Iron-core versus air-core woofer inductor intermod test.

[Al Klappenberger]

Bob,

Those Parts express inductors are actually made by "Erse". I am looking at the 3.9 mHy #16 wire version for the new extreme-slope network for Khorn. The Solen air core is .49 Ohms DCR. I might order one sometime to evaluate. I have a coupling issue to resolve sing two coils are mounted near each other. The can't couple each other. I know how to control coupling with two air-core coils but if one is an "I" core things will be different.

BTW: I did an impulse test on the 2.5 Mhy AA inductor and the 2.5 Solen earlier. Again, no difference!

AL k.

[BEC]

Al,

Just playing around with the different inductors on the bench, the two different iron core types both seemed pretty much imune to interference by at least iron objects around them. The air coils seem to be much more sensitive to iron objects at least around them. Even a screw in my workbench could be seen to change the readings on an air coil.

Bob

[Al Klappenberger]

Bob,

Yes! metal near an air core will mess it. Case in point: hat stupid steel crew on the AA tweeter filter inductor!

Actually though, my problem is coupling between two inductors. My Khorn extreme-slope design has two inductors in the singal path of the low filter. One is at the input and the other at the output. If they talk to each other the 25 dB arc top in the rejection band goes to pot! I need to check to see that doesn't happen. If you mount air-core inductos 90 degrees to each other and exactly on center the coupling is virtually nill. I need to verify that an "I" core and air core near each other won't "communicate"!

Al K.

[Al Klappenberger]

Bob,

Here's a picture of the two coils. Mounted like that they don't couple each other at all. The big one on the bottom is 3.9 mHy and has almost .5 Ohms DCR. It's the candidate for a iron core. The filter actully has 3 coils in series to the DC path. They all add up to a lot of DCR. It works fine, but it could be marginally better if I could get the DCR down.

Al K.

[Deang]

Even though those coils are sitting damn near perfect, I'm surprised you don't get at least some mutual inductance.

I'm starting to see some real hair splitting here with the iron core/air core usage in the LF part of the circuit. Just an observation based on everything said so far. What I find especially interesting is how the iron cores don't seem to be effected by being in close proximity to other metals. This is pretty important to me since I work with the PCBs, where the parts are all mushed together, and the inductor is usually within inches of the autoformer, as well as the leads from the caps and resistors.

One thing that has surprised me is the lack of discussion relating to the DCR of the LF inductor and its relationship to the Q parameters of the driver and the enclosure. Ideal LF performance is based on the relationship of the three, and I still think that arbitrarily lowering the DCR of the LF inductor doesn't deliver "better" performance at all -- but really just screws up the alignment between the driver and the enclosure.

[BEC]

Dean,

I think the iron core inductors are so massively effected by the large amount of iron inside the coil that they are mostly immune to effects from outside. Just hard for them to "see" any other iron around them.

As far as DCR is concerned, the iron core ones will be the lowest in that measurement because of the small amount of wire in them compared to the air coils. The two iron core ones we have been discussing are pretty close together and close to the one that PWK used. The air cores are somewhat higher. I don't know what the difference in DCR would mean to sound but I think AL is looking for a way to decrease that number in the 3 mH coil he is using which comes in at about .5 ohms (I believe). An original Klipsch 2.5 iron core inductor is about 0.279 ohms.

Bob

[Deang]

But you have the PERFECT inductor! It's the one that's mathematically correct, and lines up the box with the driver. There is simply no reason to use anything else!

BTW -- I'll take six.

I need a half dozen of those 3619s too.

Oh please, oh please.

[BEC]

Dean,

I will sell them to you.

Contact me.

bobcrites@centurytel.net

Bob

[Al Klappenberger]

Dean,

I did a lot of fooling with two air core inductors to see what configurations coupled least. The one I show in the picture is second best. The absolute best is 90 degrees and the edges facing each other. The way they are in the picture require precise alignment. The fields from one coil induce opposite phase energy into opposite sides of each other.

As to DCR and Q parameters, I'm interested. Tell me what I can do to test the effect of the network on the tuning. BTW, the Vance Dickason loudspeaker "Cook book" says to tune the speaker with the network IN the circuit. That supports your theory. When I tested the tuning of my "big Heresys" I did it with the network OUT. That means I should look at it again anyhow.

