This is the problem you are up against.
Most ESLs have an impedance that ranges from whatever peak they have in the bass to about 1/9th to 1/10th of that at 20KHz.
Now imagine a well built solid state amp that can double power as impedance is halved.
Now comes the tricky bit- ESLs don't follow the same rules that box speakers do. To start with, there's no box :) To further complicate things, their impedance curve is based on a capacitance (hence the falling impedance at higher frequencies). What isn't obvious is that in a box speaker, the expectation is that the impedance curve is also a map (to a certain degree) of its efficiency. For example, the woofer has a resonance in the box; this resonance is represented by an impedance peak. Into that peak the amp has to throttle back its power, which a solid state amp does because if it doubles power as impedance is halved, that means that it also halves its power as impedance is doubled. That works out nicely for flat frequency response when dealing with many box speakers.
But ESLs are different as I pointed out. Their impedance curve is not a map of efficiency; to reproduce a 100Hz tone at 90dB takes about the same amount of power as it does to reproduce a 10KHz tone at 90dB. You can see what will happen: an amplifier that can double power as impedance is halved will be too bright because the impedance is much lower at 10KHz as it is at 500Hz.
Martin-Logan has been sort of getting around this problem by making the impedances of their speakers very low; starting at 4 ohms in the bass and going down to a fraction of an ohm from there. Since many solid state amps (their intended market, since there are more solid state amps on the market than there are tube amps) can't continue doubling power into such low impedances, they sort of get away with it.
But with a Sound Lab, which is arguably the state of the art in ESLs, the 20KHz impedance is between 1.5 and 3 ohms depending on the position of the Brilliance control. A lot of solid state amps can drive impedances like that- but on the flip side, the Sound Lab is typically about 30 ohms in the bass, which means that solid state amps can't make power. So on a Sound Lab, a 200 watt tube amp can usually keep up easily with a 600 watt solid state amp on this account!
To further complicate matters, the human ear/brain system detects sound pressure by listening for higher ordered harmonics of any sound. For this reason the ear is keenly sensitive to higher ordered harmonics as it has to have a range of over 120dB. Both solid state and tube amps make distortion, but if you look at measurements you will see that the fundamental test frequency is usually only 60 to 100Hz, so we don't get to see what the amp is doing at 5-7KHz, which is where the ear is most sensitive (the Fletcher-Munson curve, if you're still reading this far). So the higher ordered harmonics aren't shown. But if you are wondering why tubes are still around its because they sound smoother, and the higher ordered harmonic structure is why (not bandwidth!).
The ear converts all forms of distortion into tonality. The higher orders are treated as brightness and harshness. If you've been paying attention, then you know where I'm going with this- in addition to a brightness caused by a frequency response error on the part of most solid state amps on an ESL, you also get higher ordered harmonic content which the ear will also treat as brightness. Its a double whammy.
This is why ESLs and tubes have had a very friendly relationship going back to the very first ESL. For more on this topic see
http://www.atma-sphere.com/Resources/Paradigms_in_Amplifier_Design.php
Most ESLs have an impedance that ranges from whatever peak they have in the bass to about 1/9th to 1/10th of that at 20KHz.
Now imagine a well built solid state amp that can double power as impedance is halved.
Now comes the tricky bit- ESLs don't follow the same rules that box speakers do. To start with, there's no box :) To further complicate things, their impedance curve is based on a capacitance (hence the falling impedance at higher frequencies). What isn't obvious is that in a box speaker, the expectation is that the impedance curve is also a map (to a certain degree) of its efficiency. For example, the woofer has a resonance in the box; this resonance is represented by an impedance peak. Into that peak the amp has to throttle back its power, which a solid state amp does because if it doubles power as impedance is halved, that means that it also halves its power as impedance is doubled. That works out nicely for flat frequency response when dealing with many box speakers.
But ESLs are different as I pointed out. Their impedance curve is not a map of efficiency; to reproduce a 100Hz tone at 90dB takes about the same amount of power as it does to reproduce a 10KHz tone at 90dB. You can see what will happen: an amplifier that can double power as impedance is halved will be too bright because the impedance is much lower at 10KHz as it is at 500Hz.
Martin-Logan has been sort of getting around this problem by making the impedances of their speakers very low; starting at 4 ohms in the bass and going down to a fraction of an ohm from there. Since many solid state amps (their intended market, since there are more solid state amps on the market than there are tube amps) can't continue doubling power into such low impedances, they sort of get away with it.
But with a Sound Lab, which is arguably the state of the art in ESLs, the 20KHz impedance is between 1.5 and 3 ohms depending on the position of the Brilliance control. A lot of solid state amps can drive impedances like that- but on the flip side, the Sound Lab is typically about 30 ohms in the bass, which means that solid state amps can't make power. So on a Sound Lab, a 200 watt tube amp can usually keep up easily with a 600 watt solid state amp on this account!
To further complicate matters, the human ear/brain system detects sound pressure by listening for higher ordered harmonics of any sound. For this reason the ear is keenly sensitive to higher ordered harmonics as it has to have a range of over 120dB. Both solid state and tube amps make distortion, but if you look at measurements you will see that the fundamental test frequency is usually only 60 to 100Hz, so we don't get to see what the amp is doing at 5-7KHz, which is where the ear is most sensitive (the Fletcher-Munson curve, if you're still reading this far). So the higher ordered harmonics aren't shown. But if you are wondering why tubes are still around its because they sound smoother, and the higher ordered harmonic structure is why (not bandwidth!).
The ear converts all forms of distortion into tonality. The higher orders are treated as brightness and harshness. If you've been paying attention, then you know where I'm going with this- in addition to a brightness caused by a frequency response error on the part of most solid state amps on an ESL, you also get higher ordered harmonic content which the ear will also treat as brightness. Its a double whammy.
This is why ESLs and tubes have had a very friendly relationship going back to the very first ESL. For more on this topic see
http://www.atma-sphere.com/Resources/Paradigms_in_Amplifier_Design.php