ALK Crossovers

[mikebse2a3]

10 hours ago, Marvel said:

Aren't you talking about two separate things. One, is the use of the swamping resistor to give a consistent load to the amp (better for tube amps that have a higher output impedance than SS amps) and two, making adjustments to the mid driver to tailor the sound to your room (which is what many seem to do on the stock heritage systems.

 

Hey Bruce...... let me try to clarify my points.

 

First as regards to the ... (better for tube amps that have a higher output impedance than SS amps)  most tube amplifiers have enough damping (ie: low internal impedance) that the swamping resistor has little to no audible or measurable effect on them as far as altering the frequency spectrum of the combined amplifier/loudspeaker due to impedance variations of the [Balancing Network/Loudspeaker]. The exception would be tube amplifiers like zero feedback SET types which will exhibit some frequency spectrum shifts with impedance variations of the [Balancing Network/ Loudspeaker] and a very important thing to understand is this spectrum shift could be perceived as audibly better versus worse and as I have said before could be a part (but not the only reason) of why someone might prefer them with a given loudspeaker over other amplifier types. Also just in case anyone thinks "better for tube amplifiers that have a higher output impedance than SS amps"   includes being hard on them or possibly damaging them in some way because of them trying to drive a higher impedance network/loudspeaker design like klipsch can rest easy because it would be a extremely poor amplifier design that couldn't handle the load presented by them...!!!

 

Second point is a quality Analog EQ or DSP processor would be a much better tool and much more flexible at tailoring the sound to your room for those that have limited or no options at dealing with the "real problem" which is the room's acoustics and integration of the loudspeaker with the listener and the room and not the stock crossovers used in the Klipsch Loudspeakers..!!!

 

miketn

[Marvel]

Mike,

That sounds about right to me. I find no fault in your logic and appreciate the further explanation. I switched out my AL networks for the DHA2 design from John Albright, mostly because I didn't like the AL networks. It seemed I had the magic word "synergy" going on and I was really happy with the system. Since I sold that pair of LS, everything has been quiet and I have been running H IIs on SS. My refurbishing some LS has taken far too long but I am making progress to get back to MY nirvana.

 

As always, I appreciate your input.

 

Bruce

 

PS. If money would allow, I would at least try going active, It is in the plans, but pushing into a semi retirement, it may not happen.

[mikebse2a3]

On ‎10‎/‎16‎/‎2016 at 5:20 PM, wdecho said:

I have tried most of the available X-overs with swamping resistors and the A's and AA's and I honestly could not tell that much if any difference between using the swamping resistor if everything else was basically the same. If I were to use a X-over with an autotransformer I would use one with the swamping resistor so that I could change the attenuation to adjust the sound for my room. If you have a stock X-over and like the way it sounds one of Bob Crites X-overs would be a good choice. As for as losing some signal, I do not think it is a serious point. My understanding is PWK liked using an autotransformer for this very reason, no loss with L-pad attenuation. I have 1 1/2 watt SET amps with plenty of power in my room with my LaScala's using only resistors in a L-pad attenuation. Any 5 watt SET should have plenty enough power for 98% percent of listeners with any of the available X-overs. There is plenty of believers in what ALK has to say and probably more believers in what PWK had to say on the subject. One thing for sure the argument will not be settled here. 

 

We will have to agree to disagree on this

 

Most noticeable was on some very good acapella recordings which had captured the dynamic range of the human voice very well and I had no issue when using the AK-3/Klipschorns but the ALK Universal produced audible clipping in my situation and others should know it could be a possible issue with low power amplifiers.

 

miketn

[dBspl]

10 hours ago, mikebse2a3 said:

 

I would have to disagree with the comment "Most of his work was when tubes ruled"  and from Roy's own accounting he was very interested and working with steep slope filters in his later years.

 

For those interested here are the Dope From Hope papers in a searchable pdf:

which includes the DFH Vol 16 No 8 I believe wdecho is referring to.

