Design question on passive crossover networks

[bracurrie]

Because my speaker system had no crossover networks I have used active biamping to balance the gain as well as the crossover point and slopes. How do you attenuate a speaker component in a passive network? I understand transformer based attenuation, I think, but in my unskilled and untrained way I am unable to identify that part in photos of many networks I have seen.

I would love to try passive crossovers, now that I am happy with my active setup, but wonder how to go about getting it done.

[Deang]

You're running cloned Jubilee bottoms with the Faital HF200 drivers and Elliptrac 400 horns.

There are a lot different ways to go about this, but I do believe Bob has a passive network for this combination.

The part you are talking about is an autotransformer. It looks like an overgrown iron core inductor with tabs and wires hanging off of it.

You can also use a single resistor in series (between amp and network), or an L-pad.

[bracurrie]

Thanks for the answer, and yes I know about Bob's crossovers. But in a hurry i mangled the question. Given that I like my active setup, how can I get a network built the way my active is setup?

[WMcD]

Some of this depends on just how complicated your active filtering system is.

Active filters using op amps can create 2nd or 3rd or 4th order filters (or most anything) with small, inexpensive components setting the parameters. However, you rally have to design the circuits with preset goals in mind. Delay to the extent we need it (a thousand of a second or millisecond -- mS) is not really possible. Notches, broad or narrow can be done.

Active filters based on digital signal processing can do the same do some tricks which are more amazing. All you have to do is fool with software. Often the DSP unit has a measuring microphone to test the results. Delay of even 5 milliseconds (about a delay of 5 feet) is a click away. Notches broad or narrow are easy.

In both of these active techniques, the electrical impedance of the driver doesn't come into play. The active circuits drive a power amp which drives the speaker driver.

What you're asking is the extent to which what might be a complicated results achieved in the active can be translated to passive elements. It can be difficult and without some very expensive test equipment to get it right.

There is a blog which describes this process of tailoring the passive crossover for the Palladium, note it is in three parts.

http://community.klipsch.com/blogs/technical/archive/2009/06/24/the-art-of-palladium-crossover-part-i.aspx

You'll see that actual audio output has been measured and then reduced at various freqs to make it flat in the middle and rolled off at the edges. The blog gives the values of the passive components which are used. What is not discussed is the operation of the program (most probably LEAP) which calculates the values needed. It also has measured the electrical impedance of the drivers.

Also, look at the component count needed. These are expensive items.

So, if you have simple demands for tailoring, it probably can be done without the program. Otherwise, it is a big job, and requires many expensive passive components.

WMcD