Quad ESL 57 speakers impedance?

[gogofast]

anyone know what impedance speaker tabs i should use for ESL 57's? i'm going to completely rebuild the speakers, but before i do it, i wanted to get familiar with their before condition. while i have read pretty much most publications about these speakers out there, no mention of impedance. maybe i'm missing some basic facts here.

the reason i ask is when i connected it to a cheap ss receiver, it sounded quite nice but when rebuilt eico hf-81 was connected, it distorted badly. i used 8 ohm speaker terminal.

thanks for any input from ESL owners.:thmbsp:

 

[mhardy6647]

16 ohms. It is a funny load, though; impedance is very low at high frequencies. You risked damage to both speaker and receiver by what you tried; stick with tube amplifiers or get ESL-63s which are better-protected.

The ESL-57s are extremely insensitive; they will not play loud with an EL-84 push-pull amplifier... but they will sound great within their limits.
Mine are usually used with an HF-81 or a Dynaco SCA-35. This past weekend I used them, pretty successfully, with a cute little pair of Radio Shack 6V6 push-pull mono integrated amps (10 watts per).

 

[bricktop]

They are nice speakers, had a chance to listen to a refurbished pair once (panels replaced). Quad did make a nice tube amp to pair with the ESLs, if you can get a pair at a reasonable price.

 

[ARRAY]

....................

 

[frperdurabo]

The 57s impedance is all over the place - really, really high in the bass, dropping to .5 ohm in the treble.

The low impedance in the trebles means you want to use the 4 ohm tap. Otherwise, the treble will be rolled off, and the bass will be too ripe.

I'm sure I've seen a scan of an ESL sales brochure which claimed the 57 had an "effective efficiency" of 92db/1 watt. (I'm searching for the link to this...I'll post when I find it again)

15-20 watts is plenty, but SET amps won't work because their output impedance is too high; I've tried. But my AES Superamp gave nice results.

 

[HiFiCanada]

The 57s impedance is all over the place - really, really high in the bass, dropping to .5 ohm in the treble.

The low impedance in the trebles means you want to use the 4 ohm tap. Otherwise, the treble will be rolled off, and the bass will be too ripe.

I'm sure I've seen a scan of an ESL sales brochure which claimed the 57 had an "effective efficiency" of 92db/1 watt. (I'm searching for the link to this...I'll post when I find it again)

15-20 watts is plenty, but SET amps won't work because their output impedance is too high; I've tried. But my AES Superamp gave nice results.

I thought it would be the other way around, impedance drops in the bass and goes up very high in the treble.

 

[frperdurabo]

I thought it would be the other way around, impedance drops in the bass and goes up very high in the treble.

I still can't find the link I'm thinking of, but this Quad brochure states the the impedance falls with rising frequency:

http://www.dc-daylight.ltd.uk/Valve-Audio-Interest/Manufacturers-Averts/Quad-ESL57-1978.pdf

This database isn't the link I was thinking of either, but claims an efficiency of 93db/watt:

http://www.radiomuseum.org/r/acoustical_esl_electrostatic_loudsp.html

Art Dudley's article in Listener magazine is a good overview.

 

[gogofast]

The 57s impedance is all over the place - really, really high in the bass, dropping to .5 ohm in the treble.

The low impedance in the trebles means you want to use the 4 ohm tap. Otherwise, the treble will be rolled off, and the bass will be too ripe.

wow, .5 ohm?! how does any amp handle that kind of fluctuation? is biamping the best or only way to go then?

is this why some claim or complain about high roll off? has anyone tried biamping with different impedance? will that work?

16 ohms. It is a funny load, though; impedance is very low at high frequencies. You risked damage to both speaker and receiver by what you tried; stick with tube amplifiers or get ESL-63s which are better-protected.

The ESL-57s are extremely insensitive; they will not play loud with an EL-84 push-pull amplifier... but they will sound great within their limits.

so, 16 ohms? i'll try that later. i've heard my friend's pair with EL84 push pull and they sounded better than any speakers i've heard. i don't usually listen at that high level so moderate volume is perfectly ok with me.

 

[Vint Age]

Great speakers! I enjoyed mine for about 12 years until the SPL on the bass panels dropped to where they were barely audible. Now they're waiting for a rebuild that I never seem to have time for. :tears:

 

[gogofast]

thanks for the graph!

anyone biamping?

