 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
There are some posts in this forum and in Full range one about Karlson boxes. A good url was pointed :http://home.student.utwente.nl/e.j.j.ulfman/
In the Articles link, there are one of a German man : Mr. Beck which analysed Karlson coupler and noticed Karlson Tube for compression driver.
The result with a TAD seems to be very good, the frequency respons is closed to an exponential horn or Tractrix horn, no need of equalisation. The tube looks easy to DYI building, very simpler compared to all DYI horns which could be found in the WEB. But I didn't find any more people using this sort of coupler other a german speakers.
For Public adress, i suppose that the trouble compared to horn design is the directivity which could not be controled as with a horn. OK !
But for HI-FI used, I ask to the group, what are troubles with Karlson tube for loading a compression driver, and also for loading speaker, in general ?, A design similar to Hedlund horn, could be imagined to back loadind a FR, but with a Karlson Tube.
Thanks to all
Eric
Follow Ups:
Hi Eric -I wish I could examine the slit tube with regard to acoustic input Z and loading - for what its worth you can see that Ernst Beck manufactured several injection-molded tractrix horns 20 years ago and has plenty of 'real-world' experience with acoustic devices.
I recently tried a 'one cut' paper K-coupler on a Eminence horn driver - the cut was a guess and probably misses an exponential expansion. (Looking at Beck's pipe it's around 1/4 wave @~1Khz and has a 1" ID)
My impressions are that the device is usable - does not have the gain of a large horn but does exhibit a wide horizontal dispersion with little timbre change throughout the wide angle - I had to tilt my 'toy' tube a bit forward to achieve some balance - you can see that Mr. Beck's arrangenent had the tube and TAD 4001 driver on pivots which allowed for angling/'aiming'
I need to make my K-coupler better - as it stands mine is only 2-unlaminated thicknesses of printer paper so its lossy - even so it plays quite loud - I didn't bother to either measure the length (which turned out to be ~6.5") or calculate an exponential taper. Even so it sounds resonably good - on some materal such as harpsichord the effect is very nice.
It's not a horn but I do think an idea worth trying - there is one gentleman in the Karlson forum working away on a program to calculate the throat Z of the K-couplers, and I think given time, more practical understanding of how to make K-couplers will be available.
In PA work you can see that Beck suggested the Community M4 driver for a midrange K-coupler - for a home experiment I think something like a JBL LE-5 or possibly small Fostex speaker could be used with some 4" pvc pipe - possibly 15-16" long for midrange.
I think by the late 1950's John E. Karlson had PA models - I wish I had the literature to see what products and when... The late 1960's brought on the Karlson Asymmetric Projector series - the 15" model AP-100 Karlson-(Oliver) Klam was 45" deep, 20.75" wide, and 18" high and used the Altec 421. After Karlson's death (I believe 1972)I think the Klams were pulled out of manufacturing - I don't really know the story...
I hope to play with this idea more - as it is I can enjoy most music with the K-coupler tweeter on top of a 12" Karlson (loaded with EV SP-12)- you can experiment like I did using paper - I'm sure by laminating several layers of paper that the tube will be sufficently rigid. It should be fun - with a cheap driver like mine you may have to play wih the tilt and possibly use a series pad reiestor with cap-bypass to get the HF balance,
best wishes
Freddyps - after listening to the slit-tube - I kinda laugh that it works as well as it does!
Forgot to mention that in my 2-way setup the paper K-coupler is using only 1.25uF on the 16 ohm driver- easily keeps up with the EV SP-12. I suspect it might get pretty darn loud if crossed over lower. That's when I need pvc instead of 2-sheets of psper!Freddy
Hi Erik and Freddy!Freddy has made several mentions of the K-Coupler on this forum, and has actively experimented with this design. He and I had a discussion several weeks ago about them, and I suggested the a whitepaper available on an earlier, archived post:
http://www.audioasylum.com/forums/HUG/messages/6153.html
This paper describes, in general, some of the response characteristics of chambers used in a K-Coupler configuration.
