be better for J material, the lower mu having a higher cutoff frequency, and so allowing somewhat higher flux density for the same total core loss. But keeping it to 0.1T should still be very safe. For a flyback, or for the forward's filter choke, neither material is very good. A solid ferrite toroid is a non-starter: the mu must be low, because the core must store energy. Energy storage is inversely proportional to mu. #26 also has a frequency cutoff around 50kHz, so it's basically all loss at higher frequencies. It can still be used there, but the ripple fraction must be kept very low so that there's more DC output than AC switching ripple. Same as with the transformer---the (AC) magnetizing current is basically all loss so you need to use lots of turns (lots of inductance, giving a low ripple fraction, and low delta B) to keep loss down. Preferably, use better core material, like #8, or a more expensive mix like Sendust/Kool-Mu, or a ferrite with air gap. Tim |
| 09d08:
Thanks for the informative answers, I'm really new to magnetics. I assume I need a ferrite core instead of an iron powder one since a half-bridge converter doesn't use the transformer to store energy. I know most manufacturers specify the intended applications for their core materials but what should I exactly do to determine which material is best suited for my application? Which properties of the materials should I check? |
| T3sl4co1l:
Magnetics type P, Fair-Rite #77, TDK/EPCOS N87, Ferroxcube 3C90 and so on are very typical power transformer ferrites. mu ~ 2000, Bsat > 0.3T, and moderate to low losses, are typical of these materials. You're using a filter inductor, right? Tim |
| Navigation |
| Message Index |
| Next page |