 

miketn

 

DOPE_from_HOPE!!!_searchable.pdf

 

 

There's another Dope from Hope article that ridicules the whole idea of constant impedance loudspeakers.  I believe it's called "The Ultimate LSH Loudspeaker".  There's nothing like good PWK satire!

 

And I can also confirm PWK spent a lot of time in his later years developing extreme high slop passive filters.

 

Kerry 

[Marvel]

7 hours ago, dBspl said:

... extreme high slop passive filters.

 

You haven't seen my construction abilities, have you?

 

Bruce

[tigerwoodKhorns]

Here it is.

 

I don't understand.  Is he saying that adding this simple L pad is going to reduce efficiency that much? 

DFH Constant Impedance.pdf

[Doug]

18 hours ago, mikebse2a3 said:

For those interested here are the Dope From Hope papers in a searchable pdf:

which includes the DFH Vol 16 No 8 I believe wdecho is referring to.

 

miketn

 

DOPE_from_HOPE!!!_searchable.pdf

 

This is super cool! Thanks for sharing!

 

17 hours ago, mikebse2a3 said:

Most noticeable was on some very good acapella recordings which had captured the dynamic range of the human voice very well and I had no issue when using the AK-3/Klipschorns but the ALK Universal produced audible clipping in my situation and others should know it could be a possible issue with low power amplifiers.

 

In one of AL K's unnecessarily abrasive rants on his website, he points out that the total efficiency of a Klipsch heritage speaker is determined by the woofer, not the squawker since the mids are always attenuated in some way to make up for higher efficiency of the squawker vs. the woofer.  Since there is no attenuation in his universal network between the input and the woofer, it seems to make sense (at least in my physics 101 level of electronics understanding) that the swamping resistor shouldn't decrease the overall efficiency of the loudspeaker.

 

Mike, since "default" setting on the ALK universal seems to be a one tap of additional attenuation on the autotransformer compared to the stock Klipsch networks, is it possible that you simply were listening to the vocal recordings at a higher overall volume with the universals since the mids (where most of the music would be) would be more attenuated with the universal than with the AK3 networks?  Maybe one could simply choose a less attenuated tap on the autotransformer in the universal for this type of music?

[dBspl]

2 hours ago, tigerwoodKhorns said:

Here it is.

 

I don't understand.  Is he saying that adding this simple L pad is going to reduce efficiency that much? 

DFH Constant Impedance.pdf

 

Yep, that's it. Thanks for posting.

 

This should go without saying, but the whole article is tongue & cheek.  It was just to highlight the absurdity of the idea that constant impedance and high power handling are required for good sound reproduction.  This completely goes against everything that PWK believed. 

 

A couple of the footnotes in that article are hilarious.

 

Kerry

[mikebse2a3]

1 hour ago, Doug said:

In one of AL K's unnecessarily abrasive rants on his website, he points out that the total efficiency of a Klipsch heritage speaker is determined by the woofer, not the squawker since the mids are always attenuated in some way to make up for higher efficiency of the squawker vs. the woofer.  Since there is no attenuation in his universal network between the input and the woofer, it seems to make sense (at least in my physics 101 level of electronics understanding) that the swamping resistor shouldn't decrease the overall efficiency of the loudspeaker.

 

It's not decreasing the efficiency of the loudspeaker but it does add an additional load on the amplifier in the squawker frequency range. It's this additional load to low power amplifiers that the users of these amplifiers should be aware of and should take that information into consideration with their unique requirements/expectations from their systems.

 

1 hour ago, Doug said:

Mike, since "default" setting on the ALK universal seems to be a one tap of additional attenuation on the autotransformer compared to the stock Klipsch networks, is it possible that you simply were listening to the vocal recordings at a higher overall volume with the universals since the mids (where most of the music would be) would be more attenuated with the universal than with the AK3 networks?  Maybe one could simply choose a less attenuated tap on the autotransformer in the universal for this type of music?

 

Doug, I had the ALK networks over a year and tried several attenuation levels including one that closely matches the AK-3 level so no it wasn't a factor in my observation. This is simply a case of a limited power source (ie: Low Power Amplifier) being asked to drive an additional load (ie: swamping resistor) and the fact that there will be a reduction in available headroom in the squawker region with this method.

