# printed inductors

T

#### Tom Del Rosso

Jan 1, 1970
0
When an inductor is printed on a PCB in the shape of a 'square wave' or
square zigzag, each time it zags the trace doubles back on itself. Doesn't
this cancel the magnetic field to an extent? How much does the coil
structure matter to a coiled inductor, or would the straight wire have the
same inductance?

B

#### brent

Jan 1, 1970
0
When an inductor is printed on a PCB in the shape of a 'square wave' or
square zigzag, each time it zags the trace doubles back on itself.  Doesn't
this cancel the magnetic field to an extent?  How much does the coil
structure matter to a coiled inductor, or would the straight wire have the
same inductance?

perhaps it is a high frequency inductor which is more like a
distributed transmission line inductor?

P

#### Phil Allison

Jan 1, 1970
0
"Tim Wescott"
When it doubles back it'll reinforce the magnetic field, and only
destructively interfere on the next line over (with the current going the
same way).

** Tim has no idea just how stupid he is.

..... Phil

G

#### George Herold

Jan 1, 1970
0
When an inductor is printed on a PCB in the shape of a 'square wave' or
square zigzag, each time it zags the trace doubles back on itself.  Doesn't
this cancel the magnetic field to an extent?  How much does the coil
structure matter to a coiled inductor, or would the straight wire have the
same inductance?

We make spiral coils on PCB to 'shim' magnetic fields.
I'm not sure what you mean by zig-zagging back... maybe a picture?

Re: a coil versus a length of wire. You get more inductance from the
coil. I think if you have a fixed length of wire and wanted to make
the largest inductance then you'd wind it as one big single turn
coil. (hopefully someone will correct me if that is wrong.)

George H.

B

#### Baron

Jan 1, 1970
0
George Herold Inscribed thus:
We make spiral coils on PCB to 'shim' magnetic fields.
I'm not sure what you mean by zig-zagging back... maybe a picture?

Re: a coil versus a length of wire. You get more inductance from the
coil. I think if you have a fixed length of wire and wanted to make
the largest inductance then you'd wind it as one big single turn
coil. (hopefully someone will correct me if that is wrong.)

George H.

I suspect the OP is refering to the meander lines on a pcb used to match
signal timing over varying trace lengths.

T

#### Tom Del Rosso

Jan 1, 1970
0
John said:
The ones you see on PC boards may be delay lines. When you route a
differential pair and turn a corner or something, one trace gets
longer than the other. So some people wigwag the shorter one to make
the electrical lengths equal.

I should have said that this definitely is an inductor. It's a choke
connected to the power trace on a 4GHz board.

G

#### George Herold

Jan 1, 1970
0
On 5/14/2013 8:02 AM, George Herold wrote:

> We make spiral coils on PCB to 'shim' magnetic fields.

PCB inductors appear to have poor Q.
They are convenient to make, but lacking performance.

Yeah just DC coils.
> I'm not sure what you mean by zig-zagging back... maybe a picture?
>
> Re: a coil versus a length of wire.  You get more inductance from the
> coil.  I think if you have a fixed length of wire and wanted to make
> the largest inductance then you'd wind it as one big single turn
> coil.  (hopefully someone will correct me if that is wrong.)

That is incorrect. Using same piece of wire, max. inductance is achieved
with max. number of turns.

OK thanks... (of course I have to go and check it now... grumble)

George H.

G

#### George Herold

Jan 1, 1970
0
How many channels? I saw a 40-channel NMR shim system once. It was hard to tune.

Hard to tune... I bet. This has just 4 coils, one each for the X, Y
and Z gradient and then a Z^2 gradient. (Z is the diectron of the
static field.)
There is a solution for that somewhere. Turns out to be a stubby solenoid..

OK I'll have to check... gotta run

George H

S

Jan 1, 1970
0
P

#### petrus bitbyter

Jan 1, 1970
0
Tom Del Rosso said:
You mean for a given length of wire? Doesn't seem useful, since you can
use a different length.

Sure. But if you need to produce some 100,000 coils, wirelength may become
an important parameter.

petrus bitbyter

G

#### George Herold

Jan 1, 1970
0

A Brooks coil is pretty close to optimal.- Hide quoted text -

- Show quoted text -

What about a toroid? (I always want to put to r's in torroid)
Or do you get more B field by spraying it everywhere?

George H.
(I love SED, say something stupid and learn something new, thanks)

T

#### Tim Williams

Jan 1, 1970
0
George Herold said:
What about a toroid? (I always want to put to r's in torroid)
Or do you get more B field by spraying it everywhere?

You get more B by focusing it into a smaller space, but that doesn't
usually give you more inductance -- it takes a lot of work to make B,
because B is energy density (in fact, e ~ B^2).

Dunno about the length required for an air-core toroid. Should be easy to
find from the thin, infinite-turns toroid formulas though.

Tim

G

#### George Herold

Jan 1, 1970
0
You get more B by focusing it into a smaller space, but that doesn't
usually give you more inductance -- it takes a lot of work to make B,
because B is energy density (in fact, e ~ B^2).

Yeah, I get to integrate the B field over all space so letting it
spread out is a win. (In other ways toroids seem like a nice inductor
shape.)

George H.

P

#### Phil Allison

Jan 1, 1970
0
"Spehro Pefhany"
Optimal coil length is a bit less than the coil radius.

A Brooks coil is pretty close to optimal.

** You have just linked a page from a Kiwi radio ham ??

WTF ??

The wanker does NOT even mention that his question is restricted to single
layer coils.

With any number of layers permitted, the solution is more like a sphere with
a hollow centre.

Coils made for loudspeaker crossovers were sometimes wound like a ball of
wool in a "beehive" shape for maximum utilisation of the wire used.

..... Phil

R

#### Robert Baer

Jan 1, 1970
0
Tom said:
When an inductor is printed on a PCB in the shape of a 'square wave' or
square zigzag, each time it zags the trace doubles back on itself. Doesn't
this cancel the magnetic field to an extent? How much does the coil
structure matter to a coiled inductor, or would the straight wire have the
same inductance?
The only geometries i saw were spiral in nature on one side (shaped
either circular or square); inner end of spiral to via withstraight exit
to outside of area.

J

#### josephkk

Jan 1, 1970
0
When an inductor is printed on a PCB in the shape of a 'square wave' or
square zigzag, each time it zags the trace doubles back on itself. Doesn't
this cancel the magnetic field to an extent? How much does the coil
structure matter to a coiled inductor, or would the straight wire have the
same inductance?

That is not an inductor, that is a delay line (transmission line type).

?-)

J

#### josephkk

Jan 1, 1970
0
But is it wrong to call it an inductor?

Its operation depends on its magnetic field, so what is the effect of having
each segment of conductor nearest to segments conducting in the opposite
direction?

At 4GHz it is not so clear. If you really want to know the tools are
expensive (in part to really encourage you to buy platform enough). Like
serious 3D EM solvers. Ask Dr. Hobbs.

?-)

J

#### josephkk

Jan 1, 1970
0
PCB inductors appear to have poor Q.
They are convenient to make, but lacking performance.

That is incorrect. Using same piece of wire, max. inductance is achieved
with max. number of turns.
Not always. The relationship between A (area enclosed) l (length of coil
transverse to its axis) and n (number of turns) is not that clean. Crank
the formulas and see for your self where the first and second order terms
are.

Replies
10
Views
2K
Replies
0
Views
1K
Replies
5
Views
2K
Replies
5
Views
2K
Replies
2
Views
1K