toroid core length

[kinarfi]

I am thinking about winding a coil or two and I have some toroid cores and was wondering if anyone can tell me what the effect of using a longer core vs a short core will have. Intuitively, I suspect that if one core is twice as long as another, it would have twice the inductance, number of turns and wire size being equal, but the wire length would nearly double. I couldn't find on Google, maybe I didn't know the right question.
Kinarfi

 

[cowboybob]

Search Google for toroid core size.

The first two Non-ad listings should help.

 

[jpanhalt]

Inductance is proportional to the square of the number of turns and the core cross-sectional area. See: https://issuu.com/magnetics/docs/powdercoredesignmanual/11 Thus doubling the length of the toroid will double the cross-sectional area and inductance.

Unfortunately, the current format of that site prevents me from copy/pasting it here.

John

 

[ronsimpson]

Thus doubling the length of the toroid will double the cross-sectional area.

Are you sure?
Maybe I don't know what length means. Is it the same as circumference? This would allow for 2x more wire.

 

[ronsimpson]

I went to attach a picture and posted by mistake.

Length? Is that the same a magnetic path length?

 

[alec_t]

Google for "solenoid coil inductance formula". The (magnetic path) length of the coil is the mean circumference.

 

[kinarfi]

Looking at the suggested sites, specifically the attached thumb nail by Ronsimpson, if the length is related to the magnetic path length, then the longer, the core, the lower the inductance. I guess I need to investigate the terms in the formula a little more, but so far, nothing I have read and understood say absolutely what effect the length has.
Kinarfi

 

[ronsimpson]

We started out talking about toroid core length. I use toroid and EE, EI cores. I think I know length.
In post 3# "doubling the length of the toroid will double the cross-sectional area". I just can't see that in any core I have. The cross-sectional area is not effected by length. (length as defined by me)

 

[jpanhalt]

The cross-sectional area is the area of a section of the ring. That is, the cutting plane is perpendicular to the face and passes through the axis of the center of the cylinder. The magnetic path length referred to by ronsimpson is concentric with the circumference of the toroid. The effective magnetic path length can be calculaed as (Source: Arnold Magnetics):

l[SUB]e[/SUB]= pi(od-id)/ln(od/id)

where l[SUB]e[/SUB] is the effective magnetic path length; od is outside diameter; and id is inside diameter.

Thus, I believe my original post is correct.

John

Edit: Here's another reference with pictures. Scroll about half way down. You will notice that inductance is directly proportional to height for toroids with rectangular cross-sections: http://www.nessengr.com/techdata/toroid/toroid.html

 

[kinarfi]

Look at and read my drawing please,

 

[jpanhalt]

Cross-section. It is not concentric with the outer diameter.

John

 

[ronsimpson]

We see the world slightly differently. apples / oranges

You said:"Thus doubling the length of the toroid will double the cross-sectional area" in post 3

I have C cores, (could be called U cores). Putting two together gets a core something like a toroid.
I can get C cores with a cross-section area of 1cm^2 and a length of X.
I can get a C core with the same cross-section area with a length of 2X.
OR
I have EI cores and EE cores where the EE has twice the length as a EI. (approx) Again the cross-section is the same. (For those that don't know, the E of a EI is exactly the same as the E of a EE core.)
OR
Using post 9:
I have a toroid with id=4, id=6, thickness is 1, If cut in half the cross-section is 1x1.
Le=pi(od-id)/ln(od/id)
Le= pi(6-4)/ln(6/4)
Le=pi(2)/ln(1.5)
Le=pi(2)/.4
Le=pi(5)

I have a toroid with id=9, id=11, thickness is 1, If cut in half the cross-section is 1x1.
Le=pi(od-id)/ln(od/id)
Le= pi(11-9)/ln(11/9)
Le=pi(2)/ln(1.222)
Le=pi(2)/.2
Le=pi(10)

If I did the path right both toroid cores have a cross section of 1cm^2.
The length is 2x.
I do not see how having the length go 2X effects the cross section.

 

[kinarfi]

I think I've got it, check this drawing, please.
Another thing I don't understand is what is happening with my self wound inductor.
I started with 46" of ~20 ga wire and measured it's inductance, 1.41 uh, put one pass through the core, measured it, 3.01 uh, another pass,7.77, another pass,15.2, then 25.7, then 39.19 and one more gave me 55.6uh.
When I tried to plug it into a circuit that was working with a 34uh inductor, is wouldn't work anymore. I had tried it with some commercial inductors of 22 uh, 33uh, 44uh and even the 22 and 33 in series and the all worked, but not my self wound, even though it measured similar values.
Did saturate the core? did I use the wrong wire, poorly wound, any ideas? See photo, please.
Thanks for help,
Kinarfi

 

[ronsimpson]

You need to know what core material you are using.
The black core looks like a EMI filter. If so it has hi loss. (low Q) Not used to store energy. It has a huge area, but will saturate.

