Interpreting These Crossover Measurements - ALK Universal

[naaac]

So, I am slowly restoring an old set of Khorns. I just took some measurements of the latest configuration and I'm wondering if this is looking like it's supposed to. 

Set up: UMIK 1 / REW / about 3 feet off the ground / 3.5 feet from front of Khorn / 1/6 smoothing

Woofer: Stephens 103LX (16 Ohm, DCR ~ 10 Ohms) // mounted on slide-in motorboard, with 6"x13" slot
Mids: University SA-HF + K-5-J
Highs: K77 Tweeter (square magnet)

Crossover: I built the ALK Universal crossover with Solen caps and Litz inductors on a k-500-5000 crossover board (1954). I'm using the old autotransformer on the board and the squawker is connected to the -3 and 0 tabs on it.

Here are the measurements. Is the woofer response supposed to look like that? It looks like it's not being properly attenuated on the upper end.

 

[mboxler]

What value inductor are you using for the woofer?  Looks like you would need a 6.4mh to cross a 16 ohm woofer at 400hz.

[naaac]

7 minutes ago, mboxler said:

What value inductor are you using for the woofer?  Looks like you would need a 6.4mh to cross a 16 ohm woofer at 400hz.

I am using a 1.3 mH inductor. I tried it with the old 5 mH inductor from the k-500-5000 crossover, and it didn't change the upper woofer taper. 

[mboxler]

I was hoping you had the stock 1.3mh inductor, as that would be way too low, and would account for your measurements.  But the 5mh should have been better.

Any way to measure the 5mh to verify it's value?   

[naaac]

1 hour ago, mboxler said:

I was hoping you had the stock 1.3mh inductor, as that would be way too low, and would account for your measurements.  But the 5mh should have been better.

Any way to measure the 5mh to verify it's value?   

 

You were right! I measured it with DATS V2 and it was actually 2.6 mH - even though it was wrapped at 4.9 mH in 1954. Looks like I need to get a bigger inductor and try it out....

[Wirrunna]

naaac, a couple of links that may be handy if you haven't already found them:

https://community.klipsch.com/index.php?/topic/110366-why-is-the-alk-universal-network-universal/&

https://community.klipsch.com/index.php?/topic/124362-alk-universal-crossover/

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

[mboxler]

9 hours ago, naaac said:

 

You were right! I measured it with DATS V2 and it was actually 2.6 mH - even though it was wrapped at 4.9 mH in 1954. Looks like I need to get a bigger inductor and try it out....

That is old!  If you want to stay at 500hz, 5mh should work.  6.4mh for 400hz.  If you get a 5mh, you could wire it in series with the 1.3mh, giving you 6.3mh, and see if you like it.

[JohnA]

On 9/22/2018 at 11:20 PM, naaac said:

 

You were right! I measured it with DATS V2 and it was actually 2.6 mH - even though it was wrapped at 4.9 mH in 1954. Looks like I need to get a bigger inductor and try it out....

 

 

 

Did the 4.9 mH inductor originally have a steel hold down screw in the middle?  😉

 Try running your low-pass calcs using 20 ohms instead of 16.   The bass horn raises the woofer's impedance a lot.  Al Klappenberger showed a Belle Klipsch' bass horn doubled the DCR of a K-33-E within its operating range. 

[naaac]

15 hours ago, JohnA said:

 

 

Did the 4.9 mH inductor originally have a steel hold down screw in the middle?  😉

 Try running your low-pass calcs using 20 ohms instead of 16.   The bass horn raises the woofer's impedance a lot.  Al Klappenberger showed a Belle Klipsch' bass horn doubled the DCR of a K-33-E within its operating range. 

Ha - no it didn't. I tried adding a proper 5mH inductor in series with the 1.3mH and now the response looks much better.
Updated woofer response in blue. 

 

[DeanG Crossovers]

Response is largely dictated by the limits of the horn. You should be able to use NO inductor, and see response dropping like a rock after 400Hz. 

 

What is happening here?

[naaac]

Well - it looks similar to the response of the K33-E woofer shown on ALK's website for his universal crossover?
 

[DeanG Crossovers]

Those are electrical simulations, not acoustic plots.

 

Here are anechoic curves of both the Klipschorn and Jubilee bass bins.

 

 

 

[naaac]

Is it because when I took measurements of the three drivers, I disconnected the other two from the crossover?

[DeanG Crossovers]

I honestly don't know, but it you're measuring one section of the loudspeaker, you should be terminating the other sections with resistors. Use a 15 ohm resistor for the midrange, and an 8 ohm resistor for the tweeter.

[WMcD]

Deang,

 

Thanks for those.  Generally it seems to me they follow the AES article.

 

Could you tell us a bit about your testing conditions and microphone placement.  LMS has a gated function you might have used.

 

I'm not sure that the AES article describes the conditions.  

 

I'd think that the anechoic chamber at Hope is only valid down to 200 Hz or so.  JBL and NASA probably have larger ones.

 

Of course the issue is  to establish some "valid" or agreed to condition to test bass horns. There may be an outdoors trihedal corner at Hope.  But then no room effects.  If in a room, which room?  Smile.

 

Microphone placement is another issue.  I've seen some suggestions of putting it close to the mouth (but where?).  Maybe on the floor one, two, or three meters in front, or at some elevation.  

 

Your thoughts are very welcome.

 

WMcD

 

 

[Schu]

The Klipschorn bass bin is that close to the Jubilee?

 

The 402 is good to about 500Hz isn't it?

[DeanG Crossovers]

I'm not sure how Roy did the measurements, I had just assumed it was in the chamber. I remember being down in Hope while he was still tweaking the network, and much energy was expended moving the Jubilee in and out of the chamber.

 

What we know for certain is that the K-33 doesn't go out to 20kHz while loaded in a Klipschorn, Jubilee, LaScala, or Belle Klipsch. ;-)

[Marvel]

42 minutes ago, Deang said:

What we know for certain is that the K-33 doesn't go out to 20kHz while loaded in a Klipschorn, Jubilee, LaScala, or Belle Klipsch. 😉

 

I wouldn't think it would go out to 20k in a Cornwall, either. Even wired straight to the woofer 😉