Early folded horn questions

[mungkiman]

Figure 1 (see attached), is a patent drawing for an early Klipsch corner basshorn, utilizing a 12" driver, and designed for 40 to 400 Hz performance with a "minimum loss in fidelity of the low frequency tones" to 30 Hz. The front panel size is near identical to the Klipschorn, while the top and bifurcated pathways are somewhat different.

Has anybody seen, or more importantly HEARD one of these? Any information would be appreciated. Thanks,

Chris

[HDBRbuilder]

That IS the Klipschorn...the bifurcated pathway is THE SAME...as is everyhing else in the design. Certain mods and changes were made as time went by until the speaker began to be produced for consumption, but that is the K-horn BASS-BIN patent pic.

[mungkiman]

Thanks HDBRbuilder,

You must also be familiar with Figure 8, of the same patent, showing an alternate method of construction using twin drivers. This afforded greater power handling, for theater applications, while keeping the exterior dimensions/appearance the same as in Figure 1. The construction of Figure 8 would be simpler than that of Figure 1, but I question the end result.

Can anyone speak as to the practical affect of this mod? Sonic characteristics? It seems of much simpler construction, with twice the power handling, yet the same volume/dimensions. Note: these drawings reference a 12" driver.

Thanks again

[D-MAN]

Howdy...

The point of the patent's 2-driver version was to provide an increased power-handling capability (and output) than the single driver approach. Remember that the patent was from 1946 when tube-driven wattage was expensive (it still is!) and the ability of drivers to dissipate heat was limited compared to todays standards. I think that special mention was made in the text to provide a more-powerful version intended for theater and large format use. In the patent, PWK also mentioned using a single larger driver but the patent drawings seem to all feature the 12" variety. The drawings faithfully represent what we know today as the "Klipshorn" which survived 50 years without serious design changes. Many other designs have been attempted and marketed; but they have all gone extinct...

In the case of the 2-driver version, the shortening of the exponential horn would theoretically raise the fc of the horn, i.e., it would not go as low as the single driver version which utilizes a longer horn but having double the throat/cone area and the natural coupling of 2 woofers would more than make up for that in raw output horsepower.

Since the bass output would be greater the midrange and tweeter setup would also require higher output to keep up with it, again not intended for smaller spaces.

Within the bandwidth that it would reproduce, it would be louder than a single driver, of course, and have twice the power-handling capability of a single.

DM

[WMcD]

I've never seen anything to indicate that the two-driver version was ever built. Under patent law, it is not necessary, as long as it is clear that it can be.

I can't agree with the thought that shortening the overall length of the horn decreases the Fc. It does if you have the same throat area. But here there are two drivers, with therefore double the throat area.

Gil

[mungkiman]

I was hoping it could go lower.....

[D-MAN]

Don't think that I am against this project, as I am not.

I am just postulating some possible performance issues based on some experience with building and researching horns on my own...

Some more thoughts about the 2 driver version...

There are a couple of drawbacks to this approach; it appears to be a modification of a previous and proven design, and is not a true "purpose-built" design from the ground up. It has cut some corners, in other words (pun intended!).

I'm guessing but the possible results of this could be (and this is probably a worst case):

(1) a "less" smooth response in that the remaining exponential horn is not as long as the expansion rate would suggest. Essentially, the expansion rate is for a longer horn and this design just "chops" the front part off. This usually produces "peaky" or uneven bass response.

(2) higher low-frequency cutoff (fc) due to a throat/length/mouth mismatch due to expansion rate and length of horn not being "correct" for a 40Hz fc horn (which I believe that the horn mouth and expansion rate would suggest).

(3) higher modulation distortion due to restricted size of back chamber ; however, insulative material can effectively increase the "apparent" size of the BC by a factor of 1.4 times as large as it really is, so that might be used to counter this effect.

In the 2 driver version, I wouldn't think that the back chamber volume is capable of supporting the use of 2-15" drivers as the effective throat area would also be doubled (I believe that the formula concerning this is in the patent, too); rather it may be better with 2 - 12" drivers otherwise the fc of the horn would be raised even higher (when too small of a back chamber area is used) considering the throat opening sizes on the 12" drivers to be substantially smaller.

Additionally, more mid-bass and mid-range waveforms will be transmitted through the horn in that is is shorter (than the traditional 1-driver version) and has less FOLDS to contend with. Also when using 12" drivers, this could/would be even more exagerated as they would probably have a higher natural upper-frequency cutoff (or ability) than a 15" would. This would, in turn, effect the design of the crossover points and efficiencies of the upper frequency lashup employed.

Thirdly, where would you get a pair (or 4 total) of 12" drivers that can perform adequately in the horn? The 15" has historically won out on this point... in that the Klipsch K33E is $130.00 each. I wouldn't want to settle for less than nominal Khorn performance if building such beasts- they are already the same size, etc...

