Crossover Roadshow

[BEC]

Bob what all of Als or what were the details of the crossover network did you use and what were your thoughts on in.

That was Al's AP12-350 network. I thought it was OK. My network and his drew almost identical acoustic curves and sound was about the same. When testing just his crossover with the spectrum analyzer, his tested bad at first, but not his fault. The swamping resistor drew so much current from my test rig that the voltage level dropped off as the frequency went up. I had to isolate his crossover load from the test equipment using an amplifier, then the test went fine.

[NOSValves]

And, for those who like to listen to 1khz test tones, on a Khorn, I think that by the time you got to about 113db out of the Khorn midrange, you would cook the resistor.

Sure, that makes sense.

I used about 100 watts into the crossover for about an hour.

PLaying music? Do you mean 100W peaking on an amplifier meter?

I mean the equivalent to 100 watts average as measured in voltage at the input to the crossover. See, we were trying to prove that the Faital HF200 driver could take the amount of power it could see used in a Khorn.

So let me get this straight. You pumped 100 watts into this test rig with the actual sound it produced coming out of the speakers? Were you in the room? Did you promptly go out and buy some hearing aids?

What is business getting so slow you need to come into this thread dumping this hogwash? I've owned a set of super AA for about 5 years. At times I push the SLP level pretty good. The swamping resistor NEVER gets even warm in the real world listening to music. For instance my system has been on playing music for 48 hours burning in a vintage amp we just installed some 70 new components..........I just felt the swamping resistor after 10 minute of letting it rip while I took my dog for a walk.... yup so loud I would not even want to be in the room with it...... the swamping resistors are at room temperature.

The fact of the matter is at 100 watts playing music into any Klipsch heritage speaker the distortion of the speaker itself would be so high no normal person could stand it for more then a brief few second regardless of the quality of the amplifier used. Do the math for the maximum power required to reach the maximum SLP rating of the Klipsch heritage speaker..............it is not a 100 watts and even if it was the distortion to any person with reasonable hearing ability would be impossible to stand.

[BEC]

And, for those who like to listen to 1khz test tones, on a Khorn, I think that by the time you got to about 113db out of the Khorn midrange, you would cook the resistor.

Sure, that makes sense.

I used about 100 watts into the crossover for about an hour.

PLaying music? Do you mean 100W peaking on an amplifier meter?

I mean the equivalent to 100 watts average as measured in voltage at the input to the crossover. See, we were trying to prove that the Faital HF200 driver could take the amount of power it could see used in a Khorn.

So let me get this straight. You pumped 100 watts into this test rig with the actual sound it produced coming out of the speakers? Were you in the room? Did you promptly go out and buy some hearing aids?

What is business getting so slow you need to come into this thread dumping this hogwash? I've owned a set of super AA for about 5 years. At times I push the SLP level pretty good. The swamping resistor NEVER gets even warm in the real world listening to music. For instance my system has been on playing music for 48 hours burning in a vintage amp we just installed some 70 new components..........I just felt the swamping resistor after 10 minute of letting it rip while I took my dog for a walk.... yup so loud I would not even want to be in the room with it...... the swamping resistors are at room temperature.

The fact of the matter is at 100 watts playing music into any Klipsch heritage speaker the distortion of the speaker itself would be so high no normal person could stand it for more then a brief few second regardless of the quality of the amplifier used. Do the math for the maximum power required to reach the maximum SLP rating of the Klipsch heritage speaker..............it is not a 100 watts and even if it was the distortion to any person with reasonable hearing ability would be impossible to stand.

Craig,

Did you read that this was a test, proving a concept that a driver and some crossovers could survive through a torture test? And did you read that they passed the test? If you are going to comment at least pay attention to what is written.

[BEC]

"Do the math for the maximum power required to reach the maximum SLP
rating of the Klipsch heritage speaker..............it is not a 100
watts..."

Craig, here is that math done for you by PWK.

[Deang]

Craig is referring to Klipschs' spec that states a maximum output of 120dB. Using the way most calculate this out, that can be acheived with around 50 watts. A hundred watts pushes it past this rated spec. Kind of a curious thing I've always wondered about.

104dB, 1w
107dB, 2w
110dB, 4w
113dB, 8w
116dB, 16w
119dB, 32w
122dB, 64w
125dB, 128w

You never really said it was a "torture test", but yeah, I guess it sure sounds like one. I can't tell what's being tortured the most though -- the loudspeaker or the listener

"My network and his drew almost identical acoustic curves and sound was about the same."

I'm having some trouble with this, especially the last part.