When you said "length" back in post #1 I bet you are talking about the black core verses the blue core. I was talking about the magnet length or how fare the flux travels. ( 3.14 x diameter)

Your wire has 1.414uH and one turn increases that by about 1.5uH. total = 3uH
Your wire has 1.414uH and 2 turn increases that by about 6.2uH. total = 7.7uH
Your wire has 1.414uH and 4 turn increases that by about 26.2uH. total = 25.7uH

If you look at what the core adds you have a turns squared effect. 1.4 x 4 = 5.6 ( 2^2=4)
6.2 x 4 = 24.8 Looks right to me.

 

[kinarfi]

Thanks, Thanks again, yes the black core was/is an EMI filter and you are correct about length, you did help me to learn more about inductors. They are a wonder, aren't they.
Kinarfi

 

[jpanhalt]

If you are worried about core saturation, here is a link to a nice inductor test bench: http://www.dos4ever.com/flyback/flyback.html#ind2

[edit]Here are two links to Dick Cappel's site with some interesting projects with inductors. one to his whole site and one to a nice oscilloscope probe.

http://www.cappels.org/dproj/Home.htm
http://www.cappels.org/dproj/aciprobe/ACCurrentProbe.html

If you are interested in the test bench, I will post its schematic here. [end edit]

I made a similar device using different drivers to look at core saturation for a project. Given a supply with sufficient current capability, core saturation will occur with fewer turns, not more. I was able to get small rf filters to saturate using just a few turns (5 ro 8 or so). If you have an oscilloscope and small value resistor (non-inductive) you can easily check for saturation.

Without knowing anything about the circuit you are using, it is impossible to say why it is not working with your homemade inductor. Unless you are pushing several amps, core saturation seem unlikely from what I can tell from your pictures. The suggestion that it might be core related is probably more likely.

Please show your circuit and give details like frequency and current.

John

 

[cowboybob]

and one more gave me 55.6uh.
When I tried to plug it into a circuit that was working with a 34uh inductor, is wouldn't work anymore.

(MY EMPHASIS).

Please show your circuit and give details like frequency and current.

Ditto.

Just off the top of my head, if you've replaced the original inductor (34uh), and it was in a resonant circuit, using a 55.6uh inductor (since a commercial replacement 34uh toroid worked), you've now got a circuit that no longer will oscillate (or suppress) at the frequency the circuit was designed for.

Although I may have missed something.

Just a thought...

 

[kinarfi]

I've been working on SMPSs for LEDs for a while and the last post of one of my threads is https://www.electro-tech-online.com/threads/14-to-40-or-50-volt-boost-circuit.124138/#post1035722 and it is the 33 uh inductor in that schematic that I was replacing, I also tried in this circuit, **broken link removed** I prefer to do my inductor testing with this circuit, **broken link removed**, same circuit as in my post, because it is made with inexpensive, easily obtainable parts.
My philosphy is that since LEDs need to be current regulated, if I control the over voltage of the current regulator to minimize it's power consumption, the SMPS will generate the voltage and current needed to drive the LEDs. Current through the LEDs is 956ma., regulated via the 1.3 ohm resistorof the LM317 and/or the 1.3 ohm series resistor in the LT1270 circuit.
Thanks for your help,
Kinarfi

PS I believe Ron is correct about the inductor I wound, "The black core looks like a EMI filter. If so it has hi loss. (low Q) Not used to store energy. It has a huge area, but will saturate." and that's why it failed.

 

[jpanhalt]

I hope you have seen this application note by Jim Williams: https://www.electro-tech-online.com/custompdfs/2012/02/an35f.pdf

He gives a lengthy and practical discussion of inductor selection. While he doesn't discuss Q specifically, he states emphatically the need for good inductors. Based on that need, he decided not to try to make the inductors in house and to use only inductors from outside vendors that are designed for switching regulators. He gives some recommended sources. I have used CoilCraft.

While it might be fun to experiment with making an inductor, for your power supply, I would suggest following William's advice.

John

 

[cowboybob]

Thanks for the clarification.