But, I am all for DIY projects! Please proceed and let us know about it! It does look very interesting!

DM

[WMcD]

I guess I got that backwards. But the answer is still the same, the Fc remains the same. This is because the rest of the horn structure going toward the mouth (big end) is the same.

It is important to keep the same throat (small end) condition. It is not quite true that the K-Horn doubles in area in each of the three sections . . . last time I looked at it. But it should have a uniform doubling lenght, all exponential functions do this.

Essentially you can go to a lenght where the area has doubled, put in two of the same drivers, and there is a proper throat match.

Looking at Fc is only part of the issue in finite length horns. The mouth size is equally important. It is actually why horns work. Just as in direct drivers, the bigger the mouth, the better for bass. There is an upper limit to things getting better. However, the K-Horn mouth is smaller than optimum, even given the effect of the corner.

Gil

[D-MAN]

Well, dudes, take this all with a grain of salt.

GIL:

You are usually right about things (and might be, after all), but I feel the need to explain and clarify my point and in this case, (GASP!) refute yours.

I have read numerous times that the LF waveform is propagated most efficiently by a slow (for a low fc) and mathematically steady expansion rate from a given throat size (which is integral in determining the expansion rate and efficiency for a particular fc) and the subsequent length of the horn to the mouth and one cannot shorten the horn by simply moving the throat opening around and get the same result; if it worked that way, the Khorn itself would be quite a bit smaller in size as would all horns. Of course, the size of the throat would have to be increased at the point that we are considering for the 2-driver version, but the cone size in question would be the controlling limit on that. I think that the horn area cross-section at that point is larger than the available cone area. I am assuming an unaltered Khorn dimensioned horn, as dimensions are not specified in the diagram. None-the-less, I will stay with what I said in the previous post. But as always, you bring up a good point, so I will have to chew on it awhile and do some research.

MUNGKIMAN:

Agin, just a guess, but if you go with dual 15" drivers, you should theoretically be able to just power your way past the low freq cutoff point to a certain extent at least more than a normal Khorn does. PWK published an article about horn low fc as not being an absolute limit.

But I don't think that you can find a good pair of 12"s that could for as little investment.

My final bit of reasoning: If it out performed or even came close to the performance of the Khorn,

and it is certainly easier to build, then why did PWK continue to manufacture the more difficult design.

I rest my case.

DM

[mungkiman]

Your Subwoofer is smaller than your Klipschorn basshorn. It's all about compromise. I wonder what the price point of drivers VS. labor/plywood was.

Note: the back chamber volume is also much greater in this early version.

[mungkiman]

If you scaled up the original twin driver dimensions for 15" woofers, the basshorn front panel would be close to 24" x 48". If the top of the basshorn were as depicted in Figure 1, you could place two Altec 511B's on top, and have room for 2 T-35's mounted horizontally or vertically. Not too big, maybe big enough. Any guidelines for what kind of performance might be expected? With all twin drivers, would any appreciable differences need to be made in the networks? Thanks,

Chris

[djk]

"and the subsequent length of the horn to the mouth and one cannot shorten the horn by simply moving the throat opening around and get the same result; if it worked that way, the Khorn itself would be quite a bit smaller in size as would all horns."

Both the Klipschorn and the Jubilee have 100hz initial taper rates.

[WMcD]

I believe we are close to being in agreement.

The issue of the twin driver in the patent version is muddled because of the use of the rapid flare (Fc = 100) in the first section and a slower one in the next two (Fc = 50 or 48 Hz). However, when PWK went to the 15 inch driver, the initial flare came down to about 50 Hz, making it a classic exponential overall.

If we look at the area where the double drivers are placed, I expect it would be appropriate for a 15 inch driver. Whether a 15 inch driver would fit is another issue.

Harry Olsen has some diagrams of throat impedance for a 100 Hz horn. In one set he varies the mouth size while keeping the throat constant. In the other he varies the throat size while keeping mouth constant.

When you do this, you vary the overall size of the horn while keeping Fc constant. Again, in the exponential horn the lenght in which the area doubles is a constant length which is determined by Fc. In a mathematical sense, the exponential equation "says" the area doubles in "lahmed" (The Hebrew letter L, which I've proabably mis spelled. My Catholic education did not include Hebrew, smile.)

Again, let's look at a pure exponential. The LaScala is an easy example. You can consider it as having three lahmeds. (You just have to keep track of the combined paths.) The throat size is 0.5 square feet. After a distance of 12 inches (out to the back corners), the area has doubled to 1 square foot. In another distance of 1 foot (2 feet total) the area is 2 square feet. (This is going forward to the mouth.) In another distance of 1 foot (three feet total) the area is 4 square feet. This last is the mouth (big end) size. It is 2 feet by 2 feet.

