Novice attempting to bi-amp. Help

Choose wisely, Luke Skywalker! That's the best advice for bi-amping. I would NOT use Carver amps at all for anything except maybe Sub Bass. Why? Because they are not particularly low in distortion or pleasing in subjective quality to my ears...

Having said that, I agree with Sean about the relative merits of "vertical" vs. "horizontal" amp application.

You can check your crosstalk by putting a dummy load on one
channel of your amp, the speaker on the other, play the signal ONLY into the one channel with the dummy load at normal listening levels, and see what you hear (if anything) from the speaker. That's your crosstalk in practical terms.

It should be >30-40dB down worst case.

I'd try to find a very high quality amp for the top end... since that's whats going to effect the overall quality of the sound that you hear. The highs have an inordinate contribution to what you perceive out of a speaker...

If you bypass the crossover in the speaker, and go with an electronic crossover, you might be pleasantly surprised at how much better the ribbons can sound without the Carver
parts in there... Since the ribbons themselves without the xover are almost completely resistive, it is a very easy load to drive, so you only need a sweet sounding amp, not one that is both sweet and can drive wierd loads!

similarly, I'd swap out the caps if they are not already Polypropylenes in the HP section immediately regardless...

If you do decide to use the internal crossover, I'd set up a little box with a 1st order high pass (a cap and some resistors) to feed the HF amps (or do it internally) so that that section sees a 6dB/oct rolloff from about 1 octave below the Carver's crossover point. This will keep all the LF energy out of those amps. Automatically less intermodulation, no matter what.

Regardless, look for some nice sounding amps for the ribbon part of the speaker, imho.

Thanks to all for the in depth advice. I believe I understand the set up to do the horizontal bi amp,
how would the hook up for vertical bi amping differ as
this is the set up most use with the carver amazing speakers
according to my research on the Audioreview site?
Thanks again for all the imput.

Mar4004

Bear, good idea about the dummy load and listening for "bleed through". Simple and quite effective. Do you think that removing the interconnect from the input of the "dead" channel and using something along the lines of a Cardas cap be of any further assistance while doing this test ?

While i have never done any "controlled" testing and taken measurements, i would think that -30 to -40 db's "should" be pretty easy to achieve. What are some figures of amps that you would consider to offer good to excellent channel separation ?

I agree with getting the passive crossovers out of there. It is amazing how much even a single "high grade" cap in a simple 6 db crossover can destroy the sound of a speaker.

I also agree with your comments regarding the Carver amps. Those speakers are capable of performance well beyond those specific amps, especially if the passive crossovers are bypassed.

Mar, in vertical bi-amping, let's say that we have amp A and amp B. You would hook up amp A to one speaker and amp B would go to the other. Amp A's right channel would drive the woofers and the left channel would drive the ribbons. Amp B's right channel would drive the woofers on the other speaker with the left channel driving the ribbons.

Obviously, i selected left / right channels just for demonstration purposes as one channel should be the same as the other and they don't have preferences as to what they drive. Sean
>

Sean, the unused channel should probably be grounded, although that doesn't actually simulate operating conditions. So, I'd suggest alternately terminating the interconnect with another bit of local gear (like the output of a CD player not playing but on...

In lab tests the other channel is usually grounded.

Obviously the best separation is going to be a set of monoblock amps with separate supplies and chassis. That should be nearly infinite. Everything else is less.

I don't have any figures at hand for commercial amps, but they should be close to the noise floor unless there is a problem modulating the power supply at high loads (there can be). But this should not show up until you really suck some power. Or, on occasion there is radiation due to the current drawn through a conductor.

(you can make a nifty thing using a loop of wire as a transmitter, like around the room, and an inductor on the input of a preamp fed to an earphone - you can "broadcast"
directly on audio that way)

I would think that shunting the input should negate all but the highest levels of crosstalk. After all, the input is where the "leakage" would have the most chance to be picked up and amplified as that is where the circuitry would offer the most gain. Shorting the input jack to ground would in effect shunt the "leakage" to ground at the same time, making the test far less realistic in terms of real world operating conditions. That is why i thought a Cardas Cap might be worthwhile, as you would only hear the "internal leakage" within the amp itself with minimal influence from outside sources.

One other thing just came to mind. I would guess that someone with HIGHLY sensitive speakers might hear more "crosstalk" simply due to the fact that the speakers allow more to be heard with less amplitude signal sent to them. As such, someone with K-Horns, K.A.R's, etc... may think that their otherwise "excellent" amp is "leakier" than someone with an amp that is actually much poorer with a pair of 86 db speakers. I guess that is why it is "good" to have test equipment and dummy loads to document such things. You at least end up with consistent and repeatable conditions that can be used from component to component. This allows one to have a specific baseline to work from. Sean
>