Crossover points for AB-2 (Belle Klipsch) and AL-3 (LaScala)

[ragjr]

Just curious if anyone can tell me the crossover frequencies of the

AB-2 network for the Belle Klipsch and the AL-3 network for the LaScala?

[Al Klappenberger]

ragjr

Belle AB-2: - 500 and 6000 Hz

LaScala AL-3: 400 and 6000 Hz

Approximately

Al K.

[BEC]

The AB-2 and AL-3 are identical crossovers except that the AL-3 has a bit extra tweeter protection (poly switch) added to its output. I don't know what Klipsch claims for crossover points, but here is a trace I ran on an AL-3 which would be the same as the AB-2 and the AK-3.

Bob

[Al Klappenberger]

Bob,

Thanks for point out that the AB-2 and AL-3 are identical. I check and you'r right. I never realized that. It just point out another wrinkle in Klipsch networks. The LaScala has a 400 Hz horn and the Belle has a 500 Hz horn. Just like the "AA". A gentle-slope filter just lets the horns decide the crossover frequency. This means there is no good way to answer the question of where the low crossvoer actually is!

Al K.

[BEC]

Al,

I would go ahead and include the AK-3 as being essentially identical to the AL-3 and AB-2 also. Only difference there is a slightly different value of woofer cap in the AK-3. I doubt you could tell that difference listening.

Bob

[Al Klappenberger]

Bob,

I did a lot of computer simulation on the AK-2 and AK-3 woofer filters and found that the 100 uF of the AK-3 is the best value. It makes a very nice 3rd order filter if you include the 1 mHy voice coil inductance and 6 Ohms resistive load of the K33 as the 3rd element. Tha't why I am offering the "AP12-AK3" as one of the options in my "AP12" sereis networks. My guess is that Klipsch discovered that by trial and error though. The part quality Klipsch used is fine for a woofer filter, so why not use what's already under that hatch if the AK-3 is already installed. Poping that hatch is the bigest pain in the butt when installing one of my networks. This way the customer doesn't have to do it!

Al K.

[Al Klappenberger]

Bob,

Here's the AL-3 and AB-2 woofer filters. Notice the peak at about 175 Hz. At 350 Hz it's 6 dB down. That's 3 dB farther down than the AK-3 filter. Neither is bad but I think the AK-3 is better for the Khorn. It already has a peak at 200 Hz without the woofer filter exagerating it! I really doubt the Belle or LasScala really needs that peak either.

AL K.

[Chris A]

Bob/Al (and others who may be conversant with this topic),

Many thanks for the information here-you guys are great contributors here. Note that I recently hooked up my center EV Dx38 to triamp my "new" '82 Belle center with stock drivers, and the difference was pretty astounding. I corrected driver delay and relative gain on the drivers.

I note with some distress that the midrange in your plots for these two crossovers from Klipsch are not attenuated/low-passing in the tweeter's band. Al, I see your discussion on this subject on your site.

This also shows up when you listen to the Belle triamped (24 dB/Octave Linkwitz/Riley crossovers at ~500 Hz and at 4.5 KHz). The sound is now very different (better). I'm currently evaluating the relative gains on the drivers, but the first indications are that I've got it okay according to the measurements on-axis ~10 feet back: -6.5 dB on midrange relative to woofer, and the tweeter down -3 dB relative to midrange. See my profile for where the speaker is located in the room.

The question that I have: does the Belle's tweeter NEED the help above 4.5 KHz? My gut tells me that this is a really bad idea in terms of polar nulls and other interesting near-field artifacts. Would you recommend that I increase the gain on the tweeter channel since the midrange is now being low-passed at the 4.5 KHz point, -24 dB/octave, or leave it as-is (assuming that I got it right to start with)?

Chris

[Al Klappenberger]

Chris,

I think your asking if the K77 needs help from the K55 above 6KHz. It's kind of a moot point considering it gets very little! The K55V poops out above 6 Khz, The Klipsch stock networks depend on that to form a crossover. The K55M doesn't go quite as high as the K55V. I think the K77 does fine by itself except above about 16 KHz or so where it poops out. I doubt that matters to most of us though. I can't even hear 15 Khz any more let alone anything higher!

There is definitely a different sound when the mid-range high range is actually limited by a filter rather then allowing it to roll off naturally. That was driven home to me when I did a side-by-side comparison of a Forte I with the stock network A/B switched with one having my network upgrade. If that difference is "help" I suppose is a matter of opinion. I consider it driver interference from multiple sources since they are by no means in time alignment!

Al K.

[Chris A]

Thanks for the quick response.

Do the relative driver gains that I mention above sound about right? I remember reading from your posts something about -6.2 dB on the midrange, and a another on the -3dB on the K77 vs. a small positive gain on the Crites-Eminence) tweeter.

I assume that the CT-125 tweeter goes lower and that makes up for its lower sensitivity vis-a-vis the K-77, and its small FR dip at ~12 KHz. This allows for a lower crossover frequency with the midrange, and avoids the large phase shift due to use at the end the midrange horn/driver capability.

