Improvements to the Khorn design

[greg928gts]

My K Horn project will use an adjustable top to angle the mid and tweeter at the sweet spot.

This is going to be part of my new design. Separately adjustable midrange and tweeter positioning/aiming.

Other designs that do not need corners are in my future for sure, but right now, I'm concentrating on the fully horn-loaded, corner speaker design.

Greg

[greg928gts]

Greg,

You've already seen the following in a private email message, but the other forum members have not.

Both the KHorn and the Jubilee have approximately 400° of folding angles. They also have split mouthes, which cause "double-slit" diffraction well within their upper frequency ranges.

Reduce the folding to two 90° bends, and eliminate the split mouth. Use two Altec/GPA 515-8GHP woofers in a cabinet approximately the same size and shape as the KHorn. Do that, and you'll have something that looks like the attached rendering.

Optimize the design to compete with the standard KHorn, and the first Hornresp plot results (KHorn in gray).

Optimize the design to compete with the standard Jubilee, and the second Hornresp plot results (Jubilee in gray).

(another) Greg

Greg, I'm going to email you regarding my new design, but I wanted to start here first. I like your bass bin design, but I have some questions about them, which I will ask directly.

Greg-R

[greg928gts]

Greg, you built a pair of Dana's folded horns, didn't you? What was the result of those?

Bruce

Yes, Dana's horns are the Jamboree's, and they are the nicest sounding bass bins I own. There is a patent on them, but I've talked to Dana about that, and he's open to letting me build and sell them for a nominal fee per pair. I may use them someday, but for now, I'm concentrating on the Khorn design.

I think there is enough room for improvement in the design and construction of the Khorn bass bin, that the gap between a typical Khorn bass bin and the Jubilee/Jamboree can be closed significantly.

Greg

[deafbykhorns]

It has already been done,,, from 18hz to 22k hz +- 3db ,,Two way and looks great,,, TANNOY Westminster Royal.

Those are beautiful speakers, I have you heard these before?

Greg,

Someone mentioned time align drivers, I think this a weakness and would be a great improvement. See http://www.phasetech.com/darts-new.html,this system uses a digital amplifier to time align each and every driver in the system. The digital amplifier is also built by a well known high end company (which I swore never to mention). I heard this system last year and it was absolutly amazing. If this could be applied to my Khorns, I would purchase it in a heartbeat. An active system in lieu of a digital amplifier would be better since some of us like to use tubes.

[Coytee]

An active system in lieu of a digital amplifier would be better since some of us like to use tubes.

Is there any possible way to get signal alignment without using a digital crossover? (I think Greg has a predisposition towards passives and away from a digital crossover, I don't know about analog active)

I don't know enough about this stuff to know if you can use an analog active to delay the signal.

[greg928gts]

Someone mentioned time align drivers, I think this a weakness and would be a great improvement. See http://www.phasetech.com/darts-new.html,this system uses a digital amplifier to time align each and every driver in the system. The digital amplifier is also built by a well known high end company (which I swore never to mention). I heard this system last year and it was absolutly amazing. If this could be applied to my Khorns, I would purchase it in a heartbeat. An active system in lieu of a digital amplifier would be better since some of us like to use tubes.

For now, I'm going to leave the electronics in the speakers passive, but they will be bi-amp or tri-amp friendly. The crossovers will be easily accessible and will have the option of being displayed if the owner chooses.

Greg

[Invidiosulus]

I don't know how right it is but my understanding of some of the articles by Bruce Edgar would indicate that the K33 is not as ideal for horn loading as one could hope.

Bruce Edgar determines the horn loading bandwidth as being the frequency range between the driver Fs and the mass rolloff frequency.

The mass rolloff frequency(Fhm) is determined(according to Keele) as being Fhm = 2fs/Qes

For the K33 that gives us,

2 x 34Hz = 68Hz

68Hz / .410(Qes) = 165.something Hz

165 Hz is it would seem the upper end of where the driver is effectively horn loaded.

A driver with Lower Qes or a slightly higher Fs or both would seem to have a better bandwidth for horn loading.

There are also calculations that are supposed to work out the ideal throat size(referred to as St).