 

 

 

miketn

[dBspl]

12 hours ago, mikebse2a3 said:

 

It's not decreasing the efficiency of the loudspeaker but it does add an additional load on the amplifier in the squawker frequency range. It's this additional load to low power amplifiers that the users of these amplifiers should be aware of and should take that information into consideration with their unique requirements/expectations from their systems.

 

 

Doug, I had the ALK networks over a year and tried several attenuation levels including one that closely matches the AK-3 level so no it wasn't a factor in my observation. This is simply a case of a limited power source (ie: Low Power Amplifier) being asked to drive an additional load (ie: swamping resistor) and the fact that there will be a reduction in available headroom in the squawker region with this method.

 

 

 

miketn

 

It does decrease the efficiency of the loudspeaker system (which includes the network).  It doesn't have any bearing on the efficiency of any individual driver component (by itself).  

 

15 hours ago, Doug said:

This is super cool! Thanks for sharing!

 

In one of AL K's unnecessarily abrasive rants on his website, he points out that the total efficiency of a Klipsch heritage speaker is determined by the woofer, not the squawker since the mids are always attenuated in some way to make up for higher efficiency of the squawker vs. the woofer.  Since there is no attenuation in his universal network between the input and the woofer, it seems to make sense (at least in my physics 101 level of electronics understanding) that the swamping resistor shouldn't decrease the overall efficiency of the loudspeaker.

 

The woofer doesn't determine the total efficiency of a loudspeaker - it essentially establishes a baseline for the sensitivity of the loudspeaker. 

 

Loudspeaker sensitivity and efficiency are two different things.

 

Kerry 

 

[Honeybadger]

 

This splitting hairs over and insignificant loss is silly.

 

Any system with an Lpad  has the same or greater losses than a swamped auto former.

 

I don't see anyone here barking over the capacitive reactance of the 13uF cap in the AA vs: the 48uF in the super AA. (because it's silly)

 

I have enough power to deal with insurmountable loss of a swamping resistor, I feel its benefits outweigh its deficits.

 

If your system doesn't have the power to drive an 8 ohm speaker with an sensitivity of 104 db, you need a better amp.

 

To me this is typical Klipsch clergy VS: vendors!

 

HB

[muel]

What ever sounds best to you is fine with me.  

It does seem to be a pertinent discussion because there does seem to be a slight change in sound.

[mikebse2a3]

HB said...."This splitting hairs over and insignificant loss is silly."

Darn them silly facts....

 

HB said...."Any system with an Lpad  has the same or greater losses than a swamped auto former".

Your statement is to broad to make your assumption valid in every situation.

 

HB said...."I don't see anyone here barking over the capacitive reactance of the 13uF cap in the AA vs: the 48uF in the super AA. (because it's silly)"

No....silly would be to bring it up since we were discussing the facts about what happens when using the swamping resistor in the ALK?

 

HB said...."I have enough power to deal with insurmountable loss of a swamping resistor, I feel its benefits outweigh its deficits."

Would you think I was being silly if I said I'm so happy for you..?

 

HB said...."If your system doesn't have the power to drive an 8 ohm speaker with an sensitivity of 104 db, you need a better amp."

All you low watt amp guys listening......

 

HB said...."To me this is typical Klipsch clergy VS: vendors!"

Now this Is Just "SILLY" IMHO but of course your free to have your own opinion...!

 

Silly Me,

miketn

[Deang]

Mike, I've built many of these type networks, and you're the first person to ever complain about this "problem" in 15 years. I have people running 45 and 2a3 SET with no reported issues. 

 

I understand that there is always a cost involved, but in this particular case it's quite small. I paid John Warren to run a series of tests for me, including a test to determine the impact of the swamping resistor - it was minuscule - and he was unable to reproduce the results shown in the published plot submitted in "The Problem With Attenuators". 

 

Al and me may have our issues, but I still like the earlier Universal and Super AA designs. They provide those who like to experiment with different horns and drivers a very simple and effective means of balancing the outputs of the drivers. 

