Planar Speaker Asylum: Re: Quad 57's with SET amps by 13th Duke of Wymbourne

Here are some comments on the drive capability required from an amp to
drive ESL 57s from my experiments to enable me to design an amplifier
to drive my own pair. I presently use a pair of amplifiers from a
Meridian M3 active speaker - I bypass the active filters and use the
bass amp of each channel full range. The amps are, as far as I can tell,
class AB transistor designs using the Lin topology with feedback (don't
know how much). It works fine and sounds great but I want to try a zero
feedback, class A design (I am now refining the design in simulation).

Because I'm designing the amps to only drive one type of loudspeaker a
generic power rating is not necessary so I made measurements of the
voltage and current output needed to fully drive the Quads. I connected
a 1ohm resistor (10x 10ohms in parallel) in series with the positive
terminal and monitored the terminal voltage with an oscilloscope (not a
storage 'scope but by playing with the brightness and timebase I could
get a good idea of the maginitude of the peaks). Also, I could use two
channels to monitor the differential voltage across the 1Ohm resistor
giving a direct reading of the current flowing into the speaker.

The biggest peaks were found using that favourite dem. piece 'Private
Investigations' from the recent CD re-master of Dire Straits 'Love Over
Gold'. With the volume set about as high as I would ever want to listen
to this piece I recorded slightly more than 30V peak voltages and 6A
peak currents during the finale. The crescendo near the end of the first
movement of Solti's 1964 Mahler 1 (again on CD) registered near 30V
peaks but the current maximum was ~ 2.5A. Chamber music hardly
registered on these scales though, ironically, it sounded too loud when
turned up whereas Mahler and Dire Straits never sounded too loud.

A further experiment was to add a 130Hz 2nd order high pass filter on
the amplifier inputs to filter out the low bass. I want to try using a
Gradient-like dipole subwoofer with the Quads (probably something very
close to the Linkwitz Phoenix) and was curious to see if that would
relax the amplifier drive requirement for the same input volume level -
it did not. I still got > 30V and 6A peaks with Dire Straits which are
due to high(er) frequency transients and not low bass excursions.

In conclusion, the Quads are a harder load to drive than I'd imagined. I
had assumed that the current required would be low because of the
relatively high speaker impedance, except at high requency but here I
expected the signal energy to be low. That did not prove to be the case
and rather than facing a 16 Ohm load the peak demand is better
represented by something like 5 Ohms which places considerable
re-emphasis in the design of a class A amplifier!

Of course, that doesn't mean that output current limiting on transients
will necessarily degrade the sound - it may even give a more pleasing sound?
I'm sure you'll enjoy the Quads whatever you drive them with.

Regards
13DoW