Now you'll see the LS varies a bit from the above approximations, but not very much at all.

Conceptually, you can see that you could place two drivers after the first expansion. They'd each be looking into 0.5 square feet of area. I'd have to look at whether this causes an issue of back chamber volume. But, we have reduced the overall lenght of the horn while keeping the same Fc and mouth size.

There is no getting away from the fact that the mouth size is the big challenge. The whole point of the horn is to have a big mouth and with that, a big radiating area.

If you look at the equations for optimum mouth size (in free space), they predict that the optimum mouth size is about 8 times larger than it actually is. This issue is solved by placing the LS in a corner.

Conceptually, you have to make a comparison of free space (no walls nearby) to a corner. Imagine the LS is sitting in the corner of an interior room in an apartment house. Look at the corner point on the floor. It is shared by 8 other rooms. 4 on that floor. 4 on the ceiling of the 4 rooms below.

Another way of looking at the 8 room theory is that the actual single corner is another section of horn grafted onto the mouth. It does not follow an exponental expansion. However, it does extend into the room for a great distance.

The corner loading presents an interesting thought experiment. Question: How big is the LS or K-Horn while sitting in a corner? If you see that the room itself is another horn section, the bass horns are very big.

I'll post Olsen's (Olson?) throat impedance diagrams.

The final issue may be why a driver has a specific optimized throat size. One reason is that the energy transfer is optimized. The other has to do with reflections from the mouth. When we look at the throat impedance, we'll wonder why it is so ragged. The answer is there are echos from the mouth when the acoustic match there is not perfect. The proper driver/throat absorbs them and prevents a further echo back up the horn. That is a deep subject though.

Maybe I've gotten obnoxiously pedantic here. Sorry. But the LS and corner are good reverse engineering explanations of a difficult subject.

Gil

[mungkiman]

With an easy modification to the top section, the Klipschorn horn mouth could be as tall as the overall cabinet. The mid horn and tweeter would nestle between the rising halves of the basshorn. Different look for sure. Better low end?

Thanks to everyone for their interest and participation.

[D-MAN]

Hi, again.

Gil was right: As far as placing the 2-driver throat openings further "up" the horn as compared to the single driver version, as long as the cavity opening effectively doubles the orig. single opening at the appropriate expansion point in the horn, the geometry and expected performance of the horn is retained. So this does not alter the horn's fc. I was wrong on that point.

2 drivers doubles the efficiency which is reported in the single driver version to be approaching 30% if I remember correctly. So effectively, you have to double the midrange and tweeter horns and drivers to keep up with the bass output. Maybe that's the reason PWK didn't manufacture these!

The back chamber volume is up in the air for me... since we are not enlarging the throat area or the horn geometry then I assume that no changes are required for the back chamber since its effect is to match the resistance of the air in the exponential horn itself to provide equal motion of the cone front and back. So perhaps no changes are needed here, too.

The doubling of efficiency would have an apparent effect on the rolled off bass freqs, but the overall response curve of the horn would not change. It would just be louder at the same given wattage, and by being louder, the "rolled off" low freqs below the fc would also be twice as loud. But they still get rolled off below fc proportionally. Altering the horn geometry to lower the fc is the only way around that.

I would also estimate the 2-driver version to have far more PUNCH (twice as much) compared to the single diver unit and with a double set of upper horns, the transients could kill. Perhaps that is what you want as there is not much lower bass available on most music source material that is not single-note bass (for cars) or "noise" and/or "sound effect" on a movie soundtrack. I'm sure that the 2 driver version would definately kick some serious tail from here to next week.

BUILD THEM and THEY WILL COME...

DM

[D-MAN]

Gil brought up another great thought-provoker concerning corner loaded mouth size:

I figure that the apparent mouth to be 1/8 the calculated size of a straight horn mouth when corner loaded. This is a kluge, but seems to work. Normal straight horn rules-of-thumb get kind of "seriously modified" when it comes to folded corner horns...

However, I have read that corner loading produces EIGEN-TONES which are supposedly not good.

What the heck is THAT? Khorns sound great to me... am I missing something?

DM

[Marvel]

Gil (or anyone else for that matter, Dennis?),

You go through the size of a LaScala, but I don't see where the second part of the horn (from the back of the cab on each side toward the front) gets any larger. Yes, the first left/right expansion is obvious, but the next section is 3.5 inches wide x 22.25 inches tall x 7.5 inches long. (77.875 sq inches) Same size for each side that extends the 7.5 inches forward. They don't expand at all. What does that do for the wavefront?

Marvel

[WMcD]

Be careful about this thought of two drivers increasing efficency. There are some games to be played with the electrical loading. E.g. reducing the electrical input impedance by putting drivers in parallel.

None the less. If it was possible to increase optimized efficency (which is 50% in this application) by doubling, and redoubling drivers . . . then we could get above 100% efficency. But that can't happen.