 

I agree with Honeybadger - this is largely hair splitting, and making something out of nothing. 

 

I didn't see any "facts" presented, just subjective opinions. I posted Al's white paper on this topic - and no rebuttal was offered.  

[muel]

Yep, I don't have much in the way of facts to offer... just my experience with almost every model of ALK network, your "super" AA, and various original networks with various different caps.  I've been glad to offer my subjective opinion while trying to offer pertinent details such as other equipment and room information.  I haven't always been good about changing just 1 variable at a time so sometimes I could not be sure of which variable affected the sound... only the net effect.  Still, I offer my anecdotes because it has been that same kind of information offered to me that has brought me so far towards great sound.

 

There is just as much place for anecdotes here as there is for "facts."    We just shouldn't present anecdotes AS fact and we should hesitate to generalize too quickly (My experience might differ from yours).  It should be obvious that we need to understand the science (of whatever) but it isn't the only thing.

[dBspl]

5 hours ago, Deang said:

Mike, I've built many of these type networks, and you're the first person to ever complain about this "problem" in 15 years. I have people running 45 and 2a3 SET with no reported issues. 

 

I understand that there is always a cost involved, but in this particular case it's quite small. I paid John Warren to run a series of tests for me, including a test to determine the impact of the swamping resistor - it was minuscule - and he was unable to reproduce the results shown in the published plot submitted in "The Problem With Attenuators". 

 

Al and me may have our issues, but I still like the earlier Universal and Super AA designs. They provide those who like to experiment with different horns and drivers a very simple and effective means of balancing the outputs of the drivers. 

 

I agree with Honeybadger - this is largely hair splitting, and making something out of nothing. 

 

I didn't see any "facts" presented, just subjective opinions. I posted Al's white paper on this topic - and no rebuttal was offered.  

 

We can all enjoy this hobby in whatever way that makes each of us happy. And if this is a solution you're happy with, then so be it. But I'm not seeing how you take a normal 30 ohm load down to 8 ohms (from 500 Hz up) and claim it's insignificant, or minuscule.  It looks like a significant percentage of the amplifier power above 500 Hz is simply being dumped in the resistor.  It just hits me as a crude solution.  

 

Kerry

[mikebse2a3]

Deag said: "Mike, I've built many of these type networks, and you're the first person to ever complain about this "problem" in 15 years. I have people running 45 and 2a3 SET with no reported issues." 

 

That's great I'm happy for them

 

 

6 hours ago, Deang said:

I understand that there is always a cost involved, but in this particular case it's quite small. I paid John Warren to run a series of tests for me, including a test to determine the impact of the swamping resistor - it was minuscule - and he was unable to reproduce the results shown in the published plot submitted in "The Problem With Attenuators". 

 

In what way was it's impact minuscule? 

Please share these test otherwise it would be pointless to discuss something if I don't have all the information.

 

6 hours ago, Deang said:

Al and me may have our issues, but I still like the earlier Universal and Super AA designs. They provide those who like to experiment with different horns and drivers a very simple and effective means of balancing the outputs of the drivers. 

 

That's all well and good but my comments have been about a possible drawback for low power amplifier users when using the ALK (with swamping resistor) versus various Klipsch Balancing Networks where the impedance is allowed to rise resulting in an easier load in the midrange and more headroom as a result.

 

6 hours ago, Deang said:

I agree with Honeybadger - this is largely hair splitting, and making something out of nothing.

 

I'm so glad you both have found some common ground considering some of the post exchanges I've read between you both....

 

6 hours ago, Deang said:

I didn't see any "facts" presented, just subjective opinions. I posted Al's white paper on this topic - and no rebuttal was offered

 

Dean do you need to borrow my reading glasses because they were in plain sight in my post as well as post by Kerry (dBspl)... ?

 

miketn

[mikebse2a3]

Let me try to demonstrate my point about why wasting power with a swamping resistor (which is placed in parallel with the Autotransformer/Squawker of the ALK network) should be taken into consideration as it pertains to available power if you have low wattage amplifiers.

