High Frequency Driver Extension

[Art J.]

Hello Richard, Yes I would like to see your model of the 30923.
My main concern with it, was the phase shift by adding passive
components to an active system. Not the best of ideas but I
still like the results. Will your model show phase shift?
When I did a sweep, it showed a gradual phase shift from
about 600c, peaking 30 degrees at 6k, then returning to unity at 20k.
This caught me by surprise at first but it must just be acting as a series
LC circuit with the inductance of the horn coil as opposed to an 8 ohm resistor load.
30 degrees max is not as bad as I expected. The average is much less.
I didn't spend the time with it that I would have liked to, I had to get the lab rack
out of the living room for the holidays. Ill play with it more next year but my main
concern will be getting the 'white noise' out of the active crossover. You can barely
hear it but it looks real bad on the scope and probably effects all my readings.

----------------------------------------------------------------------------
"Doctor, Doctor!, My horn rings like a bell
when I hit it like this!"
Doctor : "Well then, Don't hit it like that"

[endeeinn]

Does anyone have the formulas so I can calculate the values for 16ohm drivers?

[Art J.]

For going from 8 ohm to 16 you use half the value
of capacitance or 6uf to 3uf and then double the
value of the resistors

[Richard C.]

Art,

If you are concerned with the phase shift, you could use the "CD horn freq. correction" that some electronic crossovers have. My Behringer CX3400 has this option (simple switch). You would get the HF emphasis without the series cap in the circuit. You do need a phase shift at crossover to compensate for the difference in voice coil position - time alignment - phase alignment, etc.

Tell me exactly how you are using this 30923 circuit. Are you using any part of the LF circuitry - L1, C1, zobel? Are you connecting to C1 or C2 for the HF input?

Richard C.

[Art J.]

Richard, Im using a Rane AC-22 active bi-amped. I have the driver offset phase
correction set properly at the 500 cycle point. That is fine and no problem at all.
Infact it is really nice. The bummer is that as you increase frequency from that point,
the phase will start to roll a bit due to the 30923. That's all. Nothing I can do about it.
It wouldn't roll if it was not in the line. I'm probably over concerned about something
that cant be heard anyway. I do have issues with the Rane. I mentioned it was noisy
but their is more to it. When I first got it, I temporarily hooked it up to my main system
in another room, and it sounded pretty dull.
So I'm sure it's the weak link now in system 2.
I'm keeping an eye on some active crossover projects and a change
will take place sooner or later.


------------------------------------
How long is a short circuit?

[Richard C.]

Art,

Here are the images from my software modeling program of your 30923 attenuator. The maximum phase shift is 38 degrees at about 4kHz.






Richard C.

[Richard C.]

Well, it looks like my pictures were temporary. They were there last night. I need to find out how to upload the .jpg to this hostboard server, and NOT just the link.

Richard C.

[Art J.]

Hello Richard. Did you use UPLOAD NET?
http://www.uploadimages.net/index.php
When I put a picture in, I open it with
the last link option in a new window, then
save it to bookmarks. It seems to stay that way.

I will check my actual 'Le' reading of the
horn with my LCR meter. This may account for
our different results. Be back!

[Richard C.]

Art,
I did use uploadimages, but I didn't save it to bookmarks. I now seem to be getting the .jpg's, today.

I am using a B&K 875B Digital LCR meter, which is using 1kHz as the test freq. That signal is in the bandwidth of the driver, so it is difficult to get an exact reading. Even room reflections will change it, so I load it to the max by placing the horn down against a flat surface. Someone on this forum said that the factory places the voice coil in a test fixture to measure it. I emailed an engineer with one of the modern loudspeaker companies (forgot which one), and he told me they DO use a fixture to measure Le. This .15 mH reading is an average from 4 of my 902 drivers. I have done so much diaphram swapping, that I will have to look at one of my Model 14's to be sure. I will try to measure an 802-8G with a 34647 (I have two). My Model 19 - 802-8G's came with 23744 screw type diaphrams.

Richard C.

[GM]

FWIW I prefer to calc inductance at a frequency of my choosing, so I measure Zxo at the desired frequency (Fxo) and Re, then:

Le (mH) = [SQRT(Zxo^2-Re^2)/(2*pi*Fxo)]*1000

GM