St = 0.8FsQesVas

With the numbers from the K33 it looks like the ideal throat size would be 120 sq inches.

Again a driver with a lower Qes seems like it would work better in the Khorn.

I haven't taken the time to work out the "ideal" back chamber volume yet but having a correct back chamber can have an impact on how effective the driver is at playing down near the horns cutoff frequency.

I think I'm in over my head.

-Josh

[deafbykhorns]

Josh

Back in the 80's, we used to sell RCF drivers and speakers. One of the reps recommended I try one of their drivers in my Khorns. I remember having better bottom end but we didnt have Al K. designing networks with multiple taps so they were out of balance. And they were expensive at the time.

[jwc]

Greg.

What is your max tolerable footprint and height of the bass bin?

jc

[greg928gts]

Thanks Josh, good stuff. I get in over my head all the time!

Greg

[greg928gts]

Greg.

What is your max tolerable footprint and height of the bass bin?

jc

I would say no higher than the current Khorn, but the footprint could increase some - maybe 10 - 20% and still be workable.

Looking at the Jamboree design, the front of the bass bin could actually tilt down and offer a place for the front of the mid horn to rest. This would allow a taller bass bin, but still have the mid horn at the same or lower height. If the design allowed, this might be possible to do with the Khorn, although it doesn't look like it to me without a horn pathway change.

Greg

[Invidiosulus]

I would think that part of the problem with having a larger throat in the Khorn design is that you don't have room for a large enough 45 degree splitter directly opposite the throat since the front panel is only about 3 inches away.

Since the splitter indicates a hight of the St of 3 inches I would think you would want to maintain a St of less than about 42sq inches.

B&C has some drivers that have a nice low Qes that gets the Fhm up to around 350Hz and the St is around 34sq inches.

http://www.parts-express.com/pe/showdetl.cfm?Partnumber=294-669

They are an 8ohm driver so they would certainly need a reworked crossover besides the additional calculating of back chamber etc...

-Josh

[jwc]

Greg.

Here are my thoughts.

First of all, I am excited to see what you are doing. I wish I had the time to do the type of thing you do.......but I don't have a shred of time for customer service of something I were to make. This is why I stop building computers for sale.

This forum has all kinds of respect for the khorn. I'm sure we would love to see it get better and to see it grow after the passing of PWK. However, I'm thinking that the man did everything he could to make the Khorn:

1) sell

2) make a profit

3) improve the peformance in that FOOTPRINT.

I put in that order for a reason. There were a lot of speaker maker gurus in his time but he was able to land the speaker into your living room and make profit. So there was a mix of all the above mentioned to make PWK as great as he was...and still is to us.

If you read the AES on the Jube....you can see he was trying to improve performance but keep the same footprint. That is the key in my mind. I don't think anyone here on this forum is better than PWK to get more of what you want in that footprint for a single 15" K33.

You can take ideas and try them...but you might do a lot of building and testing. He was selling to the masses in his day....Unnfortunatley (sp)....with the Ipod GEN....small sells are made on some of his concepts.

I do think the jubilee bests the Khorn....easily in my mind. Having said that, I'm not afraid to say that I think the Jamboree is right in there...easily better than the khorn and a match for the jube.

So with the above....my vote is to attempt a dual driver design....consider what footprint you want to accept...and go with that..then build the rest upon. The foot print is the key....

24' depth only allows so much you can do...."physics" as Dana would say. He built the jamboree on this footprint. Yes....he could draw up a longer horn to trounce the jubilee....Heck...the geektard/rookie I am...I can do way more with 30" depth....but that usual won't sell (I don't have a problem with it).

You know....as I'm posting this...can we make a shout out to get Dana and Ajsons back on here. This forum was so much more interesting when the both of them were here. I know they have there "issues", ......but hey...I loved both of their posts...

jc

[Deang]

The Jubilee is superior to the Klipschorn, no question about it. It curves better, has lower distortion, and allows some flexibility with placement. It sounds better too.

