Jump to content

Analysis of Klipschorn and La Scala cabinets with several drivers


hulkss

Recommended Posts

You and I used very different models, and even very different woofer parameters, yet Hornresp comes up with very similar results.

I opted to use a similar throat model in the LaScala as you used in the Khorn. That is, calculating throat area (S1) at the "splitter" after the horn doubles into two paths. Then entering the slot area as (Atc) area of the throat chamber. I also lengthened the last horn segment and shortened the middle segment slightly maintaining the overall length. This was to prevent the "Cir > 1" warning. The help file in "Hornresp" states that results are invalid if this message appears.

I have modeled a lot of horns in "Hornresp". It takes fairly big percent changes in a model to make a significant difference in the SPL output. After all, a 6b change requires the acoustic pressure to double.

Link to comment
Share on other sites

It takes fairly big percent changes in a model
to make a significant difference in the SPL output.

I
think the plot of more interest is the acoustic impedance
response....you'll be able to see smaller differences more readily. You
gotta double the impedance to introduce a 3dB change in the frequency
response. The more jagged the acoustical impedance, the more ringing
you're gonna get... (assuming that the wave travelling down the horn satisfies the assumptions hornresp makes).

Link to comment
Share on other sites

I opted to use a similar throat model in the LaScala as you used in the Khorn.

It's a tough call -- how to handle folds when the cross-sectional area changes significantly within the fold. After I developed the LaScala and Belle models, I went back to my KHorn model and changed the throat model. The overall effect was negligible.

This was to prevent the "Cir > 1" warning. The help file in "Hornresp" states that results are invalid if this message appears.

Yes, I understand. The warning says that the throat acoustical impedance becomes invalid. I experimented with that a bit and found that changing the area enough to avoid the message had no discernible effect upon the results, so the models must be very "close to the edge".

I have modeled a lot of horns in "Hornresp". It takes fairly big percent changes in a model to make a significant difference in the SPL output. After all, a 6b change requires the acoustic pressure to double.

That matches my experience, as well. Overall, horns are pretty tolerant of moderate changes in cross section, as long as overall length, throat area, and mouth area are maintained.

Link to comment
Share on other sites

I'm still refining the LaScala and Belle models. It's not like it matters, so I think maybe I'm a glutton for punishment. I revised the throat model as suggested by hulkss, and made a few minor changes in the cross-sectional area profile. These models still get the "Cir > 1" warning in Hornresp, but if you ignore that and compute the response, and then modify the areas just enough to avoid the warning and again compute the response, the differences between the two are very small.


LaScala here; Belle in a followup post.

post-22723-13819455497062_thumb.png

  • Like 1
Link to comment
Share on other sites

  • 3 months later...

What happens if you add corner reflectors to the back of the La Scala bass horn? How does the response curve and impedance curve change?

Not much. A reflector has to approach 1/2 wavelength in size to make a significant difference.

Link to comment
Share on other sites

This graph may be misleading. Because you set the lines equal at 100 Hz, it makes the K-33 look like the worst driver and the Altec look like the best. What do the curves look like if they are not indexed to be equal at 100 Hz. How did your Ciare 15.64 curve in the real K-horn compared to the predicted response?

A Peavey FH-1 is similar to a La Scala with a smaller back air chamber and a bit longer horn. The OEM woofer distorts excessively. Any way to estimate how the Ciara woofer would behave in them? I have 4 that could use better woffers, but we need to keep the high power handling of the OEM woofer.

I evaluated dozens of current production 15" drivers and could only find a few that would outperform the stock K33 by any significant amount. Here they are all on one graph. The Bob Crites driver is the only 4 ohm impedance "direct" replacement in the group.

I tried to match the responses at 100 Hz so they would be relatively easy to compare. I'm not sure if all these drivers will fit. I'll model the same drivers in the La Scala next.

Khorn_Resp.jpg

Link to comment
Share on other sites

Here's a list of some other woofers worth checking-out in this application. I have chosen them because they all show nicely extended high frequency response in Hornresp. But I have not tried them specifically with the KHorn, LaScala, or Belle models, so I do not know how well they will perform with the Klipsch throat dimensions or rear chamber volumes.


B&C 15NDL76

B&C 15PE40

B&C 15PL100

B&C 15PS76

B&C 15PLB76

B&C 15TX

B&C 400AMX

BD Designs BD15

Ciare 15.75SW

Ciare 15NDH-4


Greg

Link to comment
Share on other sites

the real question is not what causes the 200hz hump, but rather, what causes thew 300hz dip.

wonder if the belle has the 300hz dip...we may have an implied answer.

Link to comment
Share on other sites

Pls forgive the noob post. I wish I knew what you folks were going on about. Also, with the graphs and numbers, what is "ideal" or even "better"? What are the practical applications of such differences - do they indicate which is simply the best driver or where xovers have to be changed, etc?