Horns are used primarily for impedance matching - which provides increased efficiency - and all horns have a response curve that resembles a band pass filter . When horns are employed for woofers, they are usually built with dimensions to allow their cutoff frequency to be close to the motor's cutoff frequency. Seen at this perspective - any band pass filter "viewed" from it's pass band and at higher frequencies - the horn has a response curve that is very similar to a low pass filter. The horn will match impedance for several octaves before rolloff, so the response curve is similar to a low pass filter with a relatively high stop band.
But a horn does this by increasing efficiency over its usable band.
The K-Coupler is essentially an acoustic low pass filter , with a relatively high stop band because of its dimensions. It is easy to see how a K-Coupler, which passes several octaves, will have a response curve similar to that of a horn. But the K-Coupler does this by reducing efficiency of signals presented above its filter frequency.
Use either McBean's program or Leach's analog, and you will see the response curve of any horn has a shape that "looks" like a low pass filter, if you do not condider horn cutoff - which is what makes it really a band pass system. You will also find that Leach's analog computer cannot model the K-Coupler because it is not a Salmon shape, and frankly - neither can McBean's program. But you will have no trouble at all modeling the K-Coupler with a low pass model.
K-Couplers - and all other acoustic low pass filters - have merits in a loudspeaker design. But it is not appropriate to consider them to be a horn.
Wayne
Sorry WayneI was speaking about K tube not K karlson box.
If i look at the Data frequency respons of the K tube , I observe that K tube mounted on a TAD 1" compression driver seems to be rather a high pass than a low pass.
In the beck page, the efficiency seem to be around 100dB (perhaps more, but difficult to read), so the K tube seems to act as an air transformer rather than a "wave guide".
I am not a specialist in wave theory, only high school rememberings. But, the fact that a tube with a progressive opening as important as the opening of a K-tube acts as an air transformer is not surprising for me. Of course, that's not a Salmon Horn, that's other things than a horn but it seems to act as an air transformer in a different way than a horn does.The design of a K-tube seems rather simple to DYI people, as I pointed simpler that many expensive DYI tractrix horn designs I saw on the WEB. And the design seems to be forgotten. That's the reason of my post.
A contrario, I am not a fan of Karlson box, I am sceptic about this rather complex design for Woofer. But I didn't remember to listen to such enclosures when I was younger. This design seems not to have any success on French audiophiles (20-30 years ago). But There are some old designs of french quater-wave box with progressive triangular or elliptical openings made for Supravox FR.
Eric
Hi Eric!You wrote:
> > I was speaking about K tube not K karlson box.
A slitted waveguide rather than a slitted box. Yes?
> > If i look at the Data frequency respons of the K tube,
> > I observe that K tube mounted on a TAD 1" compression
> > driver seems to be rather a high pass than a low pass.Yes - If I "get the picture" right, then you're exactly right. If it is, in fact, a waveguide that has been slit with the opening widening in the shape of an expotential curve - like the "K Box" - then you are exactly right. It will act as a high pass filter. If you will look at the document link at the previous post, you will see that acoustic high pass filters are described as well.
> > In the beck page, the efficiency seem to be around
> > 100dB (perhaps more, but difficult to read), so the
> > K tube seems to act as an air transformer rather
> > than a "wave guide".Perhaps. But this figure and the one measured by Freddy tends to confirm that there is no impedance matching, and that - like the other Karlson device - it is acting as a filter. The reason I say this appears to be confirmed is that compression motors can usually provide this kind of efficiency unassisted. With a horn, I would expect figures that were 6dB to 10dB higher.
Really, one should not expect an impedance gain from something that doesn't give a change in area betwen the motor and free space. The whole reason there's an impedance mismatch is that there is such a difference of energy required to excite the air immediately in front of the motor and what is required to excite the motor. The motor would ideally be much larger than it is, and of much less mass. So we match the impedance by effectively doing just that.
> > The design of a K-tube seems rather simple to DYI
> > people, as I pointed simpler that many expensive
> > DYI tractrix horn designs I saw on the WEB.Absolutely. And I think both of these Karlson designs that you and Freddy have mentioned deserve some attention. I think designs based on them have some merit. I also think that performance predictions will be accurate when one considers them to be filters rather than as horns.
