Electronics > Beginners

1-2mH 20-30A peak inductor - how would you make that?

(1/5) > >>

not1xor1:
Hi

I'm playing with LTSPICE and simulating a linear PSU pre-regulator that approaches the efficiency of a switching one while keeping the noise at the level and frequency of an ordinary linear PSU.

Unfortunately that would require an inductor in the range of the mH able to sustain tenths of Amps of peak current without saturating.

Since I never saw such a beast I wonder if that would be just too expensive compared to other possible alternatives (ordinary switching preregulator or multitap transformer).

Would a 50/60Hz toroidal transformer nucleus work for that? Any idea about size calculation?
thanks

Zero999:
Presumably this is a low frequency switched mode?

What about audible noise?

Yes, I expect a toroidal mains transformer core will work quite well.

Berni:
By itself 1mH is not that difficult, but yes doing that at 30A is going to require a very big inductor. Im guessing such a torodial core inductor would be about 5 to 10cm in diameter.

This is a pretty useful calculator for designing inductors and transformers: http://dicks-website.eu/coilcalculator/index.html (Note that you will likely have to go find a core you want at a manufacturer and enter the data from its datasheet)

The type of core also determines how well the inductor works at high frequency and the kind of losses it creates there. Cores such as silicon steel is only good for low frequency 50Hz operation. Ferrite cores handle higher frequencies and come in many blends that have different properties, some handle high fields (higher saturation current) some have high permeability (less turns needed), some operate well in >MHz ranges, some are very temperature stable etc. Typically a core optimized for one parameter is bad at all the other parameters.

not1xor1:

--- Quote from: Zero999 on January 23, 2019, 08:39:39 am ---Presumably this is a low frequency switched mode?

What about audible noise?

Yes, I expect a toroidal mains transformer core will work quite well.

--- End quote ---

It is a sort of hybrid between a classical SCR preregulator and a classical buck regulator.
There is a MOSFET switch that goes on after 2-3ms from zero-cross and goes off as soon as the inductor has enough energy to re-charge the output capacitor for the worst case load. A 1000-2000µF capacitor, before the MOSFET, dumps the voltage spikes and store the energy released by the transformer secondary. Obviously on low current load it is the capacitor which provides most of the energy for the inductor, so in that case it is a real, albeit quite lazy, DCM buck converter.  :D

I've no idea about acoustical noise, in LTspice it is quite quiet  ;D. Anyway current value and frequency are about the same that go through an ordinary transformer secondary when charging the electrolytic capacitors, so I hope that would not be much more than usual.

not1xor1:

--- Quote from: Berni on January 23, 2019, 08:46:09 am ---By itself 1mH is not that difficult, but yes doing that at 30A is going to require a very big inductor. Im guessing such a torodial core inductor would be about 5 to 10cm in diameter.

This is a pretty useful calculator for designing inductors and transformers: http://dicks-website.eu/coilcalculator/index.html (Note that you will likely have to go find a core you want at a manufacturer and enter the data from its datasheet)

The type of core also determines how well the inductor works at high frequency and the kind of losses it creates there. Cores such as silicon steel is only good for low frequency 50Hz operation. Ferrite cores handle higher frequencies and come in many blends that have different properties, some handle high fields (higher saturation current) some have high permeability (less turns needed), some operate well in >MHz ranges, some are very temperature stable etc. Typically a core optimized for one parameter is bad at all the other parameters.

--- End quote ---

Thanks for the link. I'll read carefully that page later.
In the simulator there are current pulses ranging from fractions of ms to several ms (according to the voltage and load) at a frequency of 100Hz (120Hz). So I think that an AC mains transformer core would be more appropriate.

I'm not an engineer and just studied electronics by myself on schoolbooks lot of years ago. My knowledge on electromagnetics is... just appropriately for an iron core... quite rusted  :D
I hope I'll be able to digest all that informations...  :)

thanks

Navigation

[0] Message Index

[#] Next page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod