Bi amping with Tube and Solid State

First you must know the cross-over frequency of your speaker (low end). On my Martin Logan SL3's, it is 250 Hz. Therefore, I would choose something less than this to make sure I preserve the integrity of the internal x-over circuit inside the speaker. So in my case, I could choose 180 Hz. You want a simple, single-pole low pass filter. Therefore, a single capacitor will suffice. You also need a resistor if the input port is not already using one. The capacitance is determined according the relationship between cutoff frequency, resistance and capacitance. You'll need to dig out your EE book or use a tech for your case. The cap will likely be something < 1 uF.

A key point is that this is a purely passive filter, below the x-over frequency. Therefore, there is no degradation to the signal whatsoever. Cap quality is less of an issue, use something "reasonable".

Murph...,

Your suggestions seem to be fine but you're still facing a "double-degradation" using passive crossovers on two sides amplifier and speaker-embedded.

A capacitor tended to bring more distortions than any active element so mainly a lot of engineers try to avoid their introduction unless for smoothing in PS or forward or reverse correction of RF in input and feedback circuitries. A further note that capacitor isn't an ideal and "strong" filter as you may expect and more or less will pass portions of "unwanted" amplitudes in such interaction.

The more complicated part that is not probably covered in EE books due to complexity of calculations and tremendous time to be spent for testing is integrating built-in speaker crossover or actually complexed speaker load with built-in crossover with filters tended to be used on amplifier side.

There is always signal degradation even if the filter is purely passive. Consequently, the quality of the caps and resistor will have an impact on the sound, since the signal traverses an equivalent C-R circuit.

In the context of cross-overs and bi- (or tri-) amplification, the word active simply refers to the fact that no energy from any one amplifier is used on more than one target driver. While this is a good thing (listen to a Meridian active speaker system for verification), it is still achieved via filtering. That is, the active cross over, which acts on the line level signal, uses capacitive components to filter the signal and thereby introduce its own contribution to "signal degradation" and distortion (however minimal).

My suggestion above on the HP filter is a means of reducing the energy outside the pass-band of the speaker's x-over circuit. I stand by the concept that the consequent performance improvement should outweigh any effect that an RC combination will have on signal frequency components above the corresponding x-over's cutoff frequency.

Notwithstanding this, I concur w/ Gmorris that using quality caps and resistors is prudent, at the very least. At the same time, man-o-man I wish this would all be easier... ;-)