[SOLVED] Printed Planar Spiral Inductor (Multilayer)

[faisal78]

Hi
I am trying to design a spiral inductor on my PCB.
I have a space of 10mm x 10mm for it, the good news is I also have 8 PCB layers to achieve it.
There a lots of formula's for a single layer spiral inductor available on the net, and by above dimensions I can only achieve about 2uH with 20 turns with 0.1mm trace width and 0.1mm spacing.

There are some articles on multilayer spiral inductor (i have only seen up to two) which can double up the inductance by inverting (counterclockwise) the sprial on the opposite layer. There are some mutual inductance coupling factor to take into account to also.
Is there a good simulation tool for this to accurately predict this?

Is there such thing as 4 layer/6/ or 8 layer sprial inductor? Would it even work? Wouldn't the magnetic fields just be cancelling each other out?

 

[keith1200rs]

Have you seen the attached? You need to keep going round in the same direction if you want an inductor. The SRF could be low, as could the Q.

Keith

 

[faisal78]

Hi Keith,
I have read the article. Thanks for it.
It however does not mention of the layout implementation of the multilayer inductor.
I would assume in the 4 layer spiral implementation, top layer (1) and 3rd layer (3) would be clockwise winding, and second (2) & bottom (4) layer would be counterclockwise winding because of the vias.

I find it surprising that no many articles in regards to this. I am not sure if these formula's on have been verified by others.
I will give it a try.

Any simulation tools that can do this that anybody can recommend prior to me making the PCB?

 

[keith1200rs]

I don't know why you think the windings would reverse direction when going from one layer to another - if you do that the magnetic fields will cancel and you won't have an inductor! You must always go in the same direction. You want a positive mutual coupling between layers. If the top layer inductance starts on the outside and moves to the next layer at the centre of the spiral the next layer will start at the centre of the spiral and move to the outside but it will still be in the SAME direction- all clockwise or all anti-clockwise.

There are other papers around including for IC design. Some of the Q factors are appalling though - around 2. If that is typical of what is achievable then you can see why it is not commonly used.

Keith

 

[faisal78]

Yes you are correct.
It will all go in the same direction CW or CCW. I got confused with myself.
The windings would be circling from the outside in, then a via, then the same winding directly under would spiral inside to out.

I am basically trying to make a inductive sensor for metallic objects. I am only trying to detect if a metal object is present or not.
The distance from the windings to the metal object would be <5mm.
How would inductor Q affect my implementation?

 

[keith1200rs]

A very low Q inductor will look like a resistor in series with an inductor. You will still see an inductive effect but the resistor will affect your results - in your case giving a smaller change when you detect an object compared to having a good inductor. This may not matter depending on your circuit design.

Do you really need high inductance? Can you use a small inductance but higher frequency?

The easiest way to find out the characteristics is to make something. Put a selection of different inductor designs on a PCB and get it made. There may be simulation software out there to help - I don't know.

Keith