Help with testing Inductor Saturation current

[BAhmad]

Hi all,

I am trying to figure out a way to test saturation current of DIY inductors , I have various toroidal cores salvaged from broken electronics. I wanted to know if there is some way to find the saturation current. Only thing I can find on the internet about this topic is to use a capacitor bank to pulse current into the inductor and capture the wavefrom with a scope, across a current shunt. See the rough drawing of circuit in attachement.

Using the captured waveform to find the knee point we can estimate the saturation current. But I am not able to get any waveforms similar to those shown in the articles.
Only thing I am doing here differently, is that instead of a proper shunt I am using some 1 ohm power resistors in parallel to get about 100m ohm shunt. The wavefoms, I am getting show a linear rise in current than it flaten, instead of a sharp keen point where the slope of the curve should change.Some times it goes up to 7-8 volts  across 100m ohm so quite high current. In some cases the slope of the wavefomr decreases gradually but nothing remotely close to whats shown in the articles.
 I have tried different size inductors with DC supply up to 20V but, I am unable to get any repeatable results.


here are some of the articles I am referring to.

http://www.vk2zay.net/article/200
http://elm-chan.org/works/lchk/report.html
http://www.dgkelectronics.com/quick-project-inductor-saturation-current-tester/

[ogden]

You shall get/buy some inductor with known specs and use it to test your tester during development.

[eblc1388]

...Some times it goes up to 7-8 volts  across 100m ohm so quite high current. In some cases the slope of the wavefomr decreases gradually but nothing remotely close to whats shown in the articles[/url]

With 7~8V across the 100m ohm current sense resistor, the current must be in the region of 70~80A. There are two problems.

If your MOSFET gate drive is ground referenced, then it is reduced by 7~8V because of the voltage drop on the sense resistor causing the Source voltage to rise. Unless you are driving the MOSFET with 20V pulses, the MOSFET might limit the current to prevent further current increase.

Secondly, 100m ohm is too high value for such current level. Use a 10m ohm sense resistor/shunt instead.

[BAhmad]

...Some times it goes up to 7-8 volts  across 100m ohm so quite high current. In some cases the slope of the wavefomr decreases gradually but nothing remotely close to whats shown in the articles[/url]

If your MOSFET gate drive is ground referenced, then it is reduced by 7~8V because of the voltage drop on the sense resistor causing the Source voltage to rise. Unless you are driving the MOSFET with 20V pulses, the MOSFET might limit the current to prevent further current increase.

It did happen and that point output across shunt was limited to 2-3 volts,
The mosfet driver signal is not ground references , I am using 12-12 isolator brick.

[Mjolinor]

Unless you are doing it as an exercise then why go so complex?

If you just want to know the saturation current then just stick an extra winding on there of one or two turns connected to a scope and connect the inductor to a variac. The trace on the scope will max out when the core is saturated.

[T3sl4co1l]

Without a clamp to absorb the inductor's stored energy, this is a good MOSFET-burner...

A diode from drain to another capacitor is a good way to handle it.  Then a load resistor from capacitor to +V.  Typical values might be 10mF and 20 ohms.

If you aren't getting enough current to saturate the inductor, then use a larger transistor (several IRFZ46N in parallel?) and a smaller resistor.

The version of this I made (years ago), with a single IRFZ46N, did up to 80A peak, which is enough to observe saturation in powdered iron inductors of just a few turns.

There are alternatives.  If you have ferrite cored inductors, the saturation point can be calculated from dimensions.  If you have toroid cores, the core type can usually be looked up.

Tim

[BAhmad]

Unless you are doing it as an exercise then why go so complex?

If you just want to know the saturation current then just stick an extra winding on there of one or two turns connected to a scope and connect the inductor to a variac. The trace on the scope will max out when the core is saturated.

Can you please explain that a bit more , or if you can link to some article or book that'd be great

[Jay_Diddy_B]

Hi,

Have a look at this thread:

https://www.eevblog.com/forum/projects/inductor-saturation-tester-alternative-route-to-dump-the-excess-energy/


There is a good discussion. I designed and built a saturation tester. This is the final schematic:



If you read the thread you will understand why it ended up like this.

Regards,
Jay_Diddy_B

[BAhmad]

If you aren't getting enough current to saturate the inductor, then use a larger transistor (several IRFZ46N in parallel?) and a smaller resistor.

The version of this I made (years ago), with a single IRFZ46N, did up to 80A peak, which is enough to observe saturation in powdered iron inductors of just a few turns.

There are alternatives.  If you have ferrite cored inductors, the saturation point can be calculated from dimensions.  If you have toroid cores, the core type can usually be looked up.

Tim

Thanks, I am actually using one IRFZ46N  ,  and so far I have only lost 3-4 transistors  .
I was testing inductors with quite a lot of turns may be I should lower the no of turns and try again.

as for alternative mathods I can't find any info about these cores , no markings on them and colour code does't make any sense. If you know something about it please tell me, I'd appreciate the help.

(edited)

[Mjolinor]

When the core saturates coupling between windings is maximised so induced voltage cannot increase any more so a sine wave going into the inductor cannot make more voltage in a secondary winding so while the core is not saturated the induced voltage in your small scope winding will increase in size until the core saturates, then it will stop increasing.

You have not really given any indication of sizes so a variac may be over kill. It doesn't matter what you saturate the core with. You can use DC and have two small AC windings for source and load if you want.

Rather than going complicated with someone else's information do some experiments with a bench power supply, windings cores, signal generators and a scope. Learn what is happening and what causes what.

Cores saturating is only the start, once you understand that then recovery times come into play and designing round the saturation.

I have a 100 MHz sine wave generator that will put out around 100 volts at several amps that I made just for looking at coupling of HF in this situation. Dangerous thing

[eblc1388]

It did happen and that point output across shunt was limited to 2-3 volts,

One also need to maintain the voltage across the inductor roughly constant during the measurement, as inductor current will varies with supply voltage too. A current sensing resistor dropping too much voltage(a few volts) will in effect reduce the total voltage seen by the inductor and cause error in inductance calculation.

   

[T3sl4co1l]

as for alternative mathods I can't find any info about these cores , no markings on them and colour code does't make any sense. If you know something about it ....

Don't play the pronoun game, we're not psychic...  

Tim

[David Hess]

When I did this a long time ago, I used feedback to maintain a constant voltage across the inductor removing the effect of the current sense resistor.

If I did it now, I would drive a second winding from an adjustable constant current source and measure the change in inductance directly.

[Zero999]

...Some times it goes up to 7-8 volts  across 100m ohm so quite high current. In some cases the slope of the wavefomr decreases gradually but nothing remotely close to whats shown in the articles[/url]

With 7~8V across the 100m ohm current sense resistor, the current must be in the region of 70~80A. There are two problems.

If your MOSFET gate drive is ground referenced, then it is reduced by 7~8V because of the voltage drop on the sense resistor causing the Source voltage to rise. Unless you are driving the MOSFET with 20V pulses, the MOSFET might limit the current to prevent further current increase.

Secondly, 100m ohm is too high value for such current level. Use a 10m ohm sense resistor/shunt instead.

Don't use a sense resistor. Use a Hall effect transducer or current transformer (watch out for the added inductance through) and put it on the drain side of the MOSFET.