If we pretend that there is no Jubilee, I would have to say that the one that brings the most improvement to the Klipschorn bass bin is to close off the back like the 60th Anniversary Edition. I don't see much point in pursuing further improvement, it's unlikely you're going to do anything that hasn't been tried before -- you can make it curve/sound different, but that's not the same as making it better. I mean, even PK threw in the towel.

Like you said, the Jubilee, patent or not -- belongs to Klipsch. I think the same can be applied to the Klipschorn. If Dana is willing to let you build the Jamboree, that's probably the best direction to go.I would call it something different though -- "Jamboree" doesn't work for me.

[Edgar]

There were a lot of speaker maker gurus in his time but he was able to land the speaker into your living room and make profit. So there was a mix of all the above mentioned to make PWK as great as he was...and still is to us.

JC, I couldn't agree more. PWK succeeded in creating a loudspeaker that fit fairly unobtrusively into a moderate-sized living room, and looked GOOD. Browse through the designs shown here and you'll see what I mean -- they're universally UGLY, at least to non-audiophiles.

Furthermore, he got a LOT of performance out of a relatively small horn volume. I design speaker cabinets for fun. For a long time I have been trying to get more performance out of the same volume as the KHorn, and, frankly, it's darned-near impossible. And I have my CAD program to change designs at will; PWK had pencil, paper, and woodworking tools. The fact that the KHorn still stands up really well 60+ years later is quite a testament.

If you read the AES on the Jube....you can see he was trying to improve performance but keep the same footprint. That is the key in my mind. I don't think anyone here on this forum is better than PWK to get more of what you want in that footprint for a single 15" K33. (...) So with the above....my vote is to attempt a dual driver design....consider what footprint you want to accept...and go with that..then build the rest upon.

Again, I agree. But I don't understand why PWK limited himself to 12" drivers in the Jubilee. Perhaps it was because 15" drivers typically can't handle the midrange frequencies as well as 12" drivers, due to mass rolloff, but there are 15" drivers that do much better in that region than even the 12" K31! I am referring, in particular, to drivers from Altec, B&C, BD Designs, Ciare, and even EV. If they can design a better driver, PWK and Roy could have, too.

I have noticed something about the design of the KHorn and Jubilee that may offer an explanation, though. In both designs, they seem to have tried to extend the LF response by perhaps 1/4 octave by resonating the driver below the horn cutoff. That might explain the use of high-mass, low fs drivers like the K33 and K31, but it comes at the expense of poorer midrange response. Nowadays, with so many high quality subwoofers available, I wonder if the tradeoff can still be justified.

Greg

[DrWho]

Both the KHorn and the Jubilee have
approximately 400° of folding angles. They also have split mouthes,
which cause "double-slit" diffraction well within their upper frequency
ranges.

Could you expand a bit on the "double-slit"
diffraction? Are you worried about distortion, or some abberation in
the response resulting from the bifurcation?

Reduce the folding to two 90° bends, and eliminate the
split mouth. Use two Altec/GPA 515-8GHP woofers in a cabinet
approximately the same size and shape as the KHorn. Do that, and you'll
have something that looks like the attached rendering.

What polar response do you expect with this design?

Recently
I've been spending a lot of time reading about and trying to understand
how to predict the polar response of horns. It seems to me that the
biggest advantage to bifurcation is the polar response that can be
achieved within a limited footprint, and the reduction in radius
difference when bends have to be made.

It's my understanding that
the Huygen reflector (90 deg bend) only works when the wavelength is
small relative to the size of the curve. When the wavelength is large,
you get more of a wrapping wavefront than a distinct reflection (kinda
like fluid flow models). Also, it seems like the Huygen reflector can't
be proplerly implemented without introducing a discontinuity in the
area expansion. I suppose at very small wavelengths it doesn't matter
because the sound has already left the horn walls, but it's gonna
introduce some passband ripple. I also don't think the higher
frequencies should leave the horn walls until the sound is in the final
flare, that way you don't start introducing the cavern effect as the
sound tries to get out of the "cave".

I think the biggest
downside to the Jubilee is that the splayed mouth behaves more like two
different acoustic sources, which starts to introduce a clover-leafing
polar response above whatever frequency has the same wavelength as the
width of the front. I don't have the dimensions or polar measurements
to say exactly where, but I'm guessing it's around 400Hz or so.
However, this width offers more pattern control at lower frequencies,
and the proximity of each mouth near the corner boundary provides a
smoother impedance match to the room corner, and you don't have to
worry about the early reflections that will introduce comb-filtering.

My
personal goal for a corner loaded horn involves moving the bandwidth of
the LF section up a bit with the notion of going 3-way to cover the
full 20-20kHz spectrum. A hornloaded sub that does 20-80Hz, a midbass
horn that covers 80-800Hz, and then a strong HF section that covers
800Hz - 20kHz. The goal for the HF would be 90x60 polars transitioning
into 90x60 polars on the midbass (which I'm torn between a straight or
folded horn) and then 80Hz and below will rely on the room corners for
the pattern control (or lack thereof), which puts you at around 90x90
until the first vertical room mode. Trying to do 90x60 down to 80Hz
requires the final build length to be under 30" so it can fit through
doors since the mouth width and height dimensions will need to be too
big.

I think another problem with the khorn is that it uses an
exponential expansion. I'm not sure that's the most optimized expansion
rate for impedance matching the driver to free air.

[Marvel]

Greg, you built a pair of Dana's folded horns, didn't you? What was the result of those?

Bruce

Yes, Dana's horns are the Jamboree's, and they are the nicest sounding bass bins I own. There is a patent on them, but I've talked to Dana about that, and he's open to letting me build and sell them for a nominal fee per pair. I may use them someday, but for now, I'm concentrating on the Khorn design.

Greg

Is the Jamboree the same as the Ascent, or another model. I have a set of the Ascent plans, and liked the design a lot.

Bruce

[DrWho]

Again, I agree. But I don't understand why PWK
limited himself to 12" drivers in the Jubilee. Perhaps it was because
15" drivers typically can't handle the midrange frequencies as well as
12" drivers, due to mass rolloff, but there are 15" drivers that do
much better in that region than even the 12" K31! I am referring, in
particular, to drivers from Altec, B&C, BD Designs, Ciare, and even
EV. If they can design a better driver, PWK and Roy could have, too.


I have noticed something about the design of the KHorn and Jubilee that
may offer an explanation, though. In both designs, they seem to have
tried to extend the LF response by perhaps 1/4 octave by resonating the
driver below the horn cutoff. That might explain the use of high-mass,
low fs drivers like the K33 and K31, but it comes at the expense of
poorer midrange response. Nowadays, with so many high quality
subwoofers available, I wonder if the tradeoff can still be
justified.

I do know they started with a 15" driver in the Jubilee, but got better performance with the dual 12's.

I
think the reason for the extra mass on the drivers is in anticipation
of the additional loading offerred by the room corner. I believe it
also allows for a slightly smaller rear chamber too. I see the expense
of midrange extension as more of a shifting of the usable/intended
bandwidth of the horn. If your tweeter or squawker is gonna cover the
sacrificed range, then it's a wash from the system perspective.

However, I do agree that there are better
drivers available...but that's never been the Klipsch way [] I also
wonder if the K-31 was used for business reasons since it
already existed and was good enough for the application. Roy has
actually hinted
a few times that he would like to upgrade the driver to a 4" voice
coil. I'm sure he would probably add a little more excursion while he
was at it too.

[Edgar]

Could you expand a bit on the "double-slit"

diffraction? Are you worried about distortion, or some abberation in

the response resulting from the bifurcation?

It's not distortion. It's exactly the kind of "clover-leafing" polar pattern that you mention later in your post. See here.

What polar response do you expect with this design?

I'm trying to keep the horizontal response as nearly omnidirectional as possible. To do that, the mouth dimension needs to be reasonably small relative to the wavelength.

Recently

I've been spending a lot of time reading about and trying to understand

how to predict the polar response of horns. It seems to me that the

biggest advantage to bifurcation is the polar response that can be

achieved within a limited footprint,

With the bifurcated design, the polar response narrows into a central lobe as the frequency rises. If the distance between the mouths is much larger than half the wavelength, then multiple lobes form.

and the reduction in radius

difference when bends have to be made.

