Gain matching of power amps

How important is it to have amps of same gain when using amps from different manufacturers in a horizontal biamp? Currently using Audiolab 8000P's(29db gain) in vertical biamp configuration, but need to tame metal tweeter top end. Looking at McCormack(40db gain), and Classe power amps(anyone know the gain of these) as a solution. Do you have recommendation for other power amps that have a warm presentation that might suit? Does it matter if amps have different power ratings? Audiolab's are 100W, would you recommend more, less or same power for tweeter?
I see no point in matching a gain. How can you controll it? Even if you find matched gain amps most-likely they will have different sencitivity. The input sensitivity of amps of different brands is different by non-negatable values by tenth fractions of volts or even by whole volts.
On the other hand you can use SET amp to drive your tweeter/midrange and 200W SS amp to drive a woofer if you have an active filter or crossover.
If you don't have an active crossover you might consider using two of the same brand matched amps for tweeter and woofer.
Your closer bet between the McCormack and Classe is the Classe amps which all have an output gain of 29.02db
Matching the gain is very important. It's true you can't control it unless the amp has a volume control. It is function of the amplifier design. But different gains will result in different powers being delivered to different drivers with the same input. Not good.

Woofers require more power than tweeters so you can use a lower power amp on the tweeters as long as the gains match. Most people would recommend using identical amps or at least from the same family. Say a Levinson 331 on the tweeter and a 333 on the woofer.

You can skip the rest unless you want the math behind it.

Gain is the ratio of input to output voltages. An amp with 29 dB of gain amplifies the voltage 28 times ( 1 volt in results in 28 volts out). An amp with 40 dB of gain amplifies the voltage 100 times. Assuming an 8 ohm load, the 29 dB amp with a .25 volt input would deliver about 6 watts. The 40 dB amp with the same input would deliver about 80 watts. This is assuming the amp is capable of producing that much power. If not, it will clip.

Using these two amps to biamp would result in 13 times more power being delivered to the speaker hooked up to the 40 dB amp. Clearly, an undesireable situation.

Sensitivity is how much voltage it takes to drive the amp to maximum output. It is really just another way of looking at the gain. It can be calculated if you know the voltage gain and the maximum power rating of the amp. If the two amps have equal power ratings, the amp with higher gain will clip first and is said to be more sensitive. So it really doesn't matter unless you drive the amp into clipping.

voltage gain = output voltage / input voltage
dB voltage gain = 20 log voltage gain
power = voltage squared / resistance

To get from dB gain to voltage gain, divide the dB by 20 then take the inverse log of the result (raise 10 to that power).

29 dB / 20 = 1.45
10 raised to the power 1.45 = 28.2
40 dB / 20 = 2
10 raised to the power 2 = 100

Sorry, that may be more than you were asking but I have too much time on my hands and I wanted to given a reason why the gain was critical.
I agree with Herman. Having worked with many multi-amp set-ups, you have to do the best that you can possibly do to gain match the amps. If you don't, you'll run into some very "strange" problems in tonal balance and dynamic range that could vary frequency range to frequency range. Sean
I've said it before, that Sean seems like a very reasonable and wise person.
Herman, I bet your opinion of me is bolstered when i agree with you : ) I bet that there are a lot of others that wish i would shut up :(

Honestly though, it's good that you and a few others take the time to present "details" and "formulas". It gives many of the others a good idea of what it takes to make this stuff tick and how engineers do their work. As mentioned, it is science that has gotten us this far in audio reproduction and forgetting that might be a grave mistake. Sean
Kudos to you guys, you answered a question that I had in a different post. But, let me see if I've got this right.

I've got a Casablanca preamp, which is basically a big Theta card cage that you can slap different D/A output cards in and configure each output the way you want. I think (I don't have my manual here) that you can pre-set a gain for each output, so output set A could be configured with, say, 3 dB of additional gain over output set B, at 0 dB gain. If I then had an amp X with a gain of 26 dB and an amp Y with a gain of 29 dB, would I then achieve a uniform gain of 29 dB by hooking amp Y to output B and amp X to output A?

