Q for Al K

[John Warren]

Al-

Have you measured the impedence vs. frequency for the K55M with and without horn attached? If so, is a post of it (graph, table) possible? (I'll take anything you've got, my signal generator is kaput).

TIA,

jw

[Al Klappenberger]

John,

Yes, I have made curves of it, but they are long gone! All I can tell you from memory is that it's average impedance is 13 Ohms or so on a horn (thirteen).

Al K.

[John Warren]

Al-Do you have a graph of measured (not predicted) resultant impedance vs. frequency on a Klipschorn for your network? If so can you post.

TIA,

jw

[Al Klappenberger]

John

No.. I don't have a Khorn to take the measurement on! I consentrated on making the network provide as constnat an impedance as I could assuming perfect loads. Measurements were taken by a friend using a Khorn that confirmed that the impedance was flater than with the stock network, but I can't post the data at his request.

Al K.

[John Warren]

How do you model the autoformer?

[Al Klappenberger]

John,

I am not sure exactly what you mean, but the autotransformer is modeled in my analysis program as a simple impedance step. The impedance ratio is the square of the turns ratio.

Al K

[John Warren]

Al-I follow what you were trying to do with your network, the 3-way constant resistance network and its equations have been around a long time. And, as you clearly know, speakers aren't resistors and constant resistance networks "aren't" constant resistance BUT they can be made to sound good. Your network looks like a textbook example of one "except" for the autoformer.

Is the K-55 V with autoformer 32 Ohm ??

This message has been edited by John Warren on 03-07-2002 at 09:39 PM

[Al Klappenberger]

Mdeneen,

I have an x-y plotter. That is like a strip recorder but operates on a single sheet of paper. I like that better. You can get single sheets of paper anyhwere!

John,

Yes, my netwrok is basically Butterworth prototype desing, but I have done computer optimization to make the N=2 lowpass work propery with the N=3 highpass in the squawker / tweeter crossover.

The transformer steps the 13 Ohm load of the K55V up as the square of the turns ratio. I then swamp it back down to 8 Ohms with the 10 Ohm resistor. That terminates the N=2 lowpass filter with the right impedance. That's very important for proper filter response. Of course, the tweeter filter sees the K77 load, whatever it happens to be. The lowpass filter sees a more stable load becasue the 10 Ohm resistor portion of the total load swamps out the variations of the K55.

Al K.

[John Warren]

I missed the resistor.

I have to work backwards here becasue I have no experimental data. if the 10 ohm resistor results in an 8 ohm squawker load presented to the amp that means that the impedance reflected back to the amp at the autoformer is about 40 ohms.

So the AA presents a 40 ohm squawker impedance to the amp, the woofer is about 5-8 ohms over the low-pass and the tweeter is 8 ohms. From memory the

minimum impedance in the Klipschorn with the AA is

~4.5 ohms near 50Hz. What are you getting for minimum impedance values in the Klipschorn?

[Al Klappenberger]

John,

If I read you correctly, you'r right. The AA network taps the K55 down on tap #4. That's a turns ratio of .707:1. Square that and invert it, you get an impedance ratio of 2.0:1. The 13 Ohm K55 looks like 26 Ohms to the filter driving it (a 13 uF cap). That should be the maximum value. The woofer goes down to 4 Ohms. That should be the spread. I have not measure the input Zo of my netwrok in a speaker, only on resistive loads.

Al K.

================

[John Warren]

the turns ratio of 2:1 is correct (thanks), the reflected impedance from the autoformer on the AA is 32 ohms from the literature (K55M is 16 ohms per Klipsch).