It's my understanding that

the Huygen reflector (90 deg bend) only works when the wavelength is

small relative to the size of the curve. When the wavelength is large,

you get more of a wrapping wavefront than a distinct reflection (kinda

like fluid flow models).

Essentially true. Reflectors become effective when the width of the acoustic path in the region of the bend is a significant fraction of half the wavelength. See here.

Also, it seems like the Huygen reflector can't

be proplerly implemented without introducing a discontinuity in the

area expansion. I suppose at very small wavelengths it doesn't matter

because the sound has already left the horn walls, but it's gonna

introduce some passband ripple. I also don't think the higher

frequencies should leave the horn walls until the sound is in the final

flare, that way you don't start introducing the cavern effect as the

sound tries to get out of the "cave".

Also true, which makes it advantageous to minimize the number of folds in the horn.

I think the biggest

downside to the Jubilee is that the splayed mouth behaves more like two

different acoustic sources, which starts to introduce a clover-leafing

polar response above whatever frequency has the same wavelength as the

width of the front. I don't have the dimensions or polar measurements

to say exactly where, but I'm guessing it's around 400Hz or so.

The calculations are straightforward. In the KHorn it starts around 200 Hz; in the Jubilee around 300 Hz.

However, this width offers more pattern control at lower frequencies,

and the proximity of each mouth near the corner boundary provides a

smoother impedance match to the room corner, and you don't have to

worry about the early reflections that will introduce comb-filtering.

I'm not convinced that pattern control is a big issue when radiating into half-PI space.

My

personal goal for a corner loaded horn involves moving the bandwidth of

the LF section up a bit with the notion of going 3-way to cover the

full 20-20kHz spectrum. A hornloaded sub that does 20-80Hz, a midbass

horn that covers 80-800Hz, and then a strong HF section that covers

800Hz - 20kHz. The goal for the HF would be 90x60 polars transitioning

into 90x60 polars on the midbass (which I'm torn between a straight or

folded horn) and then 80Hz and below will rely on the room corners for

the pattern control (or lack thereof), which puts you at around 90x90

until the first vertical room mode. Trying to do 90x60 down to 80Hz

requires the final build length to be under 30" so it can fit through

doors since the mouth width and height dimensions will need to be too

big.

I can't argue with anything you've presented here. I might fine-tune some of the numbers to my own liking, but basically I think you're on a good path.

I think another problem with the khorn is that it uses an

exponential expansion. I'm not sure that's the most optimized expansion

rate for impedance matching the driver to free air.

Frankly, for such a short horn, it doesn't matter very much.

Greg

[mikebse2a3]

However, I do agree that there are better drivers available...but that's never been the Klipsch way I also wonder if the K-31 was used for business reasons since it already existed and was good enough for the application. Roy has actually hinted a few times that he would like to upgrade the driver to a 4" voice coil. I'm sure he would probably add a little more excursion while he was at it too.

Anyone who believes Klipsch/Roy/PWK used a particular driver because it was cheap or what was in stock, IMO is seriously under estimating them and their reasons for the choices they have made.

Mike, I know Roy has mentioned that it "might be possible" to gain a little performance with a correctly chosen (4" voice coil) driver but until it is tried/tested and proven it's just a thought on what else we might can do to squeeze a little extra performance from it.

To be honest, I see alot of modeling being used and theory being talked of (I have no doubt that Klipsch and Roy have their own modeling and design formulas reached after decades of research and actual prototypes being built and measured/listened to and finally made into successfull finished products) but something I believe is very important to remember is what Klipsch does "that as far as I'm aware of no one here in this do-it-yourself area of the forum has shown the real capability to do on their own" is verify by measurements in an anechoic chamber and other advance measurement methods that they have reached their design goals! That they have actually reduced distortions and met their other performance goals and thus made a real improvement and not just a different and maybe even nice sounding loudspeaker.

Please understand I'm not saying this against anyone personelly but more as a reality check of what is really happening in this part of the forum. Measuements used properly are just as vital as controled listening test if real world advances are to be made!

mike tn[]