inductive coupling

[John Roman]

 Good early morning,
4am and I'm typing out a question to the group, that can only be described as a sickness. Oh well! I'm trying to finalize the layout of a three way passive crossover network and am concerned about potential inductive coupling. I've looked at several crossover's online and the placement of the inductors seems to close in some and unnecessarily wide in others. I'm using both air core and steel laminate inductors, all 16 gauge. The goal is to reduce the overall size of the network for aesthetic reasons but more importantly to ensure no interference. Keeping the inductors out of phase is apparently key but I'm unsure how to be certain there are no issues. Some sort of test procedure that can be utilized during burn-in. Some way to check the magnetic field around the inductors, there must be a simple way. I'm just not familiar with it. Any ideas?
Have a great Friday fellow Bottleheads.
John

[RayP]

This article should help on layout if you haven't found it already.
http://www.parts-express.com/projectshowcase/xover/xover.html

As to measuring the field, take a look at this url about half way down. It would appear that the simplest way is to use a compass and see how much the magnetic field in different orientations deflects a compass needle.

http://www.lessemf.com/dcgauss.html

ray

[corndog71]

Coils should be at 90 degree angles.

Here's my reference.

http://www.troelsgravesen.dk/coils.htm

(https://forum.bottlehead.com/proxy.php?request=http%3A%2F%2Fi576.photobucket.com%2Falbums%2Fss207%2Fcorndog642%2Fx-voce%2520upgrades%2FIMG_0906.jpg&hash=e351bdc469d11984297c7276ae4246a2f252601d)

[Paul Joppa]

...4am and I'm typing out a question to the group, that can only be described as a sickness. ... Some sort of test procedure that can be utilized during burn-in. Some way to check the magnetic field around the inductors, there must be a simple way. I'm just not familiar with it. Any ideas?...

You can induce an AC current in one inductor and measure the voltage generated in another. You must be sure to use enough current in the source to make an easily measured voltage in the receiver. How much current? How small must the voltage be? Those are tougher questions. There was an article years ago in some DIY audio magazine in which the author did some measurements and listening tests - sorry I can't locate it right now. What I recall as the bottom line was that most crossovers in actual use have audible magnetic interaction problems.

Generally, air-core inductors will have the largest external field per unit of current. It extends a long way, in practice I would say 2 to 4 times the largest dimension of the coil. You can put it in a shield box of mumetal or soft steel - but unless the box is several times the largest coil dimension it will change the inductance of the coil. (The presence of the speaker will have a similar effect if it is not far enough away.)

Using a magnetic core can contain most of the field if the air gap is small. Many crossover parts have a large air path - the "gap" from one end of the iron path to the other is long - so the external field is still nearly the same as the air core. The more the core reduces the external field, the more it introduces potential core distortion - no free lunch there. Ferrite pot cores that surround the coil and keep the air gap inside the coil are much better shielded than ferrite bobbins that are inside the coil, leaving the air path outside.

The usual recommendation is to make each pair of inductors perpendicular to each other. Actually, you want the axis of the field produced by one to be perpendicular to the field lines of the other, and the field lines are curved because they must make a closed loop. So mechanical perpendicularity only works if the two are also positioned symmetrically - like |- but not like |_ for example. You can arrange three inductors in a line and achieve this, but there are only three dimensions so you cannot do this with more than three - distance is your only other hope.

More than you wanted to know, I suspect ...  :^)

[John Roman]

Thanks everyone, that will definitely give something to mull for awhile!
John

[corndog71]

With all due respect to Paul,  I get my advice from award-winning speaker designer, Danny Richie, of GR Research.  

I would agree that it would be great to separate the coils as much as possible.  An external crossover is not a bad idea.  But most people are used to having the crossover inside the speaker cabinet where it's convenient and won't get lost or damaged.  It's up to the person building the speaker.

If you look at the picture I posted you'll notice the inductors are arranged in the best way possible to minimize interference from each other.  This is taking into consideration all of those parts are sharing a single circuit board.  I could if I wanted to be anal about it chuck the board and hardwire everything with lots of space in between but in the end I'd rather not.  The speaker sounds awesome as it is and practically speaking I just don't have the space to do that sort of thing.

I posted a link as well that shows how both lots of space, and if that's not an option, a 90 degree angle will minimize the interference of the magnetic fields.  That is the measured electrical effect.  It says nothing about the audible effect.  Danny promotes the 90 degree angle method in his designs and I'm pretty sure many other high end speaker designers would agree.

[Paul Joppa]

>> With all due respect to Paul,  I get my advice from award-winning speaker designer, Danny Richie, of GR Research. 

Yes, he's a smart guy.

>>I would agree that it would be great to separate the coils as much as possible.  An external crossover is not a bad idea.  But most people are used to having the crossover inside the speaker cabinet where it's convenient and won't get lost or damaged.  It's up to the person building the speaker.

Yes again, it's the great thing about DIY.

>> If you look at the picture I posted you'll notice the inductors are arranged in the best way possible to minimize interference from each other.

No, they are not. Pretty good, but not the "best."

>> This is taking into consideration all of those parts are sharing a single circuit board.

Yes again, that's what I meant by "pretty good."

>> I could if I wanted to be anal about it chuck the board and hardwire everything with lots of space in between but in the end I'd rather not.  The speaker sounds awesome as it is and practically speaking I just don't have the space to do that sort of thing.

That's the point I was trying - apparently without success! - to make. It is unreasonably difficult to make a no-compromise design with more than two or three air-core inductors. Engineering is all about compromises! You didn't ask for compromises, and I didn't offer any. I thought you asked for more information, which I tried to provide. I'm sorry if you thought I was being critical - I had no such intention.

>> I posted a link as well that shows how both lots of space, and if that's not an option, a 90 degree angle will minimize the interference of the magnetic fields. That is the measured electrical effect.

Those things will reduce the coupling in most cases, most of the time. They don't always represent the true "minimum" but they are good engineering compromises in many cases.

>>  It says nothing about the audible effect.  Danny promotes the 90 degree angle method in his designs and I'm pretty sure many other high end speaker designers would agree.

Obviously you think I am not a "high end speaker designer." Well, I never made any such claim. You don't need to argue with me; you are welcome to ignore my information if you like.

[corndog71]

Didn't mean to jump down your throat there Paul.  Often the printed word doesn't capture the subtleties of a conversation.  Plus I was feeling kinda uppity at the time.  Apologies for that.

I just wanted to express to the OP not to worry about it too much and give him some advice.

And FWIW, I would also consider you and Doc B to be fantastic tube gear designers. (:

[glynnw]

In the FWIW category, years ago I was using a Japanese humidifier in a room with a  Lazarus tube preamp.  This humidier worked by blasting the water into very small pieces (no heater used) which then went throughout the room as a mist.  After a couple of months I opened up the preamp to do something and there were very clear white patterns around every component in the preamp.  It was really easy to see which radiated the most "stuff".  I don't know if it was a magnetic effect or what, but it was obvious just how much overlapping there was with these fields.  Might be a good tool for someone with the patience to let it sit a few months.

[Paul Joppa]

Didn't mean to jump down your throat there Paul.  ...

Looking back on it, I guess I was pretty crabby last night too. So I'll counter-apologize and we can just say we owe each other a beer!