> > And the design seems to be forgotten. That's the
> > reason of my post.And I'm glad you posted your remarks. I'm certain that Freddy is too! He's been very enthusiastic about the Karlson designs. They're attractive and interesting. They also provide functionality similar to reactive components, without using electrical components - something that is of interest to many on this forum.
Wayne
Hello WayneYou are true to point that i am considering in my post only the frequency respons and not the impedance one.
That is true also that some compression driver as TAD could be used with good efficiency without horn. In fact TAD driver as old altec and JBL drivers has a horn beginning of about 2" length (5cm).
When i pointed in my precedent post, i am not a specialist of wave theory. But, i just had a look of the article in the Karlson link of M. Martin K. Poppe on IEEE journal. The scan is not very good but it has the merit to exist. My sumary is that the theoritical analysis and measurements of M. Popper seems to prove that a K. Tube could be compared to an exponential horn and provides "equivalent" acoustical impedance transforms. This theoritical analysis is confronted to a K-coupler experimental model. The data measured seems to be in accordance with the theory. The K-tube doesn't act as a high pass filter as an opened tube does, this result is discussed in M. Poppe's article.
Of course, wave theory is not my speciality. So i could not analyse if in this paper there is a great mistake, hmm,hmm! This would be very surprizing that IEEE lecturer comity lets great mistakes on a paper, but it could be. In sciences, what is true at the past, could be false at present.
I haven't time and material to perform sound measures. But I think it would be of interest to have comparison between K-coupler + a speaker (or Karlson tube), and the same speaker without.
If I could have your analysis of the paper of M. Poppe, I think it would really be of interest. I read some of your old posts, you have much more knowledge in horn theory than I do.
Eric
Hi Eric!You wrote:
Hello Wayne
> > My sumary is that the theoritical analysis and measurements
> > of M. Popper seems to prove that a K. Tube could be compared
> > to an exponential horn and provides "equivalent" acoustical
> > impedance transforms.Impedance is not the only measure of a horn and impedance changes do not provide a direct indication of efficiency. Filters also provide impedance changes.
One of the common analysis mistakes made in the early part of the century, was to assume that a horn's gain was directly related to the impedance load presented to the amplifier. From this, one would have expected 15dB gains from basshorns of even moderate size. But it can be shown how (and why) this is not the case.
Filters - both electronic and acoustic - also present an impedance change to the system. In the case of the K-Couplers, higher impedance is presented at the frequencies where energy is reflected back towards the diaphram. This raises impedance without a corresponding increase in efficiency. It is not unlike the effect caused by an electronic filter - where the action of the device raises impedance of the system at the filtered frequencies.
I have seen a few types of overgeneralizations that are relatively prevalent. The most common offenders are this one - an overestimation of horn gain based on impedance measurements - and miscalculation of reactive components, both for dB attenuation and phase angle.
Many expect twice the attenuation at the place where reactive impedance equals load impedance, and that's because they forget that reactive impedance and resistance are not combined as two resistances would be. They must be calculated with vector or polar (2D) math. This often also causes phase miscalculations, and just like the "time offset" baffle spacing issue - we come up with the wrong values as a result of this over generalization.
Phase differences between 0 and 90 are additive, at least partially, and those between 90 and 180 cause at least partial cancellation. The "magic number" here is 45 degrees - hence 135, in the case of cancellation - because that's the 3dB point. At 180, we're at the limits of couplingf - where if we were perfectly coupled, we'd have complete cancellation and two diaphrams in a waveguide will come very close to this at the waveguide frequency. Baffle mounted drivers, less so, but still significant cancellation at 180 degrees.
Phase angles are constantly changing in a reactive circuit, as a function of frequency. Likewise, two point sources at different locations in space have constantly changing phase angles as a function of frequency. That means one cannot expect to set a proper distance or a proper reactance to "correct" phase. There are only acceptable amounts of tolerance.
Anyway, these are the most common offenders. Filters and horns. Impedance and phase angles.
Wayne
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2025, Audio Asylum, all rights reserved ]