I have corrected (pulled down) some of the bass bin "hump" in the 150 Hz region based on some anechoic data for the Jubscala (whose bass bin is very similar to that of the Belle's). This seemed to neutralize the sound a bit more. I can now tilt the the bottom end of the Belle slightly to join up with the SPUD subs at ~70-80 Hz.

Chris

[Chris A]

If that difference is "help" I suppose is a matter of opinion. I consider it driver interference from multiple sources since they are by no means in time alignment!

I agree wholeheartedly.

The whole subject of these Klipsch Heritage crossovers is a very interesting design approach but not one that I would take I'm afraid. In terms of minimizing parts count and increasing inherent reliability of the as-built units, the design philosophy clearly makes a choice toward those performance criteria and downplays crossover design practice: namely the typical efforts to minimize the crossover interference regions (and thus the undesirable polar patterns) via steeper slopes, and trying to cross the drivers at an mutual region where their inherent phases aren't going off in opposite directions at a fast pace. This allows the electrical network to control the design performance and uncouples the resulting performance from mfg. variability of drivers batch-to-batch.

Chris

[Al Klappenberger]

Chris,

I forgot to mention the level thing. The -6 dB or so on the squawker is about right. The tweeter level is a matter of what tweeter you have. As it turns out, Bob's CT125 is just right for the wide-open tweeter channel of my Universal network requiring no external attenuator. I think his tweeter is less sensitive because of it's smaller magnet. That was necessary because a larger one would not have fit directly in place of the K77 under the Khorn top cover. That might also be why Klipsch made the K77m magnet square. It had to have more sensitivity to make up for the loss in the crumby AA tweeter filter.

Al K.

[Trey Cannon]

Al,

Are your curves on just the network with resistive loads?

[Al Klappenberger]

Trey,

Always test filters of any sort loaded with their designed impedance. I load the tweeter channel with 8 Ohms and the mid-range with 13 Ohms to simulate the K55V. The woofer channel gets loaded with 6 Ohms in series with 1 MHy. That's a slight fudge because the driver doesn't function like a series inductor to the transfer function. It just simulates the complex load the filter sees. The computer analysis does the same thing.

Al K.

[Guest " "]

Trey,

Always test filters of any sort loaded with their designed impedance. I load the tweeter channel with 8 Ohms and the mid-range with 13 Ohms to simulate the K55V. The woofer channel gets loaded with 6 Ohms in series with 1 MHy. That's a slight fudge because the driver doesn't function like a series inductor to the transfer function. It just simulates the complex load the filter sees. The computer analysis does the same thing.

Al K.

oh...interesting

[Trey Cannon]

When testing a network here looking at the voltage transfer only what is how we do it. That data will get you close to the acoustic xover but not right on it.

As you say, the networks are built with the driver in mind. You can do the same test with the driver as the load and see the difference between the two.

In the end it is still the electrical potation of the equation. When you add the drivers and air to the mix things change a bit.

When looking at just the electronics of the speaker this is very valid data.

[Al Klappenberger]

Trey,

I did a test once where I substituted the real drivers for the dummy loads when doing a network test. The results confirmed that the dummy load test results were close enough. What the drivers do whith what they get from the network is another story completely! I provide computer run tests results on all the networks I make but state clearly that it is for the network only! I do it partly to make sure I didn't forget something, sort-off like a checklist and to check matching between the two.

Al K.

[Trey Cannon]

Thats a very good habit to have and I would expect noting less.

Do you have a mic that you can use for testing?

Using nearfield and gated test you can get a very good look at what the speaker is doing without the room.

[Al Klappenberger]

Trey,

Yes, I have three good mikes. I have an old Colony Mitty Mike II (an electret), a B&K 4133 and a Larson Davis 2559. I have a B&K 4220 pistonphone for level calibration. I use them with a Hewlett-Packard 3563a analyzer. My problem is no anechoic environment, The best I can do is with a pulse, but I usually use white noise. I have had fairly good results with on-axis measurements, but measuring dispersion on horns has not been easy!

I like the noise method because I place a reference attenuator from the amp output into one channel of the analyzer set to equal the Mitty Mike output at 100 dB SPL. This makes the analyzer display 0 dB = 100 dB SPL at the mike no matter how loud I run the test. It's accurate so long as the mike is exactly 1 Meter from the speaker. The gated method is uncalibrated for level.

I usually use the Mitty mike for response testing because it's a 1/4 inch and flat to 20 Khz. I mainly use the other two mikes for distortion tests, The Mitty mike starts to overload slightly beyond about 110 dB. The other two take a jet engine to overload them!

AL K.

[Trey Cannon]

If you have a port on the speaker, test with the mic in the port then convert the presure number into a SPL number. test the woofer at 1m then change the converted presure number's level to match the woofer 1m level. Splice the two curves together for a true LF curve. Test the system at 3m gated for a sum curve. Add 9.5db to the 3m curve to make it a "1m" curve and splice it to the LF curve...now you should have a good total curve.

If the gated curve is not at the right level you can just match the LF curve level and you should be very close.

Food for the brain