All around best 8-ohm substitute for K-33?

[jwc]

Doc. Just to give you confidence in what you are plotting in WinISD, I did test tones on True RTA with the K33 in the Cornwall cabinet. This is with the Cornscalas I have and the Cornscala dbb (port split). The actual response was just like those peaky, non-flat curves we keep getting. Howevr, the sound is the usual Cornwallbass if that is appealing.

So, I believe the specs published by Eminence for the new K33

[DrWho]

Did you use actual test tones or pink noise? Sine waves will really be

exagerated by room effects which is why I'm asking (I can get you free

software that makes more advanced test "tones" if you're interested).

The more I think about it and you partially confirm this....the

exageration in the cornwall bass is what makes them sound better at

lower volumes.

Btw, has anyone ever noticed that the Cornwall II on the website says that it uses a K-34-E? I wonder what the difference is....(is it only the dustcap change or perhaps something more?)

This is kinda off topic, but I really think the Chorus II uses the

eminence Magnum 15 LF. They model identical to the posted response

charts of the Chorus II and they look insanely similar! I think I might

just have to open up one of me speakers to give them an inside look. I

know this may seem like a stretch, but the chorus II sounds just how I

would imagine the Magnum model sounding. During my chorus versus

cornwall listening test I noticed that the chorus really needed to be

cranked to fill out the low end (cranked as in over 90dB)...and this is

probably due to the flatter response and the effects of equal loudness.

For my listening habits and average SPL's I think I would want to go with a system somewhere in the middle of the two.

[D-MAN]

So translated you don't like the size of the cornwall cabinet and you want to get louder than 120dB max (which of course will reduce power compression) []

Did I say that? well, maybe.

SO is the question really "do we specialize or generalize"?

Actually, the specific attributes of control over the diaphragm are different for cones subject to radiating "against" the atmosphere than inside of a pressure chamber produced by horn loading.

The attributes of a particular driver that are BEST SUITED for one application tend to be detriments when used in the other application. This is not to say that there is no ability to crossover between applications, of course. However, it would be expected that optimum results can be achieved by selecting for and applying the attributes that best perform optimally for a particular application than to choose a driver that forms a compromise between the two (unless that is your purpose).

As a practical experiment, a pair of drivers (a "close" T/S match to the K33E) is going to replace the K33E's in a 1986 CW. We'll have to see if the owner prefers one over the other. However, the "new" driver has some of the attributes that would theoretically give it the edge in the direct radiator role. If what I said is true, then the outcome SHOULD be in accordance with what I've stated above. If the K33E stays, then IGNORE what I said here. We'll have to wait and find out.

DM

[jwc]

Yes. I did sine waves, clean sweep and I believe pink and white noise. One drawback was that it was tested indoors.

I purchased the 1/24th octave TrueRTA.

[DrWho]

So translated you don't like

the size of the cornwall cabinet and you want to get louder than 120dB

max (which of course will reduce power compression) []

Did I say that? well, maybe.

SO is the question really "do we specialize or generalize"?

Actually,

the specific attributes of control over the diaphragm are different for

cones subject to radiating "against" the atmosphere than inside of a

pressure chamber produced by horn loading.

The attributes of a

particular driver that are BEST SUITED for one application tend to be

detriments when used in the other application. This is not to say that

there is no ability to crossover between applications, of course.

However, it would be expected that optimum results can be achieved

by selecting for and applying the attributes that best

perform optimally for a particular application than to choose a driver

that forms a compromise between the two (unless that is your purpose).

As

a practical experiment, a pair of drivers (a "close" T/S match to

the K33E) is going to replace the K33E's in a 1986 CW. We'll have to

see if the owner prefers one over the other. However, the "new" driver

has some of the attributes that would theoretically give it the edge in

the direct radiator role. If what I said is true, then the outcome

SHOULD be in accordance with what I've stated above. If the K33E stays,

then IGNORE what I said here. We'll have to wait and find out.

