ALK Crossover tap settings revised

[Al Klappenberger]

Hi crossover network fans,

They say confession is good for the sole, so here goes...

I think it's time I came clean on a minor error in the data I have provided

on squawker level settings for my Khorn / LaScala crossover network. When I

initially did the calculation for the table of squawker attenuation versus

transformer tap settings I did it with the computer analysis program. I

forgot to change the assumed source impedance from 8 Ohms to the more

realistic .1 Ohms that modern amps really represent. The result of this was

an error of a dB or so in the actual attenuation. I haven't said anything

about it until now because everyone was using the attenuation numbers for

comparison and the error is not really significant. I am about to present

the corrected table and suggest that settings be referred to by tap

settings, like 5-2 or 3-0 rather then the dB level the setting provides.

There is no room for doubt that way.

I have decided to add an additional wire and plug for transformer tap #4 on

every network I build from now on (WORK!). I didn't supply data on tap #4

before. This will give a few more intermediate settings. The attenuation in

the table is the actual measured values and includes the roughly .6 dB

insertion loss of the squawker filter.

I have also revised my recommendations on the 10 Ohm swamping resistor too.

The table indicates the measured attenuation and theoretical impedance match

ratio using different tap connections and R values. I verified the

calculated impedance data using the impedance bridge and found it to be

pretty close. On the table, a ratio of 1.00:1 indicates a perfect match. Any

ratio lower than 1.50:1 can be considered good and will not upset the

response of the squawker filter. By this criteria, the worst case mismatch

of 1.22:1 is very good indicating that resistor changes are not really

necessary for any squawker level setting. The extreme limit for the resistor

value would be with no squawker driver connected at all. In that case, the

resistor value should be 8 Ohms. With the 10 Ohm resistor, that error would

only be 10 / 8 or 1.25:1.

I would like to hear from everyone who has experimented with different

settings. I would like to know what most people think is best. The reason I

ask is that the 2-5 setting is not quite the same level as the stock Klipsch

AA and AB networks were set. It's actually puts the squawker about 1.5 dB

lower (quieter).

Here's the new table:

Squawker R Atten Zo match

(-) (+) Ohms dB ratio

--- --- ------- ----- ---------

_ 5 - x 13.7 --- 1.00:1

* 5 - x 10.0 2.6 1.22:1

_ 5 - 1 12.3 --- 1.00:1

* 5 - 1 10.0 3.4 1.15:1

_ 4 - 0 11.6 --- 1.00:1

* 4 - 0 10.0 3.8 1.11:1

_ 5 - 2 10.8 --- 1.00:1

* 5 - 2 10.0 4.6 1.06:1

_ 4 - x 09.9 --- 1.00:1

* 4 - x 10.0 6.2 1.02:1

_ 3 - 0 09.6 --- 1.00:1

* 3 - 0 10.0 6.7 1.05:1

_ 4 - 1 09.2 --- 1.00:1

* 4 - 1 10.0 7.4 1.08:1

_ 2 - 0 09.0 --- 1.00:1

* 2 - 0 10.0 9.5 1.14:1

_ 3 - x 08.8 ---- 1.00:1

* 3 - x 10.0 10.4 1.16:1

* 1 - 0 10.0 12.5 1.19:1

* x - 0 10.0 15.4 1.22:1

* Settings using standard 10 Ohm swamping resistor.

Al K.

This message has been edited by Al Klappenberger on 05-03-2002 at 05:54 AM

[ChemPharma]

Hey Al. I agree with those revised numbers. I think the 4-x setting should be there too! Its right around the sweet spot (FOR ME) between 5-2 to 3-0! Next time your in the D.C. area, bring your tool apron and stop in and revise mine.. Would ya?

ChemPharma

[Al Klappenberger]

ChemPharma,

Now, NOW.. Doctors don't make house calls these days!

Actually, the setting of 4 - x is going to require an additional wire and plug. Both 4 and x are on the back side of the transformer. That means you can't just move one of the other wires. The wires to the front taps are all probably to short! I could mail you a set of wires and plugs if you have a soldering iron to connect them. The chances of me getting to DC are pretty slim!

Al K.

[Neek]

Thanks for the update Al. I've tried 5-2, 3-0, & 4-X settings using a 10 Ohm resistor on my Khorns, and prefer the 4-X setting. I don't know much about db's but 4-X sounds by far the best to me. I've also thought about installing a selector switch to quickly change taps to compensate for bad recordings but haven't got around to it yet. Has anyone tried this?

[JohnA]

For now I'm using 3-0. The squawker seems a little "recessed" sometimes and the tweeter a little hot. I've been planning to take one to a co-worker's house to do some testing on the bass horn. I'll try to get curves on the squawker, too. Al, do the new calculations change your recommended tap selection?

John

[scottmjackson]

Al:

As always I'm still in your corner. Good job! Would you mind? What are your contact points these days. Looking forward to hearing from you again. Scott

------------------

Scott Jackson

Palo Alto, CA

[Al Klappenberger]

Hi guys,

It seems that most of you like the squawker level set a lot lower than "stock". Interesting!

I have thought a lot about the idea of a switch for the level settings. It would be very nice, but it would run the price up a lot. Switches are expecsive. Also it would require a two pole switch with about 10 postitions to do it right. It would take a long time to wire it too.

