Is it possible to make air gap in pot-core which doesn't have it?

[ChildOfVision]

Hi!
I'm interested is it possible to make/add air-gap in pot-core (2 halves) which doesn't have it? On this forum I only found that this is, allegedly, easy to obtain.
So, does somebody know how to do this - if I add some thin/thick insulator between central pillars, than I will get loose contact between outer rings?!

Also, if somebody knows anything about an ancient pot-core inductor, please let me know! This is all I have: Outer diameter = 18mm, height (both halves assembled) = 18mm (approx.), Manufacturer: Siemens, other markings: 57T5, N22 (it is core material, I know), W03, 250. There is also added "683" - I believe this means 68mH.

TIA!

 

[unclejed613]

a small piece of shim plastic between the faces of the center post is usually what is used. it will change the inductance slightly (lower) IIRC

 

[ronsimpson]

In the factory, pot-cores have the center leg ground down to make the air gap.
By adding a shim or a piece of tape to all legs you get the same increase of gap but at twice the rate. The core will not be as well shielded. There will be some flux escaping the core. Tape is not as stable as gap materiel but it is fast and easy.

 

[ChildOfVision]

Yes, but how thick can that shim be? If it is, say, 1mm, I will get untight contact in outer rings, as I said!?

 

[ChildOfVision]

@ronsimpson:
But what if my inductor does not have air-gap at all (e.g. if it is made for transformers)? Is it possible to process mechanically in some way central pillars (by grinding for example)?

 

[alec_t]

how thick can that shim be?

AFAIK there's no theoretical limit, but in practice the gap is usually much less than 1mm. More like thickness of sticky tape or tissue paper.

 

[ChildOfVision]

Say I put small piece of paper (0,5mm for example) between central pillars (assuming that core is not air-gaped by factory). Ronsimpson said:

The core will not be as well shielded. There will be some flux escaping the core.

So, what can I do (if anything)?

 

[chemelec]

I gap Lots of Pot cores, just using a Thin plastic Slit Tape.
This tape is Simular to Scotch tape, but it has a better heat tollerance as it is made to be used in transformers.

And it seems to be stable over MANY Years.

 

[ChildOfVision]

I'm afraid that I can get this tape where I live. But I have very thin teflon tape - is it good enough? But, again, how to stop flux escapes from the outside of the shield?
edit:
Sorry, I forgot: so you have not-gapped core (for transformers, as you said), and make from it gapped core, and overall performance is even better because that core is originally intended for transformers - did I understand well?
New edit: I see now: not "core" but that Slit-tape is intended for transformers, sorry...

 

[MrAl]

Hi,

The factory gap is made in the center yes so that the flux stays contained better. But that's not the only way. As has been suggested already, a gap in the center PLUS a gap in the outer ring is good too, except you have to use half the thickness of the required total gap (if it was only in the center) because that way the gap appears twice along the magnetic path.
So there is no reason not to do it that way unless you have some application that really needs very good flux containment.

A place where you have to be careful for example is in switching power supplies when the inductor is close to the inside plastic of the case. If someone brings a strong magnet close to the case in that area, the inductance changes and could even cause the thing to saturate. This has actually happened a number of times with LED flashlights where the case is already small so the inductor is not deep within the case structure. Some of the lights blew out, some just dimmed in brightness. It was recommended not to try this because you dont know if it will just dim or blow out.

 

[ChildOfVision]

The factory gap is made in the center yes so that the flux stays contained better. But that's not the only way. As has been suggested already, a gap in the center PLUS a gap in the outer ring is good too, except you have to use half the thickness of the required total gap (if it was only in the center) because that way the gap appears twice along the magnetic path.
So there is no reason not to do it that way unless you have some application that really needs very good flux containment.

Well, that is something! But, should I need "very good flux containment"? I need that core/inductor (100uH) in oscillating circuit for L/C/ESR meter - about 85-90 kHz?

 

[crutschow]

Well, that is something! But, should I need "very good flux containment"? I need that core/inductor (100uH) in oscillating circuit for L/C/ESR meter - about 85-90 kHz?

So why do you need a gapped core for an oscillator?

 

[ChildOfVision]

I hope that I didn't understand that well from the beginning, but I read in forum dedicated to that meter, that air-gapped core is a MUST! Also, I found on some documentation about pot-cores in general, that cores without gap is intended for transformators, while gapped cores are required for oscillators (among other things). I really hope I am wrong. Actually, I don't have that core this moment (I will get it next week) so I can't check directly is it gapped or not. I just have that data from the first post. Btw. friend of mine, who will send me that core, is using these in inverters for energy-saving lamps i.e. from 12 to 220V (like transformer).

 

[MrAl]

Hello,

To gap or not to gap, that is the question.

But the answer does not really come from a particular application 'requirement' alone, but comes from the permeability of the core material and how well it fits the application with or without the gap. That means that one core without a gap might work well in two different applications, while another core without a gap may only work in one of those applications. There's a lot of possibilities here though, and sometimes the context of the reading material may infer some certain kind of core to start with, so it might be acceptable to state that the application requires a gap or doesnt, but in general you cant do that. Low permeability cores may get by in a lot of applications without a gap, while high permeability cores may not get by without a gap because the core saturates too easily. So if you start with a low perm core you may never need a gap, but if you start with a high perm core you may most of the time need a gap.

The whole idea behind a gap is to lower the permeability of the core so that it doesnt saturate with some amount of predetermined DC current or peak current. But if you already have a low perm core or very low perm core then you may not need a gap regardless of the application because it already has the ability to handle that DC or peak current level without a problem. Of course it also depends on the max flux density allowed for the core, but that's the main idea.

