Electronics > Projects, Designs, and Technical Stuff
Characterizing magnetic cores
NiHaoMike:
I have a few ferrite cores from old monitor deflection coils and I'm looking to reuse them as inductor cores for power supply circuits.
How would I go about characterizing them? One thing I did was to wind 7 or so turns on one of the cores, connect it in parallel with a capacitor, and use a pulse to excite it, at which point the resonant frequency (and thus the inductance) can be measured. It measured about 29uH. But that only tells a part of the story. Is there an easy way to figure out how many A*N (amp*turns) the core can withstand without saturating?
Psi:
I have the exact same problem, lots of cores and no idea what frequency they are good for.
At one point i planned to build a testing rig to scan frequency for a few different turn numbers and attempt to work out the core parameters.
So far i've not had time to build one though, i figured it was probably more cost effetive to just buy the core type i needed than try to build something to reuse old cores.
jahonen:
You could build a suitable setup for driving a pulse into an inductor and simultaneously measuring the current, that gives you saturation characteristic. I built something like that a while ago. Losses are more difficult, though.
Regards,
Janne
tecman:
Simple tests can be made by winding some turns, and using a signal generator and a scope or AC voltmeter that will measure the frequency, and a couple of resistors, you can approximate the inductance by measuring Xl of the coil. You can also predict the saturation by increasing the drive level until you see saturation. A larger number of turns would be better to see saturation. Some calculation is required to determine the saturation level.
paul
scrat:
I agree with Janne.
If you're building a switching converter, you can just use one of its switches and an MCU to generate the pulse and change its duration live (if you have a debugger). Starting from a low duration makes the test safe.
Navigation
[0] Message Index
[#] Next page
Go to full version