Al is curious about the test set up, how and what things were connected together and the signal type(s) used. Having to isolate the network from the test equipment is something he's never had to do.

Looking at John's impedance plot and comparing it to Heyser's, I think saying that a Heritage loudspeaker with one of the older crossovers having a "nominal" impedance of 8 ohms is a bit of a stretch. I think if one were to average that out, it would look more like 20 ohms. I can't help but wonder what the real world effect on an amplifier is with that number. Halve the impedance and the available power from the amplifier almost doubles, double the impedance and the available power from the amp goes down. So, if you swamp the high pass of the crossover back to 8 ohms, how does that result in more net power loss from an amplifier that was previously seeing a higher impedance?

I think your test put a load on the resistor that no one will ever see during normal usage. Even if they were to drive them to such levels, your test proved that the resistor won't fry. Your test also proved that the resistor isn't getting more than 10 watts. There has never been a resistor failure, and no one has ever complained or commented about unusual behavior from their amplifier after swapping out crossovers.

You have to attenuate to balance the output of the drivers. The resistor absorbs power as part of attenuation, it's power not being used anyway, and it's contained to one driver, not all of them like conventional crossovers methodologies. I think the AK-5 crossover has about a half dozen resistors in it and six times the parts -- and they actually raised the efficiency rating of the loudspeaker!

This SINGLE resistor helps to provide a valuable feature and so far every person except one has preferred the sound to what they had.

I have a lot of email responses from Al, and I will post them if you or anyone else would like to see them. I'm really trying to prevent a "I think thou dost protest too much" situation -- what is the purpose of defending something that already has such an outstanding reputation and track record?

[NOSValves]

"Do the math for the maximum power required to reach the maximum SLP rating of the Klipsch heritage speaker..............it is not a 100 watts..."

Craig, here is that math done for you by PWK.

I don't need to read all the posts to see the intent.... you have no business posting in this thread at all.

[BEC]

"Do the math for the maximum power required to reach the maximum SLP rating of the Klipsch heritage speaker..............it is not a 100 watts..."

Craig, here is that math done for you by PWK.

I don't need to read all the posts to see the intent.... you have no business posting in this thread at all.

I wonder why we can no longer have technical discussions on this forum without somebody getting offended. That used to be the fun part of this forum for me. Now, just does not seem worth it.

[BEC]

"Al is curious about the test set up, how and what things were connected
together and the signal type(s) used. Having to isolate the network from
the test equipment is something he's never had to do."

Have to say one more thing though.

Well, Al can just go back and look at the dozens of emails back and forth between Al and me during that testing. He has all the traces and all the test equipment setup that was being used. He and I were working together on that test.

[The Dude]

My network and his drew almost identical acoustic curves and sound was about the same.

Are you referring to your a/4500 network.

[BEC]

My network and his drew almost identical acoustic curves and sound was about the same.

Are you referring to your a/4500 network.

No, it was a 2-way I put together for the test we were doing.

[The Dude]

Some have asked my impressions of the network, I have recently played around with the autoformer taps. Over the weekend I have had the time to try 2, the first one was Friday night though I didn't get to do some detailed listening it sounded pretty good, that was at 0-1 which would be -12db attenuation, today is a little different I went a little more, thought if I made a bigger jump it would help me distinguish the difference and it did. This time I went to 2-3 which is -16.7db. At first surprisingly I liked this a little more, it made the bass more predominate, and the vocals are outstanding. After listening a little it all most makes the bass seem a little boomy, but that could be my room response as its not the most adequate room for good sound, and the cymbals seem to be overly bright, maybe not so much bright be predominate. If I have time today I will drop it back down and give either -13.7 or -15 a try. The nice thing about the EV HP640 horn is I think its big enough that it produces a big enough sound so if the mid is attenuated a little it wont rob to much mids from the system(but that's just my speculation). I all so don't feel that the swamping resistor robs enough power from the amp for me to notice or make a big deal to me. I hope this helps anyone that is thinking of giving these a try.

[The Dude]

My network and his drew almost identical acoustic curves and sound was about the same.

Are you referring to your a/4500 network.

No, it was a 2-way I put together for the test we were doing.

Thank you for clearing that up.

[seti]

"Do the math for the maximum power required to reach the maximum SLP rating of the Klipsch heritage speaker..............it is not a 100 watts..."

Craig, here is that math done for you by PWK.

I don't need to read all the posts to see the intent.... you have no business posting in this thread at all.

I wonder why we can no longer have technical discussions on this forum without somebody getting offended. That used to be the fun part of this forum for me. Now, just does not seem worth it.

