Phil Allison wrote:

"Arny Krueger"

Andre Jute wrote:

The indispensible Phil Allison tells us:

*******
The Quad ESL57 input transformer has an overall turns ratio of 292:1.

The bias voltage on the bass panels is 6.2kV and they require 18 kV
volts p-p of audio to reach full output.

The bias voltage on the mid/high panel is 1.5 kV and requires 6 kV p-p
of audio to reach full output.

The Quad ESL63 has a bias voltage of 5.2 kV on the panels and requires
16 kV p-p of audio to reach full output.

I doubt that Phil ever told such a whopper.
The audio voltage never exceeds the diaphragm voltage.

Exception:

http://sound.westhost.com/project105.htm

polarization is speced at 1.5 KV, and maximum drive is 2.5 KV RMS



** Actually, the page says 3kV rms max. This is due soley to the cheap
transformer used - apparaently it's a 100 volt line tranny for PA systems.



The turns ratio of ESL speakers is usually in the 80-100:1 range. The
maximum RMS input of the ESL-63 is given by some as 10 volts RMS. This
would lead to no more than about 1 KV.



** More complete ******** from Arny.

Here are the FACTS:

The "Instruction Book" says the ESL63 is designed to work with amplifiers
of 100 watts per channel at 8 ohms - eg the Quad 405 ( I have a copy in
front of me). Such an amp can deliver 28.3 volts rms or 40 volts peak
(undistorted) to the speaker.

The published specs for "Maximum Input" a

" Continuous input ( sine wave) voltage 10 V rms

Programme peak for undistorted output 40 V

Permitted peak input 55 V "


The ESL 63 has TWO input transformers with primaries in parallel and
secondaries in series with the centre connected to speaker ground. Each
transformer has a turns ratio of 94:1 - ie when driven with 1 volt rms
the output is 94 volts rms.

This makes the overall ratio 188 : 1 - see schematic:

http://www.quadesl.com/schematics/quad63_schematic.gif


Now, 28.3 times 188 = 5,320 volts rms

(= the max drive voltage to all sections of the speaker)

5320 volts rms = 15 kV p-p.

QED.

So Andre was repeating Phil's error without correcting it.

In Phil's descrption of the ESL63 above, one transformer drives each plate.
So with "Programme peak for undistorted output 40 V", which is 28.3v RMS,
each plate gets 2657v RMS or 3760V peak, relative to ground, which is less than
the diaphragm's HT voltage of 5250v. QED again!

As the two plates are in antiphase, you could claim a peak-to-peak
plate-to-plate voltage of 7520v but there is no excuse for doubling it again.

And while looking at quad63_schematic.gif, check out the zener voltage D40-D71.

--
Eiron

No good deed ever goes unpunished.