Al K.

[Deang]

I didn't expect to ever come back to this thread again, but have had a surprising experience. I finished my ALKs a few days ago, and as part of the build I went with some 10AWG Solen air cores which I mounted on separate boards because of their size. The DCR of the 10AWG air cores is .16, the same as the Erse Super Q steel laminates I've been using from Parts Express. FWIW, there's a pretty big difference in the bass. After an hour or so of listening I put some Super Q's into the circuit to see what it would sound like, and though still very good, sounded lean in comparison, and not nearly as full bodied and tight as the air cores. I've had the beasts laying around for a long time, and I'm glad I finally found use for them. I didn't expect to hear a difference, wasn't even listening for it, but it was pretty obvious.

[Al Klappenberger]

Dean,

Interesting that you could hear a difference. As you know, I had to go to an iron core inductor at the input of the ES400 network becasue of the DCR versus size issue. I did more measurements on the ES400 more carefully than before. I found the Intermod to be very low, but measureable. The distortion products were about 50 dB down at about 18W. That shouldn't be audable. The other two inductors are air core though.

Al K.

[Deang]

Beats me. I'm going through the same thing with caps too. The initial build used two 20uF Dayton Audios for the squawker. I had purchased the Auricaps for these ALKs, but since I had a bag-full of the 20uF Dayton's I thought I would try them in the boards first for a couple of days and then make the switch. The Dayton's sounded fine, and there certainly wasn't anything objectionable to the sound, in fact, the Klipschorns sounded great. I almost didn't even bother with making the switch since it sounded so good, I figured why bother. I finally found some time last night and just did it to get it out of the way, and sure enough, a smoother sound right out the gate. I don't know why some of you guys can't hear this stuff, I mean, it may not be night and day, but it's not hard to hear either. BTW Al, that's a great sounding network, seriously, I like it a lot.

[Al Klappenberger]

Dean,

I built a set of universal networks the other day with two 20 uF FastCaps instead of the 39+1 uF bypass. The measurements came out exactly the same. I expect to build them using 20+20 from now on. I will leave the kits using 39+1 though. I don't think the difference is worth the trouble of changing the instruction sheet! 20+20 costs me a few pennies less too! My new Stanford Research LCR meter says 20+20 has lower "D" than 39+1. I haven't done any listening tests though. I really don't trust me ears anyhow!

Al K.

[boom3]

I seem to recall that any magnetic material will have a sudden 'knee' in the BL curve where it goes into saturation. A memory from an old textbook. Has anyone tested the air core copper foil type inductors?

As we alll know, sometimes the added DCR of an air-core is a disadvantage in woofer circuits.

[Al Klappenberger]

Boom,

I tested a .2 mHy spiral inductor for use in a tweeter filter, but never a big one for a woofer filter. Anyhow, the "Q" curve peaked way to high for a tweeter filter. I think they are very much over rated!

Al K.

[krustyoldsarge]

----------------

On 5/28/2005 8:45:18 PM DeanG wrote:

I didn't expect to ever come back to this thread again, but have had a surprising experience. I finished my ALKs a few days ago, and as part of the build I went with some 10AWG Solen air cores which I mounted on separate boards because of their size. The DCR of the 10AWG air cores is .16, the same as the Erse Super Q steel laminates I've been using from Parts Express. FWIW, there's a pretty big difference in the bass. After an hour or so of listening I put some Super Q's into the circuit to see what it would sound like, and though still very good, sounded lean in comparison, and not nearly as full bodied and tight as the air cores. I've had the beasts laying around for a long time, and I'm glad I finally found use for them. I didn't expect to hear a difference, wasn't even listening for it, but it was pretty obvious.

----------------

Dean,

Thanks for trying 10AWG woofer air cores out for me!

Been thinking of doing it for my ChorusII's but the separate board thing, etc. has held me back. Now I've got the experience of someone I respect before I shell out the bucks for a pair of those monster coils.

[Deang]

A picture of the beast resting inside the top hat.