By the way, there has been no changes to the netwrok itself besides adding an extra wire and plug on transformer tap 4. The new table is just corrected data.

Al K.

[ChemPharma]

What is the "stock" setting on the AA network Al?

This message has been edited by ChemPharma on 05-03-2002 at 07:23 AM

[Guy Landau]

If the speakers are driven by an 8 ohm tap of a tube amp,should the 4-x be used as well?

[ChemPharma]

Al, I think you answered my last question in your original post. At a 5-2 setting, you said this actually puts the squawker at about 1.5 db lower dB (5-2 @4.6 dB, then stock = about 6.1dB)than "stock".This explains why the 4-x seems favorable (putting it closest, at 6.2, to the original/"stock" ouput level, except cleaner and more detailed). I would like to try that setting sometime once I decide to solder a new wire and plug. The 3-0 setting will suffice for now.

So, I think the results thus far are that we prefer the settings close to the "stock" or higher (4-x @theoretical 6.2dB vs. 6.1 stock and 3-0 @6.7 dB or .6 dB > "stock"), not lower.

This message has been edited by ChemPharma on 05-03-2002 at 11:35 AM

[Hardhead]

I think there's more to the situation than it appears, folks, and it concerns the relative outputs of the K-77 tweeter versus the K-77-M tweeter compared to the squawker.

The AA crossover had a very long life, and it was still being used when Klipsch made the change from the K-77 tweeter (the round Alnico magnet) to the K-77-M (the square "mud" magnet). Your speakers have the K-77-M tweeters, as I recall, and, as you say, the 5-2 setting on your ALK crossover lowers the squawker output to 1.5db below the K-77-M output.

Many Klipsch speakers made before 1980, however, had the K-77 tweeter installed. That tweeter has a lower output than the K-77-M by about 2-3db, so the stock ALK crossover setting has the output of the squawker down by only 0.5 to 1.5db as compared to the K-77, and that's a pretty good match. That's the reason your stock settings were just right for my 1973 Klipschorns with K-77 tweeters.

A few weeks after I had installed the ALK crossovers, I swapped the K-77's for K-77-M's. It sounded terrible because the K-77-M's were screaming loud compared to the squawkers. Now I know why. The output of the squawkers was still 1.5db down, but the output of the tweeters was now 2-3db up. In other words, the output of the tweeters was now 3.5-4.5db higher than the squawkers because the K-77-M's are more efficient. The 1.5db "problem" had been magnified.

My point is that the original 5-2 setting on the ALK crossovers may not be out of line for speakers with K-77 tweeters, and I would be willing to bet that problems with the 5-2 settings all involve K-77-M's. Of course, what your ears tell you is best, and your mileage may vary, but for those of us with K-77 tweeters, Al's 5-2 setting is a good place to start.

So, Al, you still done good.

[ChemPharma]

Hardhead,

I have the old round alnico magnets also (78' K-77's).

I think the 5-2 setting is a good place to start as well. I opted for the 3-0 however, after switching back and forth. The 4-X is right between the two, and I think Al incorporating the number 4 wire and connector from here on out is a good move! I bet the 4-x is the sweet spot for many of us! There are a number of factors though as you say; speaker model (khorns or scalas, k-77M or k-77, ampiflication-Class A, tube etc..). ChemPharma

This message has been edited by ChemPharma on 05-03-2002 at 01:14 PM

[Al Klappenberger]

Everybody,

Thanks for all the input.

You may have a point about the tweeters becasue this is what will dictate the best setting. There is more to this that you might think too becasue there is a 3 dB difference in the loss in the tweeter filter between the AA and their other networks. The tweeter filter in the AA has 3 dB MORE loss than the new networks using the so-called "elliptic filter". The tweeter filter in mine has only 0.3 dB loss at 17 KHz. This would imply a higher level setting on the transformer, but that does not seem to be the trend!

By the way, the Klipsch networks are set at about 3 to 4 dB on the squawker. The Klipsch T2 transformer has the same ratios as the one I use and it is set at 4-0. In other words, a lot hoter than the 5-2 and 4-1 settings you all seem to like.

Go figure! 8^)

Al K.

[edster00]

Well I just left my crossovers how they were when Al sent to me...now I suppose I am going to have to start "messing" with my ALKs! Is there no end to this madness ??

This message has been edited by edster00 on 05-03-2002 at 02:05 PM

[Tom Mobley]

Nope.

[JohnA]

So the setting that replicates the Klipsch squawker output, losses and all would be 4-0?

John

[Al Klappenberger]

Like I said: GO FIGURE!!

Al K.

[Neek]

Al, I was kinda thinking maybe a 3PDT switch would be less expensive. I could put my favorite setting (4-X)in the middle position with one lower and one higher on each side.

[Al Klappenberger]

Neek,

Are your networks DIY or ones I built?

Have you actually been using the 4-x setting?

Al K

[Neek]

Al-

Mine are DIY and yes, I've been using the 4-X tap. I tried both the 5-2 and 3-0 taps and like the 4-X best. I have older model Khorns with the K-77 tweeter and

K-55V squaker. I've been very impresssed the improvements your design has made, thank you! Do you think the source equipment is also contributing the different settings people are using?