I'm just asking you to quantify the differences instead of holding to

the premise that something optimized for one purpose can't be optimized

for another...Maybe PWK wanted to build a cabinet with the exact

dimensions of the cornwall and the K33 in his khorns just happened to

fit perfectly []

[Tom Mobley]

here the eminence web site's graph for the Magnum 15 LF: if I can get it to paste)

[Tom Mobley]

to me that graph looks pretty good. why is it listed 4.95 when it's supposed to be an 8-ohm driver? shouldn't it be up around 6?

37Hz

4.95 ohms

1.21mH

0.48

9.42

0.45

163 liters

5.76 cu. ft.

677cc

0.1579mm/N

16.88 T-M

119 grams

2.919N*sec/M

76.8

7.9mm

856cm²

122 ohms

30.94mm

11.3mm*

[Deang]

Probably the DCR of the voice coil.

Tom, would you please PM or email your phone number.

[DrWho]

to me that graph looks

pretty good. why is it listed 4.95 when it's supposed to be

an 8-ohm driver? shouldn't it be up around 6?

37Hz

4.95 ohms

1.21mH

0.48

9.42

0.45

163 liters

5.76 cu. ft.

677cc

0.1579mm/N

16.88 T-M

119 grams

2.919N*sec/M

76.8

7.9mm

856cm²

122 ohms

30.94mm

11.3mm*

The Re of the driver is different from it's actual impedance. If I'm

not mistaken the Re is the DC resistance of the voice coil whereas the

"nominal impedance" is an average over many frequencies (DC resistance

is different from AC impedance). Here's a plot of the impedance when in

an 8 cubic foot cabinet tuned to 35Hz

[Deang]

I hear an echo.

That's going to need a Zobel.

[michael hurd]

K-33 and K-34 have the same specifications, just the cosmetics are slighlty different. ( forget who posted that info, maybe freezin )

[Tom Mobley]

Dean, you have PM

[Deang]

Hey, turn on yer phone.

[Tom Mobley]

Yo, dude, your number is blocked and you didn't leave a msg. Sorry, I got to it on the last ring.

[DrWho]

I hear an echo.

That's going to need a Zobel.

Seems we posted at the same time []

Btw, what the heck is a zobel and why is it needed? (I'm guessing it

has something to do with changing the apparent impedance to the

amplifier...aka, getting rid of those huge peaks).

[Deang]

The second one is the one that has to be dealt with. That looks pretty hairy, are you sure about that plot?

[Marvel]

A Zobel is a series resistor-capacitor

(R-C) network that is connected in parallel with a loudspeaker driver in order to

neutralize the effects of the drivers voice coil inductance L(e).

I found this on the web. Is that accurate, in a nutshell, the straight poop?

[Jay L]

you got it marvel

[DrWho]

The second one is the one that has to be dealt

with. That looks pretty hairy, are you sure about that plot?

Well it's what WinISD spits out...

Btw, I never thought it was a big concern because you always measure

huge impedance spikes on both sides of the tuning point. I always see

numbers being thrown around on the order of 40ohms and this is 50 ohms

if I remember correctly so nothing hugely different. The K-33 has

spikes half the size btw (and I guess that makes sense...one is an 8

ohm driver and the other is 4).

[D-MAN]

Reflex Drivers

===========

For dual-driver reflex duty, I had a pair of Eminence Kappa 15LF's (used to be $90 ea) which were totally kick-butt drivers - they would change your heartbeat at 30 watts! However: they are sensitivity rated at 99db, though, so I could not get my high-end to match them, so I bailed on them for horn use. But if I were starting from scratch particularily for reflex drivers, these are just the ticket for a reasonable cost. I gave these away with my first prototype horns (the nasty plywood jobs-remember those?) and they are still kicking tail and getting the cops called.

Horn Drivers

==========

Some thoughts on horn-specific driver swapping...

I'm not sure that there is such a thing as a best all-around replacement!

It depends on 2 things, really:

1) How much money are you willing to throw at the problem

2) what areas of performance are you seeking to increase/decrease

Once that criteria is decided, then the hunt can begin.

