Hey SY, will you be joining us in Berlin end of the year? I'm thinking about doing a triple or quad quads.
Yes I also dig the triple set. Nice custom frames. Is that PK in the front? Looks a bit like him..
They do go , dynamics are excellent ...
Yes, PK did it originally, followed by other mods ... In a really big room,( custom room) 18 -22 ft from panels, orchestral music is exceptional. I do agree with you about imagining and some smear in the details vs a single pr, but they are the only FR ESl's i have heard that can sound like being there for Orchestral music.
Size ,presence and dynamics are spot on ..... 🙂
I can believe that as when in series the bass section is above 24ohm . When in Parallel the high frequency impedance is closer to 1 ohm , the speakers are quite efficient in the bass/midbass region and a stacked pr barely requires 25 watts Nominal in a moderate size room with a listening distance in the 8 -12 ft range from the speaker ..
I would think the Bass better on the 16 ohm tap, but the middle and top better on the 8 ohm tap ...
This month's Stereophile reviews the Quad ESL-2805, the latest incarnation of the Quad ESLs. I was quite astonished to see the measurements. It does square waves. It had 0.1% distortion at 86 db, and it was pure second harmonic, with a hint of 8th. JA noted it is better than most tube amps. It also had a near perfect step response. I wonder why this technology does not receive the same kind of attention and research as the conventional moving cone drivers and compression drivers.
It also brings into focus the countless number of hours and money people have spent evaluating listener preference in relation to frequency response and directivity, when there is a transducer here that does not have a ruler flat frequency response, nor good directivity, but is highly preferred.
It also brings into focus the countless number of hours and money people have spent evaluating listener preference in relation to frequency response and directivity, when there is a transducer here that does not have a ruler flat frequency response, nor good directivity, but is highly preferred.
Last edited:
Ra
Actually there are a number of speaker manf. that do follow the same principals to get step response that comes real close to the Quads, Thiel, Vandersteen, and just about anyone that follows the 6db rule of crossover design and arrival time.
What I didn't expect from the 2805 was the rolled off top end in that review, the best examples of the 57 went further out
Regards
David
Actually there are a number of speaker manf. that do follow the same principals to get step response that comes real close to the Quads, Thiel, Vandersteen, and just about anyone that follows the 6db rule of crossover design and arrival time.
What I didn't expect from the 2805 was the rolled off top end in that review, the best examples of the 57 went further out
Regards
David
Last edited:
Consider how it's being measured , reflections will do that . I wish JA would be more consistent in his measuring technique, well considering the polemics involved , maybe not ..
How was it measured? Can't find the review online...
It's notoriously hard to measure large panel transducers without a reflection free chamber. Near field measurements don't tell the whole story and indoor far field measurements are polluted with reflections.
It's notoriously hard to measure large panel transducers without a reflection free chamber. Near field measurements don't tell the whole story and indoor far field measurements are polluted with reflections.
I don't know about stereophile in a long time, but they used to have real problems with microphone positioning. The original innersound speakers used to have seriously beamy panels, and it was obvious in the measurements that they were not really on axis.
My measurements of both the 57's and 63's show that the 63's are significantly more extended.
Using a quasi-enchoic response technique (with MLS pulses), you can remove the room effects and get very clean response above a few hundred hertz.
A big flat panel can be insanely beamy, here's a 20" x 48" panel measured over 5 degrees of polar response:
http://www.quadesl.com/speaker/diyesl/polar_fine.gif
Sheldon
My measurements of both the 57's and 63's show that the 63's are significantly more extended.
Using a quasi-enchoic response technique (with MLS pulses), you can remove the room effects and get very clean response above a few hundred hertz.
A big flat panel can be insanely beamy, here's a 20" x 48" panel measured over 5 degrees of polar response:
http://www.quadesl.com/speaker/diyesl/polar_fine.gif
Sheldon
That is one nasty beaming panel 😀 Looks like the graph has been smoothed, perhaps 1/6 octave?
Yes you can window out the reflections which gives you enough data to show something above a couple of hundred Herz. But at a large distance the window becomes smaller and smaller which results in rather poor frequency resolution.
Even a modest size ESL like the Quad 57 needs a mic distance of at least 2m for the different segments to integrate properly. Preferably even 4m as stated by Quad. At 1 meter microphone and speaker height, this gives a best case time window of 1.4ms max. Which gives a frequency resolution of 700Hz and all you get are wiggling smoothed curves that don't reveal much detail.
Hence the dilemma. Near field measurements are not representative for large panel transducers, far field measurements require a small time window with poor frequency resolution.
Something else, you mention that you get more extended high frequencies than published. This could of course be the case, but I wonder if your microphone is calibrated? The often used Behringer ECM8000 for example has a large and broad response peak between 10-20 kHz... just something that came to mind.
Yes you can window out the reflections which gives you enough data to show something above a couple of hundred Herz. But at a large distance the window becomes smaller and smaller which results in rather poor frequency resolution.
