-
Feeling lucky? Check out the AK 24th Anniversary Giveaway thread and join in the fun!
K-402 horn - worth the effort to DIY?
- Thread starter BlueVUMeters
- Start date
DirtyErnie
Active Member
That cab's a LaScala bin with a stretched & Widened mouth, right? Should be easy to figure out from existing plans (if you have wood-working tools & skills).
Gigantic
A big, big love
off hand, it looks like a 15º angle on the outside, extended by about 6". should be good for horn loading a few Hz lower than a LS, with better dispersion.
Gigantic
A big, big love
that seems optimistic, I'd say 100-105 at the most.
DirtyErnie
Active Member
Stock LaScala bass horn is good to 104 or something. I've probably made a Hornresp sim of that Allen horn somewhere.
Dave MacKay
Active Member
I’d always heard that too (likely from @ClaudeJ1) but never saw any documentation about it. It seemed likelly to me, but I was taking it on faith.
If you have a simulation that shows the LS acts as a horn to that, would you please post it?
Dave MacKay
Active Member
I’d like to be able to measure that myself. I've attached a couple of plots showing how the tweeter, squawker, and woofers subsystems (i.e., drivers with horns) performed when swept directly (i.e., without crossovers). In the case of the woofer, I used REW to run a sweep between 20 Hz and 1 kHz.
This "nearfield" graph was taken with the UMIK-1 centred on each component, right at the "face" of the speaker (i.e., in the same plane as the front of the speaker).
And here's the same measurement taken at 1m on-axis:
Although the graphs are hard to read because they are spikey, the woofer plots don't show a steep fall-off in SPL just above 100 Hz. The "nearfield" measurement shows the woofer SPL dropping around 70 Hz; the 1m measurement shows that as well, but is complicated by a peak (perhaps a room mode?) around 50 Hz.
DirtyErnie
Active Member
Simulations:
First, a LaScala with the K33. K33 by the specs isn't a great 'horn driver' Qts is too high, so on and so on. Found out playing with hornresp, that's what lets it do a good job acting as a direct radiator below where the horn unloads:
Also found out that the Delta 10 is a GREAT horn driver, but won't drive anything but the horn. Notice it totally giving up below ~100. (and at that frequency, whats 5-10Hz...)
Lots of sim time, with lots of different drivers, and I finally came up with this line in a package that actually fits the living room (Hopefully to be built this spring):
First, a LaScala with the K33. K33 by the specs isn't a great 'horn driver' Qts is too high, so on and so on. Found out playing with hornresp, that's what lets it do a good job acting as a direct radiator below where the horn unloads:
Also found out that the Delta 10 is a GREAT horn driver, but won't drive anything but the horn. Notice it totally giving up below ~100. (and at that frequency, whats 5-10Hz...)
Lots of sim time, with lots of different drivers, and I finally came up with this line in a package that actually fits the living room (Hopefully to be built this spring):
Dave MacKay
Active Member
Thanks!
There’s a lot about this I don’t understand.
- Is the red line (105.89 Hz on the K-33 graph) where the SPL drops by 3dB?
- To my eye, the SPL — while lower — still looks pretty reasonable until around 60 Hz.
- What are the criteria for when the La Scala is said to stop acting like a horn and act as a direct radiator instead?
DirtyErnie
Active Member
The horn physically stops loading the driver, more like in the Delta10 graph (this is actually a pair of Delta10's in a LaScala cab). The K-33 can work really well as a direct radiator when the horn quits. Also does pretty well driving bass reflex below that. (all these graphs are sims in corner-loading) You can see this better in the Acoustical Impedance charts, which maybe I can grab later.
Another fun thing Col. Klipsch did was to ignore a bunch of 'proper horn theory' with his designs. "Ideally", they should all have much smaller volumes behind the driver, but if he did that, all his designs would cut right off at the bottom of the horn's loading frequency range. By increasing that rear volume, the roll-off is much more gradual, allowing the LaScala to have ANY bass below 100Hz, for instance. Also allowing the KlipschHorn to get below about 45Hz.
Finding an off-the-shelf driver less than 15" that can drive a horn, direct radiator, and bass reflex is a real chore.
Another fun thing Col. Klipsch did was to ignore a bunch of 'proper horn theory' with his designs. "Ideally", they should all have much smaller volumes behind the driver, but if he did that, all his designs would cut right off at the bottom of the horn's loading frequency range. By increasing that rear volume, the roll-off is much more gradual, allowing the LaScala to have ANY bass below 100Hz, for instance. Also allowing the KlipschHorn to get below about 45Hz.
