Series Inductors question

[arc2v]

I searched on this but did not find a good explanation, so I figured I'd ask explicityly:

I finally started building my open baffle center channel and wired up the crossover last night. One inductor in the midrange calls for 2mH. I have 2, 1mH inductors on hand (air core Jantzens from PE), so I wanted to wire them up, rather than shell out $20 for a new inductor. This is a prototype after all.

I know they need to be in series, but I'm afraid I'm at a loss for dealing with the mutual inductance. I've thought myself in circles and I think I put them opposite one another by accident.

For clarity sake, let's call the wires inner and outer, and the inductors A and B

Right now, I have A and B stacked atop one another. The signal comes in A-outer and A-inner and B-inner are connected. The signal goes to ground through B-outer. Again, these are stacked atop one another.

I got a weird distortion in the midrange, like the crossover wasn't working right (still breaking in, so no FR sweeps yet). so this was my first troubleshooting stop.

Now I'm thinking I need to have A-inner connect to B-outer.

Any advice here? If I get motivated, I will fire up Speaker Workshop and try some component measurements (even though it's not as accurate for inductors).

Thanks,
AC

 

[Conrad Hoffman]

Think of it as if you were winding a single larger inductor. If the direction is clockwise on one inductor, you have to keep it that way on the second, lest they fight each other. You can also think of them as two transformer windings, like a dual voltage primary. wired out of phase, it doesn't work!

 

[arc2v]

That's sort of what I was thinking. I actually thought myself in circles and just ended up doing it and figuring I'd debug it later.

Followup:
Someone forwarded me a webpage link a long time ago regarding inductors and location. It had a lot of experiments and the guy measured inductance for different orientations and separation distances and showed the results.

Does this sound familiar to anyone? If so, please resend the link, b/c I have searched for it and cannot find it to save my life.

Audioholics had a decent writeup too, but this one was more thorough.

Thanks,
Ac

 

[zigzagflux]

How about this:

http://www.audioholics.com/education/loudspeaker-basics/inductor-coil-crosstalk-basics

 

[arc2v]

that's the one I was mentioning in the post. I found that, but the other one (the one I can't find) was more experimental. The guy was actually measuring changes in mH on the inductors. I don't think it was geared towards crossovers, just electronics in general.

Thanks though.

Ugh, next time, I will bookmark everything I find interesting

A

 

[dlneubec]

Here is another good reference. I can't remember where it came from.

 

[arc2v]

That's It!!!

That was the summary from the article I was looking for.

Now if we can just find the original

It's good to know my brain isn't just making things up.

Thanks,
A

 

[dlneubec]

Here it is:
http://www.troelsgravesen.dk/coils.htm

 

[arc2v]

Thank you so much!

He calls them coils, which is why my searches for inductor never found it.

Now to bookmark and never lose it again.

thanks again all!

AC

 

[owdi]

If you stack two 1mh inductors on top of each other, they will couple and sum or subtract, giving a final inductance either above or below 2mh. If you want 2mh, keep them far apart, and turn one 90 degrees away from the other.

When you replace the pair with a single 2mh inductor, keeping wire guage constant, RE will fall.

Dan

 

[Svante]

Here is another good reference. I can't remember where it came from.

Hmm, that picture is ... sort of, how shall i put it, too cathegorical?

I mean, first it assumes that the cupling between coils should be minimized. This is usually an ok assumption, since two coils in a typical filter should not have any coupling, but in this case it is not. But if we assume for a while that the coupling really should be minimized, it is still wrong. That is, 1 2 and 3 is right. 4 and 5 are dead wrong, the coupling between these coils is near zero. 6 7 and 8 are right.

Now assuming that we DON'T need to minimize coupling, the others are ok too. They will all give some extra coupling. In the case of two 1 mH coils, positions 4 and 5 will result in 2 mH if the coils are connected in series. Depending on the orientation of the coils the other orientations will either result in an increase or decrease of the inductance. Orientation 8 will nearly result in 4 mH if the coils are turned "the same way" and nearly 0 mH if they counteract one another.

 

[owdi]

Hmm, that picture is ... sort of, how shall i put it, too cathegorical?

I mean, first it assumes that the cupling between coils should be minimized. This is usually an ok assumption, since two coils in a typical filter should not have any coupling, but in this case it is not. But if we assume for a while that the coupling really should be minimized, it is still wrong. That is, 1 2 and 3 is right. 4 and 5 are dead wrong, the coupling between these coils is near zero. 6 7 and 8 are right.

Based on my limited experiments with two air core coils, I have to disagree with you. If I had to rank them from least to most, it would be: 6, (7 or 1), 2, (3 or 5), 4 and finally 8.

In my tests I found that most coupling occurred if you can see one inductor through the middle of the other inductor. Now I realize this may not be the general case, and it could be highly dependent on the shape of the inductors, but I could not say the coupling between 4 and 5 is near zero. Just the opposite.

Dan

 

[dlneubec]

Hmm, that picture is ... sort of, how shall i put it, too cathegorical?

I mean, first it assumes that the cupling between coils should be minimized. This is usually an ok assumption, since two coils in a typical filter should not have any coupling, but in this case it is not. But if we assume for a while that the coupling really should be minimized, it is still wrong. That is, 1 2 and 3 is right. 4 and 5 are dead wrong, the coupling between these coils is near zero. 6 7 and 8 are right.

Now assuming that we DON'T need to minimize coupling, the others are ok too. They will all give some extra coupling. In the case of two 1 mH coils, positions 4 and 5 will result in 2 mH if the coils are connected in series. Depending on the orientation of the coils the other orientations will either result in an increase or decrease of the inductance. Orientation 8 will nearly result in 4 mH if the coils are turned "the same way" and nearly 0 mH if they counteract one another.

I'm afraid I don't understand your disagreement, assuming you want to avoid coupling. Did you look at the link above where it shows the testing results? Are you suggesting his test are done incorrectly?

 

[tinitus]

As I recall, Troels has made his conclutions based on measurements how induction reacts

 

[sreten]

4 and 5 are dead wrong,

Categorically speaking this is true. A bit of mental gymnastics
will reveal 4 and 5 to be the same as 6 and 7 and thus wrong.

/sreten.

 

[dlneubec]

Troels Gravesen tests

I just don't see it. Check out the link above. 4 and 5 are not the same as 6 and 7. Look at his test results. Clearly they cause the coils to interact differently. I'd say that in 4 and 5, the field of the right coil is "aimed" at the left coil. Not so in 6 and 7.

 

[tinitus]

fig 6 and 7 is the way to do it, I wouldnt think of any other

But the issue is two inductors coupled in series, it might be a totally different matter - could even be that fig 8 is the one

 

[sreten]

Troels Gravesen tests

I just don't see it. Check out the link above. 4 and 5 are not the same as 6 and 7. Look at his test results. Clearly they cause the coils to interact differently. I'd say that in 4 and 5, the field of the right coil is "aimed" at the left coil. Not so in 6 and 7.

Hi, You are correct they are different in that respect, /sreten.

 

[tinitus]

When you have 3 coils, you use a combination of fig 5 and 7

 

[dlneubec]

It seems to me that if you are putting two 1mh coils in series to reach a 2mh value, you want them to maintain their original 1mh value, so you do not want coupling to occur. If coupling occurs then their combined values may be something different than what the two individual values add up to.