 

[gogofast]

can you believe this? i thought i lived in a pretty big city...seattle. i've been to 4 places that sell electronic parts and none of them had 1n4007 diodes. not one of them....

i was planning on rebuilding the EHT block today....but i guess i'll have to order those diodes in and wait.

 

[Nat]

I believe dynamic speakers are inductive (and resistive) loads, whereas ESLs are capacitative loads. Hence the low impedance at high, not low frequencies, and hence the variability of different amp's ability to deal with them.

 

[frperdurabo]

thanks for the graph!

anyone biamping?

There's no electrical crossover, AFAIK. You'd have to open them up and attach the central - treble - panel to a new set of binding posts.

I'm not sure what the payoff is...the speakers don't need more than 20-30 clean watts, probably less. Maybe use an OTL, with its high output impedance, on the bass panels, and something else on the treble?

I'm not sure, but I think the xover region is around 800 hz. Does anyone have back issues of Listener magazine handy?

 

[gogofast]

There's no electrical crossover, AFAIK. You'd have to open them up and attach the central - treble - panel to a new set of binding posts.

I'm not sure what the payoff is...the speakers don't need more than 20-30 clean watts, probably less. Maybe use an OTL, with its high output impedance, on the bass panels, and something else on the treble?

I'm not sure, but I think the xover region is around 800 hz. Does anyone have back issues of Listener magazine handy?

reason i asked about biamping is that with such varying impedance across the frequency, i'd think biamping with different impedance - for example, 4 ohm amp for tweeter panel and 16 or 32 ohm for bass panels. that might accommodate both tweeter and bass panels to perform at optimal level. i''ve done biamping before, but am not really sure if different impedance will work or not.

 

[Vint Age]

reason i asked about biamping is that

You're going way too fast.

Note that the ESL57 is transformer coupled at the input. You'll need some serious volts to drive the panels directly. :yes:

 

[dnewma04]

I find it intriguing that most tube amps (and their fans) really want a fairly flat impedance curve because they don't handle the wildly varying impedance situation all that well. They also don't handle capacitive loads all that well, but almost everyone recommends a tube amp because they are "better" at handling the wildly varying impedance and capacitive load.

 

[Nat]

Whoever the original poster was: I hope you know of Sheldon Stoke's extensive description of the Quad and rebuilding it. It's on his SDS website, and might be helpful.

 

[gogofast]

yes, i have a paper copy that i'm studying every night.

after the power supply rebuild, they are totally different speakers. now i'm noticing bass panel rattle - which i didn't notice even at high volume before. i guess at least the dust cover refurb is needed.

I find it intriguing that most tube amps (and their fans) really want a fairly flat impedance curve because they don't handle that situation all that well. They also don't handle capacitive loads all that well, but almost everyone recommends a tube amp because they are "better" at handling the wildly varying impedance and capacitive load.

i guess good p/p tube amps handle the varying impedance better than most cheap SS amps. looking at the recommended amp list for ESL speakers, there are many SS amps.

 

[Nikko75]

I have measurements put away but these illustrate the nature of the impedance magnitude basically the same. The Quads ESLs' impedance drops at low frequencies to about 4 Ohms, exhibits a low "impedance Q" and impedance magnitude of 32 Ohms in the bass region, then within a range around 15kHz drops to 2 ohms, and rises again as frequency increases. The phase isn't shown here, but its inductive in the lower half like a conventional woofer, swings back and forth about +/- 20 degrees and turns quite capacitive at 15kHz.

For reference the ESL-63 is a bit different. It hangs around 3 Ohms at the lowest frequencies, has a slightly higher impedance Q and comparable magnitude of 35 Ohms in the bass, drops to 4-5 Ohms in the treble region then rises steeply then flattens out again. It's impedance phase is usually inductive to 2kHz, swings back a forth a small bit, then takes a narrow inductive then capacitive swing in the 10-25kHz region. It's not that extreme as some literature or sites may suggest.

I thought I might as well add this as it adds common ground in understanding what the graphs mean in comparison to other speakers. Below is a good example of a normal moving coil speaker with larger capacitive swings than the ESLs. The Totem Forest has three notable dips, one of which is in excess of 50 degrees capacitive at 85Hz and a wide capacitive swing at 4kHz.
The Quad ESL 989's impedance, pictured at bottom, is inductive through the bass as are all these Quad ESLs and that may come as a surprise. From 100Hz to 10kHz the ESL 989 is at most only 23 degrees capacitive within the audio band. From 14khz to 25khz it exhibits an inductive reactance and impedance resonance peak followed by a capacitive dip at 30kHz.