Hi Andy,
This is usually easily implemented merely by introducing an appropriately sized cap in the "hot" signal wire running from preamp to power amp.
Hi Jens,
Yes, a simple series cap in the 'hot' signal wire into the LF55 will give you a 6dB electrical HP slope ... which is theoretically the best in terms of phase screwups.
However, you said you were matching this with the acoustic rolloff of the driver (sorry, I forget whether it was tweeter or mid-range!
) Have you allowed for the fact that the LF end of a driver's acoustic response typically includes a 4-8dB hump at the driver resonance point ... then followed by a dropoff at lower frequencies. Are you calculating the uF required to give you the -3dB "filter point" of the electrical slope at the resonance frequency ... or above it ... or below it?
You seem to be attempting to add the 6dB of the cap rolloff to the natural acoustic rolloff of the driver but if the value of the cap is too large - ie. giving an "elbow frequency" at or below the driver resonant frequency - yes, it will add to the slope of the driver's acoustic rolloff but you will be left with the hump around the resonant frequency.
Conversely, if you use a smaller cap so that the filter's elbow frequency starts above the driver's resonant frequency, the hump will be tamed but you will no longer be taking much advantage of the driver's inherent acoustic LF rolloff.
So you probably need to implement a higher order filter.
However, when the preamp is a tube amp, this is slightly more complicated, as tube amps usually have an output cap. If this output cap "sees" the roll-of cap (because they are connected in series), the resulting capacitance will be not be correct.
The "trick" described in my previous post should be able to prevent the output cap from seeing the roll-off cap (at least so I'm told), so that you can control the roll-off merely by calculating the cap value in relation to the power amp input impedance.
AIUI, ss amps as well as tube amps typically have an output cap ... only those amps proudly proclaimed as "direct coupled" do away with this.
So maybe your friends "trick" will do the trick (I'm afraid I have no idea but the logic sounds good to moi!
)
Some time in the future I intend to go "fully active". I just need to find active filters of sufficiently high quality that will not cost a fortune. A friend of mine is currently in the process of implementing some new, digital crossovers that are supposed to be very, very good whilst still being at reasonable prices. I'm looking forward to his results 
"Active filters of a sufficiently high quality"!!
IMO there are 4 kinds of high-class active XOs ... and I use #1 simply bcoz that's the one I started with about 8 years ago and I can't afford a DEQX!!
1. Rod Elliott's active XO (
www.sound.au.com). Rod supplies the PCBs and instructions for these nominally 12dB or 24dB L-R XOs ... but you can tweak the circuit to give you 6dB and 18dB as well!
These are opamp-based ... so some purists will automatically dismiss them! (However, I think Hugh will agree mine don't seem to have any obvious failings!
)
2. Transistor-based active XOs. The only ones I know of are those produced by Naim ... the circuits are available on the Net but I have no idea how to tweak them for different filter frequencies or slopes. In contrast, there is a heap of information on the Web about opamp-based active XOs.
Possibly transistors will sound better than opamps?
3. Marchand active XOs - both opamp and tube-based. These are well-regarded and come in either kit form or fully made up.
4. The Oz-designed DEQX digital XO. In fact this is more than just a set of active filters ... it gives you room-correction DSP and driver phase correction too! A marvel!!
Regards,
Andy
