ESL989 has more panels = lower frequency
And it is wery wery strange that non of the resistors in series with the bass panels didn´t change.
I hope that i´m wrong but I suspect that after Walker era... no one really has the ... hmm lets say "intention" to change anything except esthetics and mechanical/structural stuff. Which is good of course... But it is really the same speaker since 43 years back. And this in itself must be an unique property of any equipment for sound reproduction that has not changed. This is plain and simple remarkable IMHO
And it is wery wery strange that non of the resistors in series with the bass panels didn´t change.
I hope that i´m wrong but I suspect that after Walker era... no one really has the ... hmm lets say "intention" to change anything except esthetics and mechanical/structural stuff. Which is good of course... But it is really the same speaker since 43 years back. And this in itself must be an unique property of any equipment for sound reproduction that has not changed. This is plain and simple remarkable IMHO
What current are you expecting here to make such an investment worth while?
Perhaps, because isn't other pair QUAD ESL989 without miles of thin copper wire for A-B test... But I'm sure, we can hear the speaker cable with a lot of trafos and inductor wires too (for example, between tube amp with OT and ESL). This is a very nice thing of our hobby.
But you are right. The tipical audiophile can try (and worries about) a lot of speaker cable, but for most of audiophiles is "restricted area" the ESL's viscera.
I have a tube amp that gladly delivers Klass A power in 2ohms load, so since there is resistors in series with ESL63 trafo 1,5ohm//220uf (which i have removed completely) and also some 1,65ohms in series with each transformer. I decided to short those also.
Instead i could use typical telephone wire (very thin) with low inductance and low capacitance but higher resistance. Why? Convenient, and if there was much of a sound change it was to the better. Even my dipole subwoofers that many years ago was demonstrated... i used very thin telephone wires there also and the added resistance gave a good Qts of 0,7 (driver had 0,65) the sound was great! Great power with clear distinct bass response. Some listeners would not believe me that sooo much power possibly could go through those thin wires. They had to investigate it to believe it... Then they tried to convince me that the bass would bee soo much better with a thicker wire!! And I said that I have tried, and this thin wire is best. So if i had not showed the wires their impression had remained GREAT PUNCHY LOW EXTENSION BASS... but after the visual impression they changed their mind.
Instead i could use typical telephone wire (very thin) with low inductance and low capacitance but higher resistance. Why? Convenient, and if there was much of a sound change it was to the better. Even my dipole subwoofers that many years ago was demonstrated... i used very thin telephone wires there also and the added resistance gave a good Qts of 0,7 (driver had 0,65) the sound was great! Great power with clear distinct bass response. Some listeners would not believe me that sooo much power possibly could go through those thin wires. They had to investigate it to believe it... Then they tried to convince me that the bass would bee soo much better with a thicker wire!! And I said that I have tried, and this thin wire is best. So if i had not showed the wires their impression had remained GREAT PUNCHY LOW EXTENSION BASS... but after the visual impression they changed their mind.
I have dealt with the "Crosby mods" on Quad 63's and they change the last 20 inches of wire for different and much larger gauge wire. I believe they charged a lot for it too. In general, made a big mess of things...
Here's some advise from the designer:
Sheldon
A related question if I may: what is the typical leakage current (or load resistance) of an ESL 63 panel?
What is the typical ripple at the bias from the voltage multiplier output?
I've searched but didn't get further then some unspecified 'fast' and 'slow' blinking of the neon.
One of the things I am working on is an adjustable bias generator to investigate what that does to the performance.
Some numbers to this would be appreciated.
Jan
What is the typical ripple at the bias from the voltage multiplier output?
I've searched but didn't get further then some unspecified 'fast' and 'slow' blinking of the neon.
One of the things I am working on is an adjustable bias generator to investigate what that does to the performance.
Some numbers to this would be appreciated.
Jan
If the ignit voltage value of the neon and the paralleled capacitor's value is known, then you might calculate the charge per blink. And if the blinking frequency of the neon is known, then you might calculate the charge suppliyed by time into the panels for reloading them. And from this value [charge by time for 4 or 6 panels] you might calculate an approximative average load resistance for a single panel. And from this resistance, also the average leakage current per panel. Assumed, you know the HT voltage. I might be wrong ???
Last edited:
It varies all over the place depending on the relative humidity. When you switch off an ESL-63 while it is playing - which you should never do at high volume! - you will hear the sound fade out in a minute or so on a wet day. On a very dry day, after an hour, it still plays almost as loud as when it was on.
Regarding the neon lamp, what matters is the difference between the ignition and maintaining voltages. Don't forget the 47 nF capacitor shunted across the lamp.
Regarding the neon lamp, what matters is the difference between the ignition and maintaining voltages. Don't forget the 47 nF capacitor shunted across the lamp.
Last edited:
Hi Jan,
The load is in the hundreds of MOhms and the four 20nF caps are in series 5nF supporting the Bias voltage.
Lets say for simplicity, Vbias is 5000Volt, resulting in a current of 5000V/100e6Ohm= 50uA.
While I=C*dU/dT and dT 20msec for 50Hz, the ripple dU becomes I*dt/C = 50e-6A*20e-3Sec/5nF = 200Volt
Apart from the influence on the max output level, that will be affected by the magnitude of the bias voltage, the speakers efficiency will also drop.
See attachment for a mains supply of resp 225V and 140Volt that translates directly proportional to the HV Bias Voltage
FR will not be affected but output drops slightly.
So the above calculated 200V ripple on 5000 volt will have no effect IMO.
Hans
P.S. the curves in the image are not simulated but acoustically measured.
Attachments