You don't trust your ears, but I don't have complete faith that the measurements are telling everything. Here's a few examples to ponder over:

BEC built a board with two Type A's on it, with a switch to go back and forth between them. One circuit was built with his GE cans, the other had the Jensens in it. Though the Jensens had substantially higher ESR than the GE cans, neither Bob or his son Mike could hear a difference going back and forth on the fly using the switch. Here is an example of things measuring differently, yet sounding the same. Should we reject the results of the experiment, and if so, on what grounds? If we accept the result, then what conclusion do we draw? Needless to say, I'm not comfortable with either choice! If the former, we must either blame the switch, the gear (unable to resolve the differences), or the ears used in the experiment. If the latter, well ...who really wants to go there?

I built up a Type A board with the GE cans, and another with the Jensens. Using mono, and pushing the Klipschorns out from the wall and somewhat against each other, I went back and forth between the two and easily noticed a difference in the sound of the squawker/tweeters. I can't tell you which was "better", because it wasn't the proper way to determine that, nor was that the purpose of the experiment. I used some old Beatles material, and some Santana (recorded in mono). Next I built up some Auricapped A's to throw into the mix, and waters started to muddy up somewhat. I definitely felt burned out at that point, and all I could really determine was that it sounded more like the Jensens than the GE cans. The cans always sounded sharper, and crisper in comparison. When I found the time to build mates for them (and after getting the rest of the Jensens back from Bob), I ended up preferrring the GE loaded Type A's for normal listening, but found them too incisive at higher SPL's. So, why are my experiences and conclusions different than Bob's? Is it because I'm only 12 feet off of my Klipschorns? BTW, no Behringer in the system while doing any of this.

I finally threw in towel with the Type A experiments, and built up some nice Jensen loaded AA's to use while I built up the ALKs. Putting them in, I thought they sounded like crap. Rough and thick sounding, and thought I had soldered something up wrong. The sound was really different than the Type A, and in a very bad kind of way. I thought, "this is the freaking GREAT Type AA?!" I checked them out, found everything in order, and turned everything back on. It was bad enough that I was just going to cut my losses and hide them in the workroom somewhere. Instead, I caved in to the audiophile in me. Half of the Jensens for these networks were brand new, so I put the Behringer in the system, turned on the Pink Noise, set the volume to 11, and left the room for 3 hours. Oh yeah, it was awful, and Debbie was ready to wring my neck. Even with the system being upstairs, with the door closed, we could hear it coming through the registers in the basement. Before dinner I went up and turned everything off, and pulled the Behringer out. After I ate, I went up, powered everything back on, and settled in for a listen. It was still syrupy, but not nearly as rough and harsh. I listened for about an hour, and each day after it just kept getting better and better. By the time I had the ALKs built up, the sound was something I could easily live with. Now, I'm not going to go for the "it was your ears getting used to the new sound" bit, because the initial sound was terrible. Terrible is terrible, and no amount of listening gets a person "used to it". To test myself, I put the Type A's with the GE cans back in, and it was immediately apparent that the Jensen AA's were much smoother in the treble and more open sounding in the midrange. This resulted in me blasting the GE loaded Type A's with Pink Noise, and adding the .10uf film and foil bypass cap to the tweeter can. I haven't put them back in to listen since doing it -- I'm just too tired to mess with anymore. The ALKs are in, and I just want to enjoy some good sound for a while.

I've tried really hard to allow the measurements to explain everything, and after the experiments with the Heresy's last month, it was getting easier. Doing this stuff with the Klipschorns really confused me, and I'm at a bit of a loss to explain it.

There is one last thing. I'm in possession of a response plot that compares the ALK, and an old set of AA's. With the exception of the elevated high frequency extension/output of the ALK's (probably because the tweeter cans in the AAs were shot), the traces from the rest of the plot are practically right on top of each other. From a meaurements perspective, there isn't much of a difference, however, I'm sure we can agree that from an audible perspective - the difference is substantial.

[Al Klappenberger]

Dean,

There is a scientifc reason for all of your observations:

It's all in your mind!

Al K.

[BEC]

Well put Al.

Bob

[Deang]

It's all in your mind!

You bet it is -- right where it matters most!