Passive Bi-amping using XLR Y-adapters


I want to try passive bi-amping using two identical amps.My preamp has two pairs of RCA outputs, but only one pair of balanced outputs.
My system sounds much better and dynamic using the balanced outputs instead of the RCA outputs, therefore I would like to use the XLR outputs to BOTH of the power amps when I purchase the second amp.I would like to use XLR Y-adapters for this set-up.
Would the best sonic results be achieved using XLR Y-adapters at the preamp using one male plug, two sets of interconnects with female plugs on the amp end of each interconnect per amp, per channel, or, using one interconnect per channel from the preamp and then split their outputs at the amps?
Would doing either of the above change any of the specs and perimeters that the amplifiers would "see" from the preamp and/or interconnects as opposed to a single interconnect per channel using only one amp, such as interconnect resistance, capacitance, inductance, or preamp output impedance, voltage gain, etc?
Thanks in advance.
daltonlanny
I too am considering this question. I've been told that preamps are generally wired inside by splitting the outputs, not by laying out separate circuitry for each output. If that is the case, then a Y-adaptor should be the way to go. If you talk to interconnect companies and salesman they all tell you to buy another pair of interconnects. In your case, with only one set of balanced outputs, I'd definintely use Y-adaptors.
Use Y-adaptors on the amp side (female to 2 males), not on the preamp side, to avoid a second set of interconnect.

Using two amps running full range into passive crossovers is NOT bi-amping. Sorry. Bi-amping requires bypassing the original passive crossovers and using an active crossover between the preamp and power amps matched to the design crossover frequencies and slopes as the original passive crossovers. True bia-amping provides multiple benefits - up to 6 dB improved dynamic range, amp-driver optimization, better driver control from removal of passive crossovers components. What you propose provides none of those benefits, but all the expense and additional complexity.