Inductor winding experiments in parallel (on the same core)

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

zebra50

Well-known member
Joined
Jun 4, 2004
Messages
2,943
Location
York, UK
Hi!

I did a couple of little experiments today with an inductor core, and have managed to tie my brain in a knot. Can anyone help me explain the difference between these two scenarios to myself?

Experiment 1.
I took two lengths of 0.5 mm enamelled wire, held them together and wrapped them together a few times clockwise around a bobbin, so that there were 14 turns  of each, forming two layers. I added the core and measured each winding at about 1.34 mH at 1KHz. Then I put the pair in parallel, and measured 1.35 mH. The two coils of wire are acting as a single, thicker coil, which is pretty much as I expected. (I could also get a very low reading by putting them in anti-parallel.

Experiment 2.
I wound a layer of 15 turns of wire around the bobbin, clockwise, and then wound a second layer anticlockwise. The windings read about 1.5 mH each. When I put them in parallel, phase aligned, I read 0.75mH - half the value of a single winding. (Or about zero when anti-parallel.) The two coils are now acting independently, as though they are two separate inductors.

So, both are two windings, in parallel, on the same core, wired so that the signals are in phase. But they behave differently.
Is the second scenario different simply that the two coils can no longer act as a single entity, because one is anticlockwise?

Thanks!

Stewart
 
interesting. there may be capacitive coupling at play--don't know if that helps explain.

wind 10 turns, and 10 more first 10 measures X mH 20 measures 4X mH dunnit;

you got X\2...
 
>wind 10 turns, and 10 more first 10 measures X mH 20 measures 4X mH dunnit;

Yes, that's what I expected, and I did indeed get the expected circa 6 mH when I wired them in series (and in phase).

The maths of parallel, separate inductors is pretty simple, and the same as for two parallel resistors - and experiment two behaves like that, giving L/2... So that is OK...

And experiment one behaves like a single coil of double thickness - and I can understand that too.

But when I consider both together then my brain starts to overheat. :eek: I'm sure this should be a textbook thing, that I ought to know.
 
OK, I've got it. The mutual inductance must of course be included

http://www.play-hookey.com/dc_theory/series_inductors.html
http://www.play-hookey.com/dc_theory/parallel_inductors.html

and

http://www.electronics-tutorials.ws/inductor/parallel-inductors.html

By flipping the direction of the winding, I am simply reversing the mutual induction.

Thanks for putting up with my dumb question. Sometimes thinking out loud helps clarify things.

;D
 
try it again with 100 turns each and you may get different results,

how big is the core?

get a bigger core and bobbin,

1 mh is pretty low for a DMM type induction bridge, barely off the noise floor,
 
Good idea, CJ - it would be interesting to see how the mutual inductance changes as the the windings get fatter, and less ideal.

Core is/was a stack of 19 x 0.3 mm lams, EE25, 80% Ni. My LCR meter happily measures microhenries - it is not exactly NIST traceable, but certainly OK for 1.00 mH, and Good Enough For Audio.

Thanks!
 
permalloy can give really weird results sometimes,

take apart a blown power trans and use a steel core,

clip leads can have many microhenries,

you need about 100 mh to get rock solid on this,

there is no way to divide inductance in half with 2 windings,

you can get zero-out of phase,
double-series
or the same-windings connected parallel,
but not half.

unless they were of un-equal turns,
 
zebra50 said:
Experiment 2.
I wound a layer of 15 turns of wire around the bobbin, clockwise, and then wound a second layer anticlockwise. The windings read about 1.5 mH each.
That's where the problem is. They are close in value, but one has one more turn; when you put them in parallels, there is one turn shorted. This short reflects on the resulting impedance. One turn difference can make a huge difference in the end.
 

Latest posts

Back
Top