Article: Solving the Klipschorn Throat Riddle (Edgar)

[djk]

The 50s vintage '16 ohm' EV 15WK voice coil measured 3,2 ohms DC, about the same as a current K33.

PWK made up a formula to justify the 16 ohm advertised impedance, an average value of the peaks and dips if you will.

Heyser measured the newer style Klipschorn quite thoroughly in 1986, this would apply to most models prior to the current AK4 production. While the woofer by itself is 3,2 ohms DC, the crossover inductor adds another 0,5 ohms, and the cabinet brings the minimum up to about 4,5 ohms AC impedance.

The smaller throat just boosted output in the 400hz region a bit. If you model an ideal horn in McBean's Hornresponse, you will find a bigger throat can have a dB or two more output in the bottom half octave. IOW, a bit of robbing peter to pay paul is going on.

In 1982 the transition from the type AA network to the type AK was made. This, and subsequent revisions went to a higher Q two pole filter on the woofer. A bit more output with a bit lower impedance, the overall response was a bit smoother.

US patent #4,237,340 "The smoothness of amplitude response over the range of audible frequencies that is necessary for high fidelity sound reproduction is improved by inclusion of peaking circuits in the LC crossover network of the loudspeaker system to enhance amplitude response in the regions of crossover frequencies."

"said peaking circuit means are resonant at approximately 350hz"

"L.sub.1, C.sub.1 produce a three dB rise at approximately 350 Hz"

"The peaking effect derives from the fact that the LC crossover network has lower input, or driving point, impedance near the resonant frequencies of the various peaking circuits. At these frequencies, the input impedance drops from 8 or 16 ohms to as low as 4 ohms. Consequently, conventional solid state amplifiers, which are characteristically designed to deliver their maximum output into a 4 ohm load, produce a high output near the crossover frequencies of 400 and 6,000 Hz. where the outputs of the loudspeakers are drooping. "

L1=2,5mH

C1=128µF

Actual values in the various AL, AK, and AB networks for C1 vary from 100µF~136µF, type A or AA do not have this part.

http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.htm&r=1&f=G&l=50&s1=4,237,340.PN.&OS=PN/4,237,340&RS=PN/4,237,340

[JohnA]

Interesting thread. I'm wondering if I might make any actual

improvement in bass response of a La Scala (Industrial) with more

modern drivers, higher power and compensatory EQ?

I'd like to select a more robust woofer and let the DriveRack

properly EQ the system so that more power might help extend the limits

a wee bit. I know I can't overcome the horn's limitation due

to mouth size, but I do stack the bass cabinets.

Or might I be stuck in Never-Never Land, searching for a correct

slot size? I went to Gauss woofers, in the seventies, in

University S-9 (Classic) cabinets and never felt as if I accomplished

much.

Look into a JBL 2226H.

[ZAKO]

JBL 2226 has a Fs of 40hz ......not good. 30hz to 35hz... better.

[ajsons]

Here is an example of a bass horn using a compression chamber (more or less) in the horn throat. Usually this is used for MIDS and TWEETERS but this is the first example that I've seen in a bass horn in this example, much like an Altec A7 (828). THe reported effect of this compression is that the midrange dispersion characteristics are pronounced horizontally in a wider pattern. The compression chamber is clearly NOT used in a Khorn, nor is it the case for the acoustic filter slot.

DM

Dana,

Here's one that pre-dates the Tattersall patent (1987) by about 33 years.. It is a closet horn (never heard of that one) from 1954. Looks like the U.S. had that convergent-divergent throat first before the Aussies.

Armando

[ajsons]

It's an interesting construction article so I'm posting the other pages.

[ajsons]

I wish I have two closets that can hold a couple of these.

[D-MAN]

Armando, yes - that's a new one on me. I wonder if anything was mentioned about the purpose of the converging throat?

Tattersall did it specifically to increase upper-bass and midrange dispersion...

Dana

[D-MAN]

An extra note on the Khorn so-called "throat riddle"...

Remember we are talking about the throat cavity opening and NOT the horn throat per se. The throat cross-section remains 78 sq. inches or 6x13" (Khorn and my horn).

I tried the K33E's on a larger (6x13") slot size and didn't like them at all.

I tried the CB15's on a smaller (3x13") slot size and didn't like them at all.

I spent some time swapping out motor boards yesterday, and after all is said and done, it is back together the way it was to start with - none of my time-consuming experiments resulted in any improvement...

The K33E is a somewhat strange animal in that, according to the Keele formulas, it is most efficient with a 115 sq. in. opening and a back chamber of 5.4 cu. feet! However, it apparently sounds best in the same horn with a 39 sq. in. opening and a back chamber of less than 2.8 cu. feet instead (Khorn)!

The CB15 was chosen specifically for the 6x13" throat cross-section of my horn using Keele's formulas for throat cross-section, back chamber and frequency corners. I am struck by the difference between the two drivers (CB15 and K33E) when the throat cavity parameters are altered.

The efficiency of the driver should become reduced with a smaller slot, of course. In my experience, the bandwidth itself doesn't appear to widen, as one would expect if one was altering the throat cross-section by making it smaller (we are only altering the cavity opening, the horn remains constant). That would also make the horn longer by necessity. Where I find the K33E doesn't apparently become less loud (slight - maybe), it definitely "tightens up" its overall response with the narrower slot size. This result is not predicted behavior IMO.

But since we are technically NOT reducing the throat cross-section, the effects of a smaller horn-feeding slot size is still somewhat up in the air. It apparently doesn't particularily act as a compression chamber, either, although technically, some aspects of that would seem to be present. If it did, then the higher end of the bandpass would be reduced somewhat, and in the case of the Khorn/K33E combination, it has slight but opposite effect. Although it would appear that a front compression chamber exists in the Khorn with a smaller slot size than the horn throat cross-section, my guess is that it isn't, based on the fact that a high-end peak was added to the response around 400Hz.

I cannot put my finger on the T/S parameters which would indicate the behavior of the K33E in the Khorn. None really seem to apply - I've come to the conclusion that this is due to the level of efficiency the K33E is forced to operate at in the Khorn configuration.

In the case where one is merely making the opening TO THE THROAT smaller, the effect is to reduce the efficiency and (I assume MOST drivers other than the K33E), the bandpass is also reduced, fairly proportionately with the drop in efficiency, or so it seems.

I assume that the behavior of the K33E under these circumstances can be explained by the fact that the Keele formulas for optimal efficiency technically do not apply. The K33E in the Khorn or an equivalent is operating somewhere quite a bit below its optimum efficiency, where the Keele formulas are no longer valid. It's behavior when loaded into a Khorn cannot be accurately modelled using formulas based on maximum efficiency and widest bandwidth.

On the other hand, the drivers specifically chosen to meet the requirements of maximum efficiency and widest bandwith in a given horn application WILL tend to follow the predicted behavior (I used Keele's formulas). I can verify that the CB15 behavior I have experienced follows Keele's predictions quite well.

DM