Thx

Link to comment
Share on other sites

Pls forgive the noob post. I wish I knew what you folks were going on about. Also, with the graphs and numbers, what is "ideal" or even "better"? What are the practical applications of such differences - do they indicate which is simply the best driver or where xovers have to be changed, etc?

Thx

Hornresp is a horn analysis and modeling program. You enter the dimensions of your horn, and the measured parameters of your woofer, and Hornresp predicts the frequency and phase response, acoustical and electrical impedance, diaphragm displacement, group delay, etc. It's only a model, so the results are only as good as the assumptions behind the model and the accuracy of the input data. But for comparing different woofer drivers in the same horn, or different horns with the same woofer, etc., it's a nice tool.

Ideally, you'd like for the frequency response to be ruler flat from DC to light. In the real world, of course, that never happens. In the frequency response plots that are being addressed in this thread, people are wondering what causes some of the peaks and dips that seem to occur in the same places in the KHorn, LaScala, and Belle Klipsch. If they are caused by the woofer, then changing the woofer to a different make and/or model might eliminate them. If they are caused by the dimensions of the horn, then changing the woofer won't eliminate them.

Ultimately, as I said before, Hornresp is only a model. While the model can predict general characteristics, the effects must be measured in the real world in order to have any credibility.

Greg

Link to comment
Share on other sites

Here's some interesting food for thought. In post #1181505 I graphed a carefully measured internal path for the KHorn, and a Hornresp model for it. Well, I've been experimenting a bit with that Hornresp model and some alternate woofers. And, if you're subwoofing your KHorns, so you can afford to give up a few Hz of KHorn bass, Hornresp says that you can improve the midrange response of your KHorn by a considerable amount.


In the attached graphic I have duplicated that earlier internal path graph, only this time I've added a green line that shows the modifications suitable for use with a BD Designs BD-15 woofer. (This is a VERY expensive woofer, but there are less expensive alternatives that perform nearly as well.) And the Hornresp graph below that shows the improvement that results from the internal modifications and woofer substitution. (Black line is BD-15 in modified horn; gray line is K33E in stock horn.)


As always, models don't substitute for measurements, but this certainly points in an interesting direction.


Greg

post-22723-13819486168538_thumb.png

Link to comment
Share on other sites

Could you show the K-33E in the new horn? And just for kicks, could you show the BD-15 in the original horn?

I do know that the area expansion of the khorn was compromised to make it buildable....same thing goes for the Jub LF (according to the JAES article). Also, these models are only showing the effects of impedance matching the driver and how it relates to the flare rate. Hornresp sorta loosely models polar response, but it's only valid for an axisymmetric horn. I've found that it's often very difficult to optimize both polars and impedance transfer, so you gotta pick your compromises.

All that to say, it'd be interesting to see if you could actually realize a horn the maintains your model without sacrificing other attributes...and being actually buildable.

Link to comment
Share on other sites

Could you show the K-33E in the new horn? And just for kicks, could you show the BD-15 in the original horn?

K33E in black, BD-15 in gray. The "ideal" K33E horn also uses the optimum throat size predicted by Keele (which is NOT the same as the optimum throat size for the BD-15) and slightly smaller than stock back chamber volume. For this plot I just used an exponential expansion instead of the piecewise conical approximation; the results are within a fraction of a dB of each other. The "original" BD-15 horn is unmodified from stock Klipsch spec.

I do know that the area expansion of the khorn was compromised to make it buildable....same thing goes for the Jub LF (according to the JAES article).

Agreed. In the "ideal" model I tried to limit the modifications to reduction of cross-sectional area, because it's a lot easier to make the area smaller than it is to make it larger! The exception is in the region between approximately 50cm and 90cm, where the area can be enlarged by making the internal "wedges" smaller.

Also, these models are only showing the effects of impedance matching the driver and how it relates to the flare rate. Hornresp sorta loosely models polar response, but it's only valid for an axisymmetric horn. I've found that it's often very difficult to optimize both polars and impedance transfer, so you gotta pick your compromises.

All that to say, it'd be interesting to see if you could actually realize a horn the maintains your model without sacrificing other attributes...and being actually buildable.

All completely true. Models don't substitute for measurements, or for construction! Frankly, I think that the addition of corner reflectors would probably have more significant effect than the relatively minor changes in cross-sectional area. The "foldback" region, where the path hits the corner and turns nearly 180° to exit the horn, seems to be particularly troublesome.

Greg

post-22723-13819486188612_thumb.png

Link to comment
Share on other sites

Just one thought to throw out there....doesn't decreasing the cross-sectional area increase velocity related distortions?...which predominantly is problematic with the lower frequencies. Klipsch and company seem pretty adament about using rubber throats in all their horns. I can't tell if it's for decreasing throat overload, and/or controlling HF polars, or maybe something else?

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...