It's really not watts or amps but volts. ESLs are mostly voltage driven devices. However, there is a reactive impedance due to the fact that they are really just capacitors (impedance drops with increasing frequency) that are driven by transformers (voltage step up devices).
Since there is a characteristsic impedance, but one that is reactive - it turns out that in solid state amps it is a good idea to have a large SOA (safe operating area) and a stable design so that the amp doesn't barf on the wierd phase angles and implied current/voltage lead/lag.
In order to move the diaphragm almost only voltage is required - however in practice there is some current wasted in the process. This is why tube amps can be direct coupled to the stators in some designs, even though the plates are actually very high impedance. Current not really required.
For various reasons, imho, there are few solid state amps that shine on ESL loads, while tubes of sufficient power (read power = voltage swing) can often give superior results in the midbass...
There are only a few different variants on the drive transformer circuits used in ESLs. Notoriously bad are the original ML CLS with a nasty dip down into the 2 ohm region, making them a tough to drive speaker in general. Depending upon what the manufacturer of your speaker did in the drive circuit (and there are only a few choices for full range cells - full range cells being like CLS, Acoustat, not Quad 57)you could have a nasty, low Z load to drive. That makes it very hard for all but the very biggest amps to handle, and that narrows the field so much that you may have a hard time finding one that also sounds good.
If you want to do some research look into the patented drive circuit for the Acoustats - it solved this problem very nicely. One can use it for personal use without violating the patents, btw...
For some ESLs it can be a problem to find an amp that will drive the speaker and sound good doing it for the above reasons.
Since there is a characteristsic impedance, but one that is reactive - it turns out that in solid state amps it is a good idea to have a large SOA (safe operating area) and a stable design so that the amp doesn't barf on the wierd phase angles and implied current/voltage lead/lag.
In order to move the diaphragm almost only voltage is required - however in practice there is some current wasted in the process. This is why tube amps can be direct coupled to the stators in some designs, even though the plates are actually very high impedance. Current not really required.
For various reasons, imho, there are few solid state amps that shine on ESL loads, while tubes of sufficient power (read power = voltage swing) can often give superior results in the midbass...
There are only a few different variants on the drive transformer circuits used in ESLs. Notoriously bad are the original ML CLS with a nasty dip down into the 2 ohm region, making them a tough to drive speaker in general. Depending upon what the manufacturer of your speaker did in the drive circuit (and there are only a few choices for full range cells - full range cells being like CLS, Acoustat, not Quad 57)you could have a nasty, low Z load to drive. That makes it very hard for all but the very biggest amps to handle, and that narrows the field so much that you may have a hard time finding one that also sounds good.
If you want to do some research look into the patented drive circuit for the Acoustats - it solved this problem very nicely. One can use it for personal use without violating the patents, btw...
For some ESLs it can be a problem to find an amp that will drive the speaker and sound good doing it for the above reasons.