measuring values of an unknown EE core

hi there, i'm trying to make a 100W sepic converter but the EE core i bought is saturating so i need to measure the permeability to be able to calculate the rest of the numbers.....the problem is that the place where i bought it only has the inductance index (which i forgot which one was mine's from the list :/ ), no brand, no datasheet, no nothing.
I read in a pdf that dividing the inductance index of the core by the value from the "air" core gave the permeability by i got something like 66 which is crazy so something must be wrong.
I wound 100 turns and got 99,2µH, with air and 6559µH with the ferrite core, 9.92nH/N² and 655.9nH/N² respectively and doing the math gave µ = 66 which doesnt make sense at all

The thing is, I have the core, an oscilloscope and a LCR meter, anything i can do to get the permeability or any usefull value to calculate the rest?

thanks in advance

The inductance of the coil increases proportionally to the relative permeability (er) of the core. er is 1 for air. It looks like your core has an er of 66... as you've stated. So either your core actually has a very low er of 66, there's an airgap in your core or your LCR meter is faulty.

If your LCR meter is broken, you can use the oscilloscope to measure the oscillation frequency of the coil connected with a capacitor in parallel. You just need to provide some excitation to the coil/cap (e.g. pulse some current through the coil with a small battery...)

I found a strong tie wrap around the ferrite can make a difference to the core, you only need a tiny gap to change things.

I built a test jig I found on a website for testing saturation, it also tells you some of the cores parameters, I advise that you make one of these, its just a hex inverter and a fet.

"I found a strong tie wrap around the ferrite can make a difference to the core, you only need a tiny gap to change things."

Absolutely true. Been there done that.

"I built a test jig I found on a website for testing saturation, it also tells you some of the cores parameters, I advise that you make one of these, its just a hex inverter and a fet. "

This is great! Could you post a schematic?

laharl2k said:

hi there, i'm trying to make a 100W sepic converter but the EE core i bought is saturating so i need to measure the permeability to be able to calculate the rest of the numbers.....the problem is that the place where i bought it only has the inductance index (which i forgot which one was mine's from the list :/ ), no brand, no datasheet, no nothing.
I read in a pdf that dividing the inductance index of the core by the value from the "air" core gave the permeability by i got something like 66 which is crazy so something must be wrong.
I wound 100 turns and got 99,2µH, with air and 6.559µH with the ferrite core, 9.92nH/N² and 655.9nH/N² respectively and doing the math gave µ = 66 which doesnt make sense at all

The thing is, I have the core, an oscilloscope and a LCR meter, anything i can do to get the permeability or any usefull value to calculate the rest?

thanks in advance

Click to expand...

Hello,

I think you meant 99.2uH with air and 6559uH with the core.

In that case the relative permeability of the core would calculate out to close to 66 that's true.

There are cores with low permeability, and cores with high permeability, so you may just have a low mu core. This can be caused by a gap or by a distributed gap or just the material type itself.

But another possibility is the excitation is just too low to properly test the core material. BH curves dont go straight up through zero they first go out a little and then shoot up suddenly. So the slope starts out rather low, then increases. the 'increased' part is the more typical mu for the core, but to measure that means you've got to force more current through the wire. So you may need a higher test current.
The opposite is also true however, that if the current is too high then the permeability again looks low. So too much current isnt good either.
If you had a way to vary the current throughout the testing, you could start low and then increase it and see how the permeability changes with different levels of excitation. In a DC to DC converter like a buck, this is not too hard to do because you can just increase the load current while watching the scope of the current through the inductor.

The LCR is fine, i've made a 100 turn air coil and then calculated the value and there was only a 10µH of difference due to the imperfections of the coil.

As for the er = 66, isn't that kind of too low for a ferrite? didn't even know there were ferrites with that value

I'll try the oscillation frecuency just to have another reference.

edit:

MrAl said:

Hello,

I think you meant 99.2uH with air and 6559uH with the core.
....
But another possibility is the excitation is just too low to properly test the core material. BH curves dont go straight up through zero they first go out a little and then shoot up suddenly. So the slope starts out rather low, then increases. the 'increased' part is the more typical mu for the core, but to measure that means you've got to force more current through the wire. So you may need a higher test current.
The opposite is also true however, that if the current is too high then the permeability again looks low. So too much current isnt good either.
If you had a way to vary the current throughout the testing, you could start low and then increase it and see how the permeability changes with different levels of excitation. In a DC to DC converter like a buck, this is not too hard to do because you can just increase the load current while watching the scope of the current through the inductor.

Click to expand...

Yeah it's 6.559 MILI H, my bad
I was thinking about that too but then the LCR would only measure the initial inductance as it would be measuring the initial permeability.....maybe DC biasing the core with another coil would work? I suppose inductance and permeability should go up till it starts to saturate no? I have a CC dummy load, i'l give it a try

edit2:
well i tester the LCR on the oscilloscope and it uses a almost 2Vpp sine wave, and also added another coil to the core, 20 or so turns and slowly increased the current, the inductance increased a little to 7,1mH but then started going down again, not sure if this worked.......also added 2 bat85 diodes in parallel just in case to protect the lcr which didn't seem to change the measurement at all
will try doing the same but with 10 or 20 turns to see it behaves

Is core gapped? If it is not it will not likely work for SEPIC switcher. Most EE core ferrites are made of very high permeability mix.

If it is gapped this would show a low effective Er in you winding tests.

laharl2k said:

The LCR is fine, i've made a 100 turn air coil and then calculated the value and there was only a 10µH of difference due to the imperfections of the coil.

As for the er = 66, isn't that kind of too low for a ferrite? didn't even know there were ferrites with that value

I'll try the oscillation frecuency just to have another reference.

edit:


Yeah it's 6.559 MILI H, my bad
I was thinking about that too but then the LCR would only measure the initial inductance as it would be measuring the initial permeability.....maybe DC biasing the core with another coil would work? I suppose inductance and permeability should go up till it starts to saturate no? I have a CC dummy load, i'l give it a try

edit2:
well i tester the LCR on the oscilloscope and it uses a almost 2Vpp sine wave, and also added another coil to the core, 20 or so turns and slowly increased the current, the inductance increased a little to 7,1mH but then started going down again, not sure if this worked.......also added 2 bat85 diodes in parallel just in case to protect the lcr which didn't seem to change the measurement at all
will try doing the same but with 10 or 20 turns to see it behaves

Click to expand...

Hello again,


Well, if you start with a core with one coil and then add another coil you then have a transformer, so the analysis has to change from an inductor to that of a transformer. So if you connect a DC voltage source to the second coil you have effectively shorted out the second winding which affects the first winding too, or at least it should. I believe you would have to use a DC current source, not a voltage source, and you'd want to apply and remove this source carefully if anything else is connected in the circuit like the meter.

The easiest way to do this is to use the coil and core in a Buck type converter circuit. It doesnt have to be a buck with feedback though, just a square wave generator with a driver transistor and an output cap with some resistive load. You'll get about 1/2 the voltage of the power supply voltage at the output, and you can check for di/dt through the inductor and then use v=L*di/dt to test for the inductance. This is especially a good test if you are going to use the coil in a DC to DC converter anyway. The drawback is you need a scope to see the inductor current, using a small value resistor in series with the inductor to measure the current. When you increase the load current (make the output resistance lower) you get more DC current through the coil and thus you can test a lot better than with an inductance meter which may not be able to provide the coil with enough current to properly measure the working inductance.

The pseudo Buck circuit is rather easy to set up even if you dont have a waveform generator (555 and transistor perhaps), so i guess the big question here is, do you have a scope or can buy one or borrow one?