Let's imagine some crazy newb at building stuff was looking at a 3-way speaker with a passive crossover and biwire connectors and thought to himself:
"You know, if I disconnect the high pass and low pass filters coming out of the wire-connection points: I could use an external crossover and bi-amp properly".
How do-able does that sound.
And if he also happened to think:
"You know: part of this is a damper to make sure the HF and LF have the same SPL: I'm just wasting wattage. I should bypass that too and adjust the gain on the amp powering that portion"
Would you get out your straight jacket then?
For that matter: if one has the schematics of a passive crossover; how difficult is it to just program that information into a DCX and replace the passive crossover entirely (obviously the wiring of that part is simple, but does the passive crossover diagram / stats give enough to just program the active crossover to emulate it?).
I assume you are talking about your B & Ws. If so, then I think this is a worthwhile project. I have never used the Behringer unit and can not vouch for acceptable quality, but active crossovers are not that difficult to design and build.
I have long thought that triamping the B & W is a worthwhile project. I really feel, excellent though the passive crossovers are, they are the biggest limitation. I have a feeling that properly done a project like this would really let the cork out of the bottle so to speak. I think B & W are remiss in not offering active versions of their 800 series. ATC and PMC a beating them up badly in the professional arena because of it.
So how to start.
1). Order the woofer tester from parts express and get the T/S parameters of woofer and mid and the impedance curves. I would not do it to the tweeters, you might damage them.
2). Carefully copy the existing crossover circuit.
3). If still available see if you can get the B &W white papers on those drivers especially the mid range. I have not checked the site for those white papers in a while.
Now it is a lot easier to do what you suggest on a good speaker than a bad one. I don't believe those drivers have any nasty out of band peaks that need notching.
Now the published crossover slopes are combined driver acoustic and electrical slopes. So when we have your data we need to reverse model the passive crossover.
Then we can work out the electrical slopes and the diffraction compensation.
Now when you go active, the drivers impedance curves do not need correction unless you plan on using class D amps. I would not recommend them at this stage of the project. Level matching is done ahead of amplification always, so as you point out you don't waste power. A crossover in a three way speaker will waste between 1/2 to 2/3 of the amp power.
In order to avoid damage to the tweeter, I strongly recommend putting a 50 mfd cap in series with the tweeter.
Active filters are built from cascaded first order filters. You use op amps and the turnover frequencies are set by the RC network you place in the gain loop of the op amp chips. You never need inductors in active crossovers, so you don't have evil ringing and resonance problems For insurance reasons I recommend you design the circuit so it can be powered from Walworts.
Jung has an excellent little book that gives you the down on dirty on how to design using op amp chips.
If you go carefully, think things through, then this a very doable project. I strongly suspect if you do it well the improvement will be truly astonishing.
Since you are an engineer and have an interest, I'm going to encourage you to embark on it, where with a lot of posts I would not.
