PFC inductor advice
I did a Power Factor correction circuit using the on-line design utility from IRF.com.
I am using IR1150S chip and now my prototype is working, my goal is to obtain 1kw max. and for now I use a T130-26 toroid core stolen from a damaged PC smps for the boost inductor(too small) wound to obtain the required 270uH inductance.
As I did not find too much information about the type of toroid needed, I would like to know what kind of core material is better for this application, I am looking at 26 or 52 material types and maybe a T157 size would be OK.
Someone can give me advice on that?
I enclose my actual schematic.
The inductor must be chosen so that it will not saturate at the maximum current it will be subjected to. The max current will set the 1/2 LI^2 (energy) capability of the inductor. The peak inductor current is pretty easy to calculate, but I don't have the formulas handy.
I would also recommend that the output caps be rated at a minimum of 250 volts, 300 would be better and 200 in unacceptable.
Consider either MPP toroids (see Arnold Magnetics and others), RF iron powder toroids (material 2 is the most popular, see www.micrometals.com ) or gapped ferrite and litz wire.
Use the search engine, there are some posts on this subject.
I will try with iron powder material 2, it is easier to get for me.
Hi Rick, you are right about the capacitor voltage, I am using 250v now.:)
Actually, if the frequency is not too,high, Micrometals mix 18 is easier to work with, being 55 perm material. It's a good compromise between permeability (to keep the number of turns down) and low loss. Micrometals #2 is low loss, but the low perm will require you to lay down a lot of turns to get the required inductance. I've been using mix 18 E cores in my 1150 PFC, running at around 80 kHz.. They worked really well, but I replaced them with Cool-Mu/Sendust E-cores for higher efficiency (~ 1/2 point better). However, if you can't get Cool-Mu, Micrometals mix 18 is a nice second choice. Micrometals mix 26 is far too lossy to use for a PFC choke, though I guess you had to start with something....
Mix 52 is lower loss, but can have long-term degredation problems that take months to surface. A design that starts out cold melts down after running a few months. This problem appears to be confined to E cores, and may be a peculiar artifact of the way the flux turns the corners in the core, or it may also be the way the iron particles and binder distribute themselves when the core is pressed. A failed core when sectioned shows a distinct pattern of discoloration and deterioration at the corners. Mix 18 doesn't have the problem, and is lower loss.
I have also found that using an E core PFC choke resuts in a much cleaner leading edge current waveform due to the lower distributed capacitance in the winding. This is shown by both performance and actual impedance plots. My summary - Micrometals Mix 18, E cores.
Aren't stray fields stronger with iron powder E cores too? Then again, it's like taking a toroid and concentrating all the turns (and the RF losses) in 1/3 of the length of the core while leaving the other 2/3 unused.
I prefer toroids for inductors. If a single core requires too many turns or losses are too high, just stack more cores... ;)
The E core I used was small (EI-375 size or EI-21 size) and the bobbin full. The toroid (T-130) had two layers of wire.It appears that the coupling effect of the toroidal core between all the inner turns may be a factor. The impedance plot for the toroid shows more capacitance and an additional high frequency resonance.
I fully realize that the concentrated turns on the E core result in more spray flux, but a shorted turn of foil around the outside of the core helps to tame this and also allows one to ground the core so that it doesn't act as an antenna. Just like anything else with switching power supplies, you take with one hand, and have to give a little something back with the other...
I'm currently optimizing EMI for my PFC, and I'm not seeing a lot of switching frequency harmonics show up like they would if the choke stray flux were coupling to the EMI filter. Actually, the 150kHz to ~4MHz EMI is shockingly well-mannered.
I preferred toroids, too, until I saw these results. Using an unsnubbed output diode, the leading edge current waveform went from intense underdamped ringing with the toroidal inductor to a clean, relatively small recovery spike (some of which may be capacitive) with just a small, small touch of damped ringing using the E core (something that could be cleaned up with a simple RC snubber). These results were obtained using the new Q-Speed diodes touted as SiC replacements. The ST PFC diodes are almost as good-looking, but the Q-Speed diode is more efficient. I'll be trying some new IR./Vishay diodes when I get the chance.
pfc inductor fflyback inductor
isn't this timely.
i have been fighting with my plasma emiting pfc. i have been using a ferrite. i guess i will change that
i found a huge problem with my other plasma emitting smp's when the current got too high. i had a bunch of so called iron powder toroids that are really ferrites.
so i cant wait to get back to the "lab" and try the mpp and iron toriods that my wife said had come.
so i hope this will solve my problems.
secondly what would be the best to use for flybacks ferrite or iron/mpp
i guess i didn't know what the cause and failure mode was supposed to be. an kept trying larger mosfets and dealing with crazy wave forms as the current started getting up...
We use this chip in our professional amplifiers and it has serious turn on issues. We also use #18 cores from Micrometals which work well but DO NOT use a core which is too small. Our smallest size is a T175 and MMetals offer a T-200B which is a double stocked T200 core.
The problem with the IR1150S is that on turn on there is not enough dead time and the power device will blow. This is more apparant at 240v AC. We do not use a MOSFETS as IGBTs work so much better. I designed an extra circuit to deal with the PFC on turn on. The input surge issue is now a thing of the past.
IR are designing a new chip which they say will take care of the inrush current issue.
Zed Audio Corp.
with the pfc inductor should it be wound with multiple wire in parallel.
or can it be a solid wire like the out put inductor of an isolated smps?
|All times are GMT. The time now is 03:46 PM.|
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2015 DragonByte Technologies Ltd.
Copyright ©1999-2015 diyAudio