Sensitivity

========

One of the BIGGIES, the rated sensitivity for a particular bass horn driver will influence the selection of ALL of the other drivers and crossover networks. In my experience, I selected only drivers with a somewhat related value to the K33E (96+db) with 97 being the upper limit, to keep the other drivers (and the crossovers) the same as I had. The important thing to think about is that manufacturer's stated sensitivity is frequency related and there is no real standard for measuring it in the industry. Be prepared for variances and try to error on the "upper" side but not by too much!

Frequency Response

=================

This is a real conundrum for a bass horn driver. Published specs are virtually meaningless in such a case. The Fs is a main concern, preferably the lower is better. Upper frequency corner is more cone-mass and magnet size related. The Fs selected should be lower than the Fc of the horn (the Khorn is stated as 38Hz (in the Jubilee paper) and 40Hz in other Klipsch brochures). You want to be below that by as much possible without getting "sloppy" on the suspension. Typically, this means staying away from roll-type surrounds. The Keele formulas for Fcl and Fcu are going to be of use in determining the upper and lower frequency response corners, instead of relying on a published frequency response curve for the driver.

Qts

==========

.30 or below (preferred) is highly desirable for a horn driver. Represents good control over the cone and is indicative of a strong magnet/light cone relationship.

G/mm

==========

Cone mass - in general the lighter the better for horn use. Direct radiators typically have a heavier cone. Look for this as a quick determiner as to its "best" application.

Throat Size

=========

The drivers will have a certain throat size (or opening) that they will be most efficient at, using Keele's T/S horn formulas. The Khorn can support up to a 78 sq. in. maximum throat opening. This is the maximum upper limit.

Vb

=======

Back chamber volume - the Khorn is approx. 4800+ cu. inches (figure around 3 cu. ft.) for anullment purposes. Alot of modern drivers will do better with less volume, so that could be an issue. Undersizing the Vb is typical (PWK did it) upto 20% in some cases. It raises the Fs though.

Extension

=========

Believe it or not, I do not find a great deal of correlation between a great deal of extension or not. I ended up preferring a driver with a Zmax of 4.8 mm (considered "normal") over the K33E (over 8mm) and others with a 6.O mm. My opinion, this is one of the less critical criterias. It MIGHT come into play at EXTREME volumes, but hopefully, you will never have to go there when loaded into the horn.

Cast or Steel Frame?

=================

Cast aluminum tends to be lighter and less subject to "ringing" than a steel frame. Get a "heavy" frame regardless. Cast frames tend to have less reflections from frame elements behind the cone due to the reduced size of the struts employed. This varies, of course.

Vc

=========

8 Ohm voice coils dominate the market. However, DC resistance/AC impedance varies widely and is frequency related. You may find that there is no relationship to reality with the manufacturer's published specs. Dual driver setups would typically use 2-8ohms wired in parallel.

Opinion

=========

Does a low-Q driver sound better? My findings indicate YES - I ended up choosing the lowest Q driver in the bunch as my favorite.

Does extended Zmax matter? Not in my experience. I chose the driver with the least extension out of the group.

Does a cast frame sound better? That I don't know, but I ended up choosing a cast frame driver. I would guess that it is less reflective.

Does 8 Ohm vc's sound different compared to 4? No difference that I can tell.

Does a lighter cone make a difference? I ended up preferring a driver with a light cone compared to others in the group, but I really cannot say what I think the difference actually is.

What area of T/S specs matter the most? They are all tightly interrelated, so NONE stand out or are anymore important than the others.

Rule of thumb: Low Qts (.3 or below), large magnet, light cone, low Fs, St (throat size=highest efficiency match), Vb match.

Keele's work will get you pretty darn close to what you want. His T/S formulas are extremely helpful in this regard. All of the drivers I selected this way would technically qualify as being "keepers" - however, within that group the performance variations were such that I preferred some attributes over others. Fine tuning, so to speak.

It remains interesting to me that the K33E would NOT be chosen for the horn by its T/S parameters! There is a great deal of leeway here!

DM

P.S. check out this provocative little number (Madisound PRO-AUDIO 15).