Intermittent treble, that's unusual. Sounds like a loose contact (solder lug) on the treble panel or perhaps a broken capacitor/resistor in your crossover.
If one bass panel is playing and the other is not then it's either a corroded contact (check the wires on the panels), and otherwise it's the bass panel itself since they are in parallel.
If one bass panel is playing and the other is not then it's either a corroded contact (check the wires on the panels), and otherwise it's the bass panel itself since they are in parallel.
Actually, the bass panels are almost always less then optimal. They will play and on very rare occasions they are still close to specification, but I'd say that's less than 1 in 20. That does not mean that you should go and fix them, just realize that they are probably not as good as they might.
Some electrostatic speakers fail dramatically when they go bad. The Quad ESL slowly deteriorates but keeps playing (most of the time). That's why a lot of people think they have a good functioning speaker while in fact it's not even close to what it could be. Some of them however are still really good after 40 years. A lot of modern stats don't even make it to 10 years. So far for progress
Aging of the diaphragm and coating. It loses sensitivity and the resonance frequency goes up. Both combined make for a rather poor experience and this is probably the main reason it acquired a reputation for 'having no bass'.
One Thing Audio/Manufacturers/Quad/History/Peter Walker
We stretch it up to the required tension and stick it in the oven. In its cold form, if you stretch it, it will creep. But by putting it in the oven under tension all the molecules will line up to the most stable position so the tension won't change. This is quite important because otherwise the resonance of the speaker would change and it is carefully designed to a particular value. If it was too slack it would hit the plates, too tight the resonance would be too high. Secret ingredient X is used as a damping layer on the Mylar and then we bond the frame to the electrode structure."
"So you stretch it, stabilise it then bond it to one side of the electrodes. It is a push pull device with two stators and one moving diaphragm in between."
We stretch it up to the required tension and stick it in the oven. In its cold form, if you stretch it, it will creep. But by putting it in the oven under tension all the molecules will line up to the most stable position so the tension won't change. This is quite important because otherwise the resonance of the speaker would change and it is carefully designed to a particular value. If it was too slack it would hit the plates, too tight the resonance would be too high. Secret ingredient X is used as a damping layer on the Mylar and then we bond the frame to the electrode structure."
"So you stretch it, stabilise it then bond it to one side of the electrodes. It is a push pull device with two stators and one moving diaphragm in between."
The text that you cut-and-pasted is about the ESL 63, which uses 3.5 micron Mylar for the panels. The Quad ESL 57 used 12 micron PVDC film for the bass panel, a material that ages quite differently! I've seen bass panels with a resonance frequency of over 100Hz. I can show you hundreds of frequency response measurements from different Quad ESL 57 bass panels. Almost all of them have raised resonance frequencies.
The text that you cut-and-pasted is about the ESL 63, which uses 3.5 micron Mylar for the panels. The Quad ESL 57 used 12 micron PVDC film for the bass panel, a material that ages quite differently! I've seen bass panels with a resonance frequency of over 100Hz. I can show you hundreds of frequency response measurements from different Quad ESL 57 bass panels. Almost all of them have raised resonance frequencies.
Same effect observed by myself: I have re built two pairs and both had a resonance frequency above 100Hz, lowering to sthg around 50Hz after.
I got a lot of information from Baxandall's book. Just lacking: gauge of each wire.
Last edited:
I answer to myself: I just dis-assembled the transformer.
Regarding each of the 2 smallest coils (for the bass) first: wire has a diameter of about 0.2mm. Overall resistivity is about 2.25kOhm.
This makes length=4158m as first estimate (a lot, there must be something wrong !), computed that way:
2.25*1000/(1.7e-8)*PI*(0.1/1000)^2
Next step: unwind and count number of turns.
Winding seems very tight; not sure I can re-do it easily with a manual winder: the coil might be too thick if not done properly.
Last edited:
Hello,
Not yet. I will do this later as this is not essential for me right now: this is a spare part.
The turns ratio is given in Baxandall; I do not it by heart. But this does not give the number of turns, just the ratio. I will get some equipment to count the number of turns as it seems quite high and thus painful to count without making an error.
The transformer is made of one main core + 2 additional cores, adding power for the bass panels. Hard to explain but just have a look to schematic (widely available) to make up your mind. Just remember that the bass panels need the 3 cores to work. That's why I think it would be difficult to split the transformer in two and still keep the same behaviour.
But is should work; it would just be different.
I used a high end multimeter (Fluke 179) thus this should not be an issue. My main error may come from the measurement of the diameter.
Anyway I will conlude once I will have unwinded the coil.
Unless someone who has done the same thing gives some figures meanwhile.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.