- - - - -

The eigen values is something which I did not address because it is a very difficult subject. Please note that we discuss the beneficial effects of corner loading. That is accurate and in fact Klipsch, Inc. does have an outside corner to run test in that condition.

However, the fact of the matter is that any speaker type in a home setting is sending pressure into a somewhat closed room with three sets of parallel walls (counting the floor and ceiling). I didn't go into this with the apartment house analogy.

To some extent, those three sets act like organ pipes . . . and we're sitting inside the organ pipe. Again, it takes some fuzzy logic and thought experiment. You have to consider the length of the organ pipes as being the distance between opposite walls to each other, and ceiling - floor. So there are going to be resonances set up determined by those distances.

You could think of this as being in the (roughly) half way point of three organ pipes. One is east-west, one is north-south, one is up-down. In the thought experiment you have to shorten them down and they become a rectangular room. But, the room has different dimensions floor to ceiling, side wall to side wall, and back wall to front wall.

So the issue becomes, how do we describe the relative distances of the the walls? That is also to say, the size of the bass organ pipes we sit within. Those are the eigen values. If the room is a cube, the eigen values are 1:1:1. In most rooms the ratios vary.

- - - - -

Gee. Let's consider the problem of driving pressure into each of the three organ pipes. The end of each is a good choice. Where does each of the pipes end so we can drive them?

You'd say we'd like to put bass drivers (three total) in the middle surface of the front, side and bottom walls. That would really get the pipes going. However we may have to live with putting one driver off to the side of the admittedly squat pipes. The side of the left right, front back, and top bottom "pipes" all occur at the corner of the room. Therefore, the corner is a good place to excite all the three organ pipes. Again, the squat organ pipes are our room.

- - - -

You may say, "Horrors and damnation. I thought I had this magnificent rig with flat frequency response and now Gil is saying that I'm sitting in three organ pipes." But that is correct.

Is it a good thing? Yes, it can be. The ability of any speaker to make bass is limited. We benefit by sitting in the organ pipes. However, this is only if all the eigenvalues just happen to cause resonances at the right place so that the room becomes a type of equalizer to boost bass just were the speaker is rolling off in bass response. This is established by the dimensions of the room. The other problem is that we're seldom sitting at the exact middle of the organ pipes.

- -- -

My old pedantic self.

Gil

[3dzapper]

Horrors and damnation

[djk]

http://forums.klipsch.com/idealbb/files/1977LaScalaDim12.jpg

"Yes, the first left/right expansion is obvious,"

Yes, it goes from 13" to 22.25" in less than 8", a taper rate of more than 100hz.

"but the next section is 3.5 inches wide x 22.25 inches tall x 7.5 inches long. They don't expand at all."

Actually it is 2.75" X 22.25" and the mean path is 9".

"What does that do for the wavefront?"

Sound transmssion in tubes 5.32 (Olson), Fig. 5.15, at high frequencies the loss is signifigent. It can be ignored in the bass.

Transmission from one pipe to another pipe of different coss-sectional area 5.33 (Olson). The change of 2.75" to 3.5" causes signifigent reflected power near the crossover point. This power can add or subtract depending on the frequency (it makes peaks and dips in the response).

One should also read Exponential connectors

5.29 and see Fig. 5.12, clearly the pipe section has no gain.

Final expansion goes from 3.5" to 11.5" in about 13".

Overall, you could call it a 100hz horn.

"the geometry and expected performance of the horn is retained. So this does not alter the horn's fc. "

Actually, you were right the first time.

Fig 5.11 A~D show the effect of a larger throat with a constant taper rate and mouth area. The throat size has no appreciable effect upon amplitude variations (that being strictly mouth size). However, the seperation in frequency between successive maxima is increased as the throat becomes larger due to the decreased length of the horn. The frequency at which the first maxima in the acoustic resistance occurs becomes progressively higher as the length is decreased. Comparing C with D, we see that doubling the throat diameter shifts the effective cut-off of the horn upwards by about 30%.

Even though Fig 5.11D has a 100hz taper rate and enough mouth area for same, its actual cut-off is is closer to 200hz. Based on its 1/4W length the cut-off should be 175hz, so we see that we really can't cheat as we will be caught.

"The back chamber volume is up in the air for me... since we are not enlarging the throat area or the horn geometry then I assume that no changes are required for the back chamber since its effect is to match the resistance of the air in the exponential horn itself to provide equal motion of the cone front and back. So perhaps no changes are needed here, too."

3.2.2 The back chamber air volume can be calculated by the following equation: (Klipsch, JAES Vol. 48 October 2000)

Vb= Vas/((Fc/(Fs)(Qts))-1

So if you used two of the same driver, the back volume would double (Vas doubles, everything else is unchanged).