Testing a ferrite transformer

If you have a ferrite transformer and you know its dimensions but ferrite core type is unknown, what would you do to test it?

It has no winding ,you will wind it !!

For testing it's permeability, you can just wind a few turns and measure inductance, then back calculate reluctance and permeability. Saturation flux density isn't difficult to measure. You want to switch it to a low impedance DC source, and measure the current waveform. By looking at where di/dt starts to increase sharply you measure get the saturation current, and you can back calculate saturation flux density from that. For that measurement I use a DIY spot welder which is basically a big capacitor bank that discharges through the inductor, and is triggered by a big SCR.

After that, things like core loss and the specific shape of the BH curve are pretty tricky and require sophisticated instruments.

Thank you for your reply.

First part about permeability OK i can understand that,
L=Uo Ur n^2 A /l

as Uo=4pi*10^-7
Ur=core permeability
n=number of turns
A= cross section area
l=winding length
So measuring L , knowing core dimensions ,n and l i can get Ur
(ignoring wire resistance).


But the second part,saturation current the circuit you suggest if i understand looks like the attached circuit.

Now what am i to change ,the firing rate? or the voltage the capacitor bank is charged to? because this di/dt is not clear

May i use a pulse generator followed by a power transformer instead? i know that capacitor bank can supply huge current instantaneously but the transformer i have is a small one.

May i use a pulse generator followed by a power transformer /(transistor not transformer)
May i use a pulse generator followed by a power transistor

koshary said:

Thank you for your reply.

First part about permeability OK i can understand that,
L=Uo Ur n^2 A /l

as Uo=4pi*10^-7
Ur=core permeability
n=number of turns
A= cross section area
l=winding length
So measuring L , knowing core dimensions ,n and l i can get Ur
(ignoring wire resistance).

Click to expand...

Yes, but keep in mind that cores have "effective" magnetic lengths and areas which are somewhat different from the geometric length and area. For common core shapes (toroid, E core, etc) you can find the formulas for effective area and length.

But the second part,saturation current the circuit you suggest if i understand looks like the attached circuit.

Now what am i to change ,the firing rate? or the voltage the capacitor bank is charged to? because this di/dt is not clear

Click to expand...

For an accurate test it's desirable for the DC voltage applied to the inductor to be constant during the pulse, therefore you want as low of an impedance as possible. That's why I use a broadband current transformer or a very low value shunt resistor for current measurement. Also you want the capacitor bank to store much more energy than what you think will saturate the transformer, so that the voltage will not sag before saturation occurs. But you don't want so much that it burns out the transformer or the SCR.

But yes, that circuit should work, though I would put the sense resistor at the cathode of the SCR.

For example, I recently used this setup to test some big inductors I made, which were 400uH and were designed to handle 100A peak. For that I charged about 50,000uF of capacitance to 30V and dump the entire thing into the inductor. The peak current is several hundred amps, and is easily enough energy to make sparks and spot weld. On the scope, I'm looking at the inductor current. At the beginning of the pulse, it will be climbing at about 75,000amps per second, as a 400uH inductor should when driven with 30V. But at some point, the slope of the current will increase, meaning that the core is saturating and inductance is dropping. Depending on the type of core material, the saturation can be very abrupt and fast, or it can be soft and slow. You just kind of have to make a judgement call on where "saturation" really occurs.

May i use a pulse generator followed by a power transformer instead? i know that capacitor bank can supply huge current instantaneously but the transformer i have is a small one.

Click to expand...

It might work if the impedance of the pulse generator is low enough to not droop when the transformer starts saturating. Driving a power transistor can work too; it's not much different from the SCR.

Thank you very much for your help,helpful post i did click

by measuring the inductance L for certain number of turns on this transformer with the help of inductance meter, you can determine its Al value (L=N*N *Al * 10E-9) which is the inductance index. and once its Al value is determined you don't need inductance meter any more. then you can easily calculate inductance for any turns using the formula L=N*N *Al *10E-9 . where N is required number of turns.