Inductor Saturation

Thread Starter

Boondocker

Joined Dec 1, 2006
13
I have an interest to learn more about inductors and their saturation point. So I created an Inductor saturator test circuit based on http://www.vk2zay.net/article/200.

Inductor test circuit.jpg

The diode was removed across the inductor on the breadboard because compared to vk2zay circuit. With it in place I couldn’t generate any change in voltage slope across the shunt R1.

Below are the results. Not seeing a “knee” up change of slope that indicates saturation but a drop-off instead. Does this drop also provide a saturation indication?
Scope 1.jpg

Turns 33
Color Black
V volts 15
kHz 20
t delta t us 3.16
v delta V 0.318
R Resistance 0.47
I Amps I 0.676596
L Inductance 70


Next, I tested an inductor marked 560uH. However, after taking the readings and running the numbers the inductance comes out at 841uH. Did I make an error in the calculation or the readings flawed?
Scope 2.jpg

Purchased Inductor 560 uH
V volts 15
kHz 20
t delta t us 23.5
v delta V 0.197
R Resistance 0.47
I Amps I 0.419149
L Inductance 841
L=V/I*t

I’m a learning hobbyist, and appreciate any comments.

Thanks,

Boondocker
 

jpanhalt

Joined Jan 18, 2008
11,087
I think most of what you are seeing is just ringing (oscillation).

I suggest that you begin first by reading Dekker's very nice description of an inductor test bench. You need to scroll down about half way.

http://www.dos4ever.com/flyback/flyback.html#ind2

Awhile back, I had an interest in this same subject. Here is a link to one post in that thread that shows some results:

http://forum.allaboutcircuits.com/showpost.php?p=67207&postcount=10

This screen shot shows saturation. It shows voltage taken across a current sense resistor. You can see the "knee" as the current takes off. The top trace is the gate drive of the mosfet. The lower trace is voltage across the sense resistor (i.e., inductor current). If you want to see larger images, go the original thread.



For comparison, this shows less or no saturation:



I also suggest that getting a standard coil, as are made for mag-amps, will help you get to a known starting place. Here is where i got mine. I used the G6422-A from Coilcraft:

http://www.coilcraft.com/pdfs/pwrmag.pdf

John
 

Thread Starter

Boondocker

Joined Dec 1, 2006
13
japanhalt,

Thanks for the links. I like how your figures illustrated the discharge of the inductor as well.

I lowered the duty cycle down to 5% (to obtain a 1:20 on/off ratio) on put the diode back in place. It worked to remove the singing oscillation you mentioned. The overall measured inductance of the coil didn’t change.

Still getting a drop off curve instead of ‘knee-up’. Note, when the inductor is removed and the circuit is closed the voltage across the resistor is 200mV.
Inductor scope 3.jpg

Boondocker
 

kubeek

Joined Sep 20, 2005
5,794
Then I would say that you don´t push enough current or charge it long enough to get it into saturation. Is it a ferrite core inductor, or some other type?
 

jpanhalt

Joined Jan 18, 2008
11,087
How much current is your supply capable of?

Do you have any small, probably bare, toroids used for noise suppression? Wrap 5 to 10 turns of insulated heavy wire (16 to 22 awg) on them. Start with a very low duty cycle. Be sure your supply is properly fused. Vary the duty cycle or applied voltage (i.e., current) to see what the curves are like. You are monitoring current as a function of time as you increase the duty cycle/voltage, right?

John
 

Thread Starter

Boondocker

Joined Dec 1, 2006
13
Thanks japanhalt,

I got an inductor to show saturation! As indicated by the ‘knee’ change in slope.

Saturated Inductor.jpg

Made the inductor change you suggested, used a physically smaller inductor. Next, I changed the switch mosfet from an irf510 to a higher power P0N06L. Now that there are more watts, reduced the resistance on the shunt down to 0.094ohms.


Yes indeed, start at a low duty cycle and voltage then adjust to ramp up the power. The watts are beginning to add-up. I’m using a current limiting power supply for fuse protection for now.
 

jpanhalt

Joined Jan 18, 2008
11,087
Looks good to me. I share your joy.

Get one of those calibrated mag-amps from Coilcraft. If need be, tell them you are a Professor at CalTech and need it to demonstrate ;) Then, do the calculations and see how close you come. You will also gain a better appreciation of how mag-amps work.

I used a 0.05 ohm resistor for my current sense. You will see quite an effect from having too much resistance.

John
 

Thread Starter

Boondocker

Joined Dec 1, 2006
13
John,

It is worth the try to make the request for sample inductor. I might have a better chance using my work e-mail and address.

I’ll need to get some lower value resistors. Already have 5- 0.47 ohm resistors in parallel to get down to 0.094 ohms.

Thanks again,

Boondocker
 

jpanhalt

Joined Jan 18, 2008
11,087
John,

It is worth the try to make the request for sample inductor. I might have a better chance using my work e-mail and address.
A yahoo.com or aol.com probably won't work. A work address might. I was kidding about the CalTech part. Although, mine was, in fact, a .edu address, and that really seems to help. It was not so much the cost of the item, but meeting minimums. Once I convinced the lady I was more than willing to pay, it came as a free sample, by FedEx no less. I don't think companies really want to stop giving samples. I think they just want to stop some of the abuses. And, in reality, the free sample was used for education.

My resistor is an Ohmite, CS3, 4-lead version (CS3FR050). Kind of expensive, but they work.

John
 

marshallf3

Joined Jul 26, 2010
2,358
A yahoo.com or aol.com probably won't work. A work address might. Although, mine was, in fact, a .edu address, and that really seems to help.
I don't think companies really want to stop giving samples. I think they just want to stop some of the abuses. And, in reality, the free sample was used for education.
You make some valid points. I rarely use my company e-mail address because I'm probably the most spam free person in the whole organization. It isn't hard to get a website for $10 a month with three domain names and almost unlimited e-mail accounts but don't even go near godaddy or most of the others as you'll end up a prime target for spam.

I wrote to my provider just now as they're always rolling out new deals and if I recall one involved me getting a referral fee credited to my account which I'll gladly split with anyone that signs up.

It's hard to live in this world without a true website that doesn't have popups and all the other crap. When I want a website I want true .ftp interface because its totally directory based and so easy to do. I can write HTML in notepad for all that matters but the free download do Microsoft's Visual Studio Express or even the web developer part of it makes it a piece of cake but somehow takes all of the fun out of it.

I'll know tomorrow (or as I look at the clock later today) as to what deals they're offering. I've got three domains for $10 a month total (including registration) but let's see what the latest deals are, if yours isn't business related it could be a heck of a lot cheaper with all the same access privileges.
 
Top