Even a modest size ESL like the Quad 57 needs a mic distance of at least 2m for the different segments to integrate properly. Preferably even 4m as stated by Quad. At 1 meter microphone and speaker height, this gives a best case time window of 1.4ms max. Which gives a frequency resolution of 700Hz and all you get are wiggling smoothed curves that don't reveal much detail.
Hence the dilemma. Near field measurements are not representative for large panel transducers, far field measurements require a small time window with poor frequency resolution.
Something else, you mention that you get more extended high frequencies than published. This could of course be the case, but I wonder if your microphone is calibrated? The often used Behringer ECM8000 for example has a large and broad response peak between 10-20 kHz... just something that came to mind.
A PK refurbed ESL57 got featured on Stereophile's website yesterday, they were covering their NY show, best sound was mentioned ....🙂
Nice, yes it should not be possible but those 50+ year old ESLs can still hold up their own against a lot of speakers. When in proper condition of course. Adding a $20.000 OTL tube amp does not hurt either 😀
Thanks for the pointer!
Nothing really today at any level or cost will beat the 57 fo that midrange detail and microdynamics
Regards
David
Regards
David
I guess I wasn't clear, I get more extended response I referred to was that the 63's have more extended response than the 57's in response to AVWERK's comment earlier in the thread.
I have a calibrated Behringer mic, as well as a panasonic electret, and for serious work, a B&K 1/2" lab mic.
Sheldon
Sheldon/Arend-jan - a slightly OT question for you both:
Which is the harder to rebuild panels on - the 63 or 57 - assuming say I'm using a kit from ER?
Fran
Which is the harder to rebuild panels on - the 63 or 57 - assuming say I'm using a kit from ER?
Fran
I find the 57's to be a little more tricky, often due to treble panel warping or deformation over years of storage. The 57 panels are not built with the same precision as the 63's. So getting matched response from panels is harder. Plus many 57 treble panels have burned areas that need repaired. And they all are older with more potential for neglect and issues. There is a ton of rivets to remove for each 57 panel, and replace with some other rivet-like fastener (I use nuts and bolts).
The 63 frames and stators are higher precision. The glue that holds the circuit board stators to the frames is the Achilles heal of the 63's and the crappy quad glue has paid for many vacations for me. There are two more panels on the 63's, but they are a lot easier to rebuild. The 63's are held together better, not relying on tape, plywood frames, and visible expanded metal grilles. Taking apart and reassembling a 63 is easier, but is tedious due to the many connections to the panels, but the connections are pretty obvious.
The electronics in the 63's is more complex. There is a clamp circuit that is fairly involved. The high voltage supply is a bit more complex. The delay lines are also more complex but are essentially trouble free in my experience.
In terms of ER kits, I think the 63 kit is more complete and overall easier to follow and produces a result that is very close to the original Quad speaker when complete. The ER 57 kit is a little bit of a departure from the way I rebuild the 57's, and the manual (as I've seen it a year or so ago) is a little less complete than the 63 manual.
It's really a matter of personal taste (57 vs 63) and what you can find to buy and rebuild. I've had a lot of both types over the 15+ years I've been rebuilding and using them. The 57's have a midrange transparency that is hard to beat, arguably the best out there in the sweet spot. The 57's don't have as good of extension in the bass and treble. The 63's have better dispersion, you can get very good tonal balance farther away from the sweet spot than with the 57's. The 63's look much much nicer than the 57's (IMHO), with the 57's looking like they were built in a garage.
Sheldon
The 63 frames and stators are higher precision. The glue that holds the circuit board stators to the frames is the Achilles heal of the 63's and the crappy quad glue has paid for many vacations for me. There are two more panels on the 63's, but they are a lot easier to rebuild. The 63's are held together better, not relying on tape, plywood frames, and visible expanded metal grilles. Taking apart and reassembling a 63 is easier, but is tedious due to the many connections to the panels, but the connections are pretty obvious.
The electronics in the 63's is more complex. There is a clamp circuit that is fairly involved. The high voltage supply is a bit more complex. The delay lines are also more complex but are essentially trouble free in my experience.
In terms of ER kits, I think the 63 kit is more complete and overall easier to follow and produces a result that is very close to the original Quad speaker when complete. The ER 57 kit is a little bit of a departure from the way I rebuild the 57's, and the manual (as I've seen it a year or so ago) is a little less complete than the 63 manual.
It's really a matter of personal taste (57 vs 63) and what you can find to buy and rebuild. I've had a lot of both types over the 15+ years I've been rebuilding and using them. The 57's have a midrange transparency that is hard to beat, arguably the best out there in the sweet spot. The 57's don't have as good of extension in the bass and treble. The 63's have better dispersion, you can get very good tonal balance farther away from the sweet spot than with the 57's. The 63's look much much nicer than the 57's (IMHO), with the 57's looking like they were built in a garage.
Sheldon
- Status
- Not open for further replies.