The reason I ask is because I always assumed that there was some nonlinearity in the gain as you increased the input voltage. I.e., that the gain you find in the specs is some average and is not a constant over the range of input voltages. Is this a dumb question?
Gain is gain, it is constant no matter what the input voltage (until clipping, that is). So yes, in theory your hookup would work. The only caution has to do with tolerances, i.e., is the gain exactly 29.0 dB or is it actually 28.2 dB? Same goes for the other amp. You would need a high quality calibrated voltmeter to be certain.
Edesilva, you are correct. Gain in dB is additive so your math is correct. You are also correct about amplifiers being nonlinear. However, a well designed amplifier will keep this non linearity to a small fraction of a dB and shouldn't be of concern.
Miketw and Edesilva: Neither of you mention whether you have considered using an active crossover, which I consider the best way to do biamping. The individual speaker drivers should not really be receiving the full frequency range from the respective amps driving them. The overall goal of biamping is to split up the signal after the source/preamp and send a limited range of frequencies to the appropriate amp and then driver.... bass signal (say <200Hz) to the bass amp and then directly to the woofer, etc. Another advantage of the active xover unit is that most allow volume adjustments, so that the signal level going to each amp can be can equalized. This helps some with the gain issue, but as Herman has nicely outlined, you still don't want amps of wildly different gain involved. By the way, I'm not convinced that it always works to have SS for bass and tubes for mid/highs, since some speakers reveal the sonic differences between such amps at the crossover point. Some experimentation is obviously required for each case.
Thanks for the info, I'm still thinking about what to do when I move in 1.5 mo. and can unpack some gear in boxes, and its food for thought. Martin Logan actually recommends bi-amping my Prodigys with tubes up top and SS below. I was going to use a pair of ARC VT100s I have, but now am thinking of watching for an old ARC D400 and using the extra VT100 elsewhere.

In my case, Ral, the Casablanca can be turned into an active x-over. All things being equal, however, I'm not inclined to go active b/c I'm not sure I can replicate the job the ML x-overs are doing and the marginal volume doesn't really mean that much to me.

I always thought the "purists" suggested avoiding an active x-over anyway to eliminate "yet another thing in the signal path"... (Obviously, doesn't apply in my case, 'cos the Casablanca is there anyway).
Edesilva: I didn't know that about the Casablanca... interesting. I see your point about not trying to recreate the ML crossover. And yes I think many purists want to avoid any signal processing between source and amp, and this is quite reasonable. However, in a few cases a good active xover can actually improve things if the speaker's passive xover system is not so great (that was my case with the Maggie 3.6R factory bass-to-mid xover box). A Marchand xover (deluxe version with upgraded parts) has worked well for me and introduces negligible sonic signature given the overall improvement from biamping. I've read that the Bryston xover is pretty neutral too. Probably there are other good ones.
Many thanks for the responses. Especially Herman, level of detail I wanted, and much appreciated. I have two matching power amps right now, so I know that this is the best gain match solution. May need to look at other alternatives to resolve tweeter problem. Currently using runs of Stealth Premier for both tweeter, and woofer. Maybe should look for a softer/darker cable for tweeter, recommendations?.
What is your view of going to a more powerful amp, and NOT bi-amping at all. Candidates, Classe 150/200, McCormack DNA 1/2, Audio Research D240/400?. Other recommendations? Would probably still want to biwire.
Sorry to take a step backwards here, but the difference between using a good quality active crossover and bi / tri / quad amping and passively multi-amping with passive crossovers is STAGGERING to say the least. More "junk" in the signal path ( active crossover ) will almost always produce results that you've never dreamed of if you have the time and know-how to dial it all in. However, you do have to take the passive crossovers out of the speaker circuit, resulting in "direct drive" from amp to speaker. MUCH, MUCH better. Sean