No kidding I guess this is no longer a public forum where anyone can post. The hostility is rediculous.

[mike stehr]

No kidding I guess this is no longer a public forum where anyone can post. The hostility is rediculous.

Hey man, you have no business posting in this thread either...

[Deang]

Everybody relax.

I was talking to Craig on the phone today and told him I was tired of talking about the damn resistor.

Bob, honestly, over the years, it seems like every time one of these networks is discussed you feel the need to come in and make some comment related to the resistive element. Your first contribution to this thread was:

"Here is what Paul Klipsch had to say about impedance. He was defending the use of the autotransformer which raises impedance instead of using resistive attenuation."

I think you've posted that document in relation to this topic about a dozen times or more over the years.

It doesn't always feel as if it's a "technical discussion" you're looking for, but more of an opportunity to undermine the performance of the design -- I believe this is what Craig was alluding to. To be honest, it just seemed out of place in this thread, but maybe not, so I responded to your concerns and criticisms just like I always do.

[Deang]

Several are asking where we're at with this thing. Here are the next six.

Rusty Melton
David McDonald
Dave Morasco
David Beeman
David Hansen
Saul Gelfand
Billy Strother

[Deang]

From Al:

Here is a computer power distribution analysis done on my Universal network. It assumes a constant RMS voltage into the network corresponding to a totally RIDICULOUS 100W at 8 Ohms swept across the audio spectrum. THIS WILL NEVER HAPPEN IN THE REAL WORLD! It represents a situation contrived intentionally to destroy a loudspeaker.

The autoformer is branch 1 on the schematic diagram at the left. It has a turns ratio of 1.87:1 which is reducing the squawker level power by 1 / 1.87 squared = 0.286:1 corresponding to an attenuation of 10 X Log10(.286) = 5.44 dB. This is an attenuation of roughly 1/4 of the input power sent to the midrange driver.

Branch 2 is the 10 Ohm swamping resistor. The red trace on the plot is the curve of power dissipated in that resistor with frequency. Note that it is absorbing roughly 75W at 2500 Hz. That is the remainder of the 100W input NOT NEEDED by the midrange driver AT 2500 Hz ONLY. 1/4 of the power to the driver and 3/4 to the swamping resistor. The power absorbed by the resistor drops off at other frequencies. It's on the down-stream side of the midrange bandpass filter of the network. It "sees" ONLY MIDRANGE ENERGY!

Note the vertical dotted line marker at 160 Hz. The numeric data to the right shows the loss (green) and attenuation (red) at this marker frequency. This shows that at woofer frequencies, NEARLY ALL OF THE AMP'S POWER IS AVAILABLE TO THE WOOFER. THE SWAMPING RESISTOR IS OUT OF THE PICTURE absorbing only 10.67W of the 100W applied to the network! The swamping resistor absorbs only the power not needed by the midrange and only at midrange frequencies. If this was one of my extreme slope networks rather than the 6 dB / octave slope of the Universal, the power to the swamping resistor would drop like a rock below the woofer / squawker crossover frequency. This is dependent on the low frequency side slope of the network midrange bandpass filter section. This is yet another reason I advocate extreme slope networks! Neither the midrange driver or the swamping resistor would see any of the power at woofer frequencies where the vast majority of power is distributed in audio program material. The speaker would handle more power!

Those who make claims that the swamping resistor is cutting amplifier power in half or that it reduces the efficiency of the loudspeaker simply do not understand its function!

[Deang]

From Al:

Here is a similar power analysis plot of the ES400 + ES5800 combination. Here also we have a 100W input across the audio spectrum. Notice that the power absorbed by the swamping resistor is only 118 MILLIwatts (0.118W) or less at and below 340 Hz with 100W going to the woofer.

[Deang]

From Al,

Here is my universal network using a universally accepted "L-Pad" in place of the transformer and swamping resistor just the way the average hack engineer would do it. Note that the same situation applies. The parallel resistor at branch 2 is dissipating 54.9W at 2647 Hz. with the same 100W input. Additionally, the series resistor at branch 1 is also dissipating an additional 12.8W. That would require an additional plot but the curve is identical. No power is being taken from the woofer, where it needs to be applied.

Nobody ever complains about a L-pad robbing power from the amp, and 100W input would obviously cook the 10W resistors here as well. However, how often do you see L-Pads using 50W resistors in home entertainment speakers? Never, because It is NOT NECESSARY!

[NOSValves]

Al rests his case..... now back to Deans road show thread.