 

Please note the following:

(1)    2.83v/8ohm  = 1 watt from the amplifier and 1 watt into the KHorn produces approximately 104db at 1 meter.

(2)    To simplify things we are going to focus on the spectrum of the midrange of the (AA Network/Klipschorn) which exhibits a rise in impedance that peaks to about 30 ohms around the 2khz region.

(3)   Now another calculation to remember for my example is 5.66v/8ohm = 4 watt from an amplifier. Low Wattage Amplifier owners using the ALK Network especially take note of this fact..!

(4)  So lets use a low wattage amplifier capable of 4 watts maximum for this demonstration.

 

 

In the case of a [AA Network/Klipschorn] if we apply 5.66v at 2khz were the impedance is approximately 30 ohms we only require 1.067 watts from our amplifier.

Result: Our low wattage amplifier isn't even close to reaching full capability to supply power in this midrange spectrum.

 

In the case of the ALK Universal Network we have a parallel 10 ohm swamping resistor with the 30 ohms (Autotransformer/Squawker) and at 2khz it is reported to give approximately 8 ohm load to the amplifier. Now if we apply 5.66v at 2khz to the [ALK Network/Klipschorn] the Autotransformer/Squawker is still seen as 30 ohms by the amplifier and will still require 1.067 watts but we have now added an additional parallel load on the amplifier of the 10 ohm swamping resistor and it will require 3.2 watts from the amplifier to obtain the approximately constant impedance of 8 ohms of the [ALK Network/Klipschorn].  Note: There is a small amount of series resistance due to the additional band pass components of the ALK that has been ignored in this example for clarity and is why if you caught it the 1.067 watts + 3.2 watts = 4.267 watts and that's because 10 ohm in parallel with 30 ohms = 7.5 ohms and this additional series resistance gets us close to the claimed 8 ohm of the ALK and we will actually be very close to the 4 watt requirement from the amplifier from my previous calculation example.

Result: Now our low wattage amplifier is at it's upper capability limit in the midrange spectrum.

 

So now how does this relate to listening levels.

Lets take our 104db Khorn at 1 watt at 1 meter and calculate using the inverse square law the spl at a listener position 12 ' back from the speaker and this = 93db at the listener position. (Note: depending on room acoustics the actual spl at the listener position will likely be a little higher than calculated for small enclosed spaces).

 

Now recordings with wide frequency spectrum and especially well recorded dynamics will stress the capabilities of the amplifier much more than my simple example so more headroom from the system should be taken into account otherwise clipping will be a real possibility.

 

One last thing I want to say on this:  Everyone has to decide what works best for them and I hope discussing the facts and how things work will help inform all those interested and to those that want to call this "hair splitting and making something out of nothing" or "Silly" with nothing better to offer thanks for this valuable input on the subject.

 

 miketn

 

Edit: By the way I haven't even mentioned what this means for distortion from the amplifier/loudspeaker but many amplifiers and especially SET types will exhibit lower distortion at lower power levels so it shouldn't come as a surprise that at 1.067 watts versus 4 watts the [AA/Klipschorn/Amplifier] has the potential to produce lower midrange distortion versus the [ALK/Klipschorn/Amplifier] in this example.

 

[mark1101]

so if the total load is 8 ohms which is normal for a speaker then the amp can't drive the speaker?  So what happens on the Klipsch network when the impedance drops to the normal levels of 8 ohms?

[mikebse2a3]

5 minutes ago, mark1101 said:

This math is definitely incorrect.  The swamping resistor in this example presents the following in parallel with the autoformer:

 

(30 x 10) / (30 + 10) ohms to the amp which is about 7.5 ohms that is the whole point of a swamping resistor.........the amp sees 7.5 ohms from the network.   Resistors in parallel.

 

Mark please reread my post.....I said "Note: There is a small amount of series resistance due to the additional band pass components of the ALK that has been ignored in this example for clarity and is why if you caught it the 1.067 watts + 3.2 watts = 4.267 watts and that's because 10 ohm in parallel with 30 ohms = 7.5 ohms and this additional series resistance gets us close to the claimed 8 ohm of the ALK"