So your core may need a gap or may not depending on if it saturates once you get the required number of turns to get the required inductance value with the required level of DC current. If it saturates and you dont want to get a different core, then you use a gap. If it doesnt saturate, you're good to go.
You can test it in your app and see if it works or not, but if there is no DC current then all you have to do is worry about the peak current if it is low enough to not saturate the core.

So if you read an article that states that you need a gap, then they are probably referring to a higher perm core. You could either hunt for a lower perm core or put the gap in. Sometimes you dont want to be bothered by calculating the permeability that would work for your application, so you get something that would work probably with a gap and then you experiment with different size gaps until you get it to work properly.

You have to keep in mind that when you have a construction that has a certain inductance required for the application and then you use a gap to lower the permeability (because that is what it does) you then have to add turns to the core to get the required previous required inductance back again (lowering the perm means lower the inductance). This means adding more series resistance to the core which means lower efficiency. Also, once you add turns you also are adding to the ampere turns of the construction and that means that even if it didnt saturate with the (now) lower inductance it may again saturate with the new added turns. Thus, you may have to go through two or three of these experiments to get the right value of both inductance and permeability.
Very roughly speaking here, the inductance goes up as the square of the turns ratio, but the ampere turns only goes up as the ratio of the turns (not the square of the turns ratio) so the inductance goes up faster than the ampere turns so eventually you may find a number of turns and a gap width that works with your chosen core. If you cant get the required number of turns on the core you have to go to a bigger core size.

 

[jpanhalt]

I think you may already have an air gap. I did some experiments a few years back with cut toroid and a ferrite bar to complete the magnetic circuit. The fit has to be very close (i.e. lapped) to get it to approach a single, ungapped toroid. It doesn't take much.

My suggestion is just to try what you have. Or, a thin piece of tape, like the Teflon you mention, will probably give enough gap.

john

 

[ronsimpson]

Normally I gap all parts of a core because the core can break if you gap only the center leg and squeeze the core hard.
I apply tape across the whole core then with a sharp knife cut away the excess tape.

I have used gaps of 2mm. At 2mm the flux leaks out some.

 

[unclejed613]

i've seen TV and computer monitor flyback cores where a small (possibly 0.1mm thick) shim was glued on the faces of the core halves. the glue served two purposes, to hold the shims in place and to hold the core pieces in place, pot core gaps are usually accomplished by the use of very thin plastic washers. if you need a complete outer shell, i have seen pot cores with a gap in the center, and a powdered ferrite-bearing glue used to join the outside halves.

machining ferrite can be difficult. ferrite is a very hard garnet material. diamond saws and diamond bits are used for cutting and machining it. it is also very brittle and can shatter from mechanical or thermal stresses during machining.

 

[ChildOfVision]

Thank you all!
Sorry for delayed reply - it is about time difference...

I consulted my friend who is "expert" in this field and also build that meter: he said that the reason for gapped core in my appplication is more precaution measure (to prevent core saturation - and thus, instability in measurements). You are all right, he said that all I need (if anything) is to insert a small piece of thicker paper, so there will be practically no "leakage" from outer rings, since the gap is only some 0,1-0,2mm. Also, as you point (thank you MrAl for detailed explanations!) I should first experiment a little, so maybe I don't need that gap at all!

I hoped that somebody, maybe, knows data for that particular (very old) core, but if somebody knows how much its core material - N22 is "good" (in general), e.g. what is difference compared with today's modern ferrite materials?

TIA!

 

[alec_t]

Here's the spec for N22 ferrite
https://www.electro-tech-online.com/custompdfs/2012/04/PDF_N22.pdf

 

[MrAl]

Cracked Cores

Hi again,

I should have mentioned that a while back i did some experiments with toroid cores. As you know, toroid cores are made up as a 'ring' of magnetic material rather than have two halves, and require special winding equipment to wind in production.

Because the toroid is a complete ring with no two halves to separate, there's no way to put in a gap other than cutting a slot in the toroid. I would have done that, but i figured that the required gap would be too small for me to cut with my diamond saw, so the smallest gap i could make would be too large for the size of the toroid i was intending to work with (small like 3/4 inch or less outside diameter).

This led me to think up a new way to get a gap in the toroid. What i ended up doing was a series of experiments with 'cracked' cores. Yes that's right Cracking the core by inserting it into a nice size bench vise and slowly turning the handle until the core actually cracks into small pieces. I found that by putting the core into a cloth first meant that after it cracked all the pieces were held in one place by the cloth, but tape around the outside would work too, and that usually i got four distinct pieces.

Once the core is cracked, the pieces are then glued back together piece by piece using super glue. The super glue provides a thin layer so the gap stays small enough. But it does then enter the magnetic path four times, so the total gap is four times the thickness of one layer of super glue. It worked though for the experiments, and i found that even this thin layer altered the permeability quite a bit and so rendered the cores more useful for some applications where before they would never work.
The improvement here though would be to find a better kind of adhesive, one that would not break down with temperature. The super glue probably severely limits the useful temperature range so another kind of glue would be better. Maybe high temperature epoxy.

The thing about the gap in any construction though is that it should be of uniform thickness along the direction of the magnetic path. If it is thinner in one place than another because of poor machining, it could cause the core to saturate. I've never experimented with this though but always meant to. It's unclear whether the core would be less useful because part of it saturates and the other part doesnt, or if the part that saturates somehow affects the part that doesnt, but i would think that some measurements would clear this issue up too. Also, the machining process is usually done pretty accurately but that's mainly to get the faces to match up near perfectly. Those machines are very expensive too