Finding an off-the-shelf driver less than 15" that can drive a horn, direct radiator, and bass reflex is a real chore.
Gigantic
A big, big love
I was wrong, my memory has been adversely affected over the past few years by my recent illness. My measurements show a resonance at around 100Hz, which is the transfer point from horn loading to bass reflex and another at 62 Hz, where it runs out of juice, unassisted by venting. the resonance and null at 180Hz and 345Hz are much more pronounced and really highlight the deficiencies of this horn without PEQ
my "compact" horn, which is effectively a 80% scale Lascala, transfers around 130Hz and falls off without assistance from the passive radiator at 55Hz, the passive extends to 32Hz, using a Dayton Audio ODEUM 12N Midbass woofer and the same passive used in the Forte. My first choice was a JBL 2206, but the one I have needs to be reconed; I expect that it would perform slightly better, similarly, the Dayton Audio RSS390-PR was unavailable when I built my prototype, so I went with the inferior Klipsch part. Green is the measurement at 1m; red is the passive at .25m at, iirc 1/12 smoothing:
my "compact" horn, which is effectively a 80% scale Lascala, transfers around 130Hz and falls off without assistance from the passive radiator at 55Hz, the passive extends to 32Hz, using a Dayton Audio ODEUM 12N Midbass woofer and the same passive used in the Forte. My first choice was a JBL 2206, but the one I have needs to be reconed; I expect that it would perform slightly better, similarly, the Dayton Audio RSS390-PR was unavailable when I built my prototype, so I went with the inferior Klipsch part. Green is the measurement at 1m; red is the passive at .25m at, iirc 1/12 smoothing:
Dave MacKay
Active Member
Thank you.
Any thoughts on the Kappa 15C as an alternative to the K-33?
A number of posts have claimed the Kappa 15C 4 ohm woofer as an improvement on the K-33 but I don’t recall seeing any measurements.
Invidiosulus
Well-Known Member
Care to share the input you are using for the LS?
your first K33 sim doesn't show the massive 148Hz hump that they are known for.
Below,
K33 in gray.
Kappa 15C in black
K33 in ported LS(DJK mod) actual measurement in room in blue before applying PEQ.
One can see that the Kappa 15 C should offer stronger performance from the 150Hz region on up compared to the K33 but it doesn't take advantage of the box volume and should provide less low end in a stock LS.
I assume my LS II measures closely to these, at least the bass bin. After switching in the Kappa 15C for the K33’s, the K55 was bumped up 2.6 dB (calculated).
This was done to more closely match woofer and the MWHL and DE120 at crossover frequency.
Retuned the low pass on sub up to 80Hz and LS is running full range.
Best they have sounded. So far. Plus are really a 105dB efficient speaker now. At 1000 Hz.
I do not notice the hump at 150 Hz. The crossover may be taming that some.
Listen at a good distance and all sound is reflected. Cannot see the speakers. This may tame the high Q spike at 150Hz.
The ported LS does not look very good, compared to stock. Even less compared to Kappa 15C.
This was done to more closely match woofer and the MWHL and DE120 at crossover frequency.
Retuned the low pass on sub up to 80Hz and LS is running full range.
Best they have sounded. So far. Plus are really a 105dB efficient speaker now. At 1000 Hz.
I do not notice the hump at 150 Hz. The crossover may be taming that some.
Listen at a good distance and all sound is reflected. Cannot see the speakers. This may tame the high Q spike at 150Hz.
The ported LS does not look very good, compared to stock. Even less compared to Kappa 15C.
Dave MacKay
Active Member
That is of interest to me. I am running a system that is fairly similar to yours.
I have tri-amped LS I with K-33 woofers, K-55/K-400 midrange, and DE12-/SMAHL tweeters. I replaced the passive crossovers with an active crossover/DSP. My subs are crossed at 80 Hz too.
The system sounds good to me now, but I’ve been wondering if the. Kappa 15-C might be an easy change that would make it better.
Would you happen to have any before (i.e., K-33) and after (i.e., Kappa 15C) measurements?
DirtyErnie
Active Member
Mine were 'Corner' sims, here's 'Corner' vs 'Wall' K33:
K33 (gray) Kappa 15C (black): The trend is there, get better upper range, lose some bass.
Where does the horn quit? The lowest crossing of Acoustical Impedance and Acoustical Reactance:
K33 (gray) Kappa 15C (black): The trend is there, get better upper range, lose some bass.
Where does the horn quit? The lowest crossing of Acoustical Impedance and Acoustical Reactance: