Hello to all ESL experts and enthusiasts!
After a house move, a pair of very old (SN 2086/7) Quad ESL63s are buzzing, humming and hissing more than I remember - and enough to stop me using them and find some answers. After a bit of research on possible causes (and in the hope that it was only electronics to blame) my first port of call was replacing caps and diodes in the EHT unit and the 1000µF 16V PS cap on the protection board.
Unfortunately, this made no apparent difference and so the most likely (and most feared) cause was now rearing its head - that I have the dreaded 'stator separating from matrix' problem. Stripping them down to the bare panels has confirmed this. So I now have two options: delivering them to a professional service to renovate the panels or having a go myself. The first is easier, far more expensive and potentially means me being without them for a long time. The second is cheaper, a lot more personally satisfying but also quite a challenge/learning curve. I say that as someone who in my 20s worked for both Rogers as QA on BBC monitors and designer of the Studio 1a, and then later as a senior engineer at KEF with the privilege of working on advanced loudspeakers with such pioneers as Laurie Fincham and Richard Small. So I'm not afraid of electroacoustics, soldering or circuit diagrams! Unfortunately, KEF went the same way as Quad but did so first i.e. bought out by a Chinese company that was more interested in the brand name than the R&D brains and so squandered a considerable talent pool and focused on marketing and moving as much production as possible to China to save money. 🙁
First of all, a couple of apologies. This is quite a long post and covers ground that has been discussed before. In particular I have read a number of detailed threads here (e.g. this monster ESL Diaphragm coating). The reason I think it's relevant are:
0.1) I'm sure details have changed over the years as people here gained more experience with important issues like adhesives and membrane coatings
and
0.2) many others who are in the the same position as me would I'm sure like to have the key 'difficult' renovation issues all collected in one place.
So here goes with what I think are the toughest things to get right:
1) Choice of adhesives
What we want is something that:
- lasts a long time, in many realistic use cases (> 10 years?)
- is not too toxic to work with
- is relatively easy to apply
- is relatively easy to get hold of and not too expensive (in the UK or EU hopefully)
And there are 3 places where it needs to be used, only the first 2 of which seem to widely discussed.
1.1) stator to matrix
1.2) membrane to matrix
1.3) nylon damping to stator
In my case, (1.3) is relevant as several of the panels have the nylon damping material peeling away from the stator and this obviously needs to be addressed carefully.
I have seen several adhesives discussed:
- polyurethane
- contact adhesive
- epoxy
- super glue/cyanoacrylate
- others?
I would like to canvas opinion here from those with experience and expertise about which adhesives (with brand names or type numbers if possible please) have stood the test of time and satisfy what I believe are the key requirements in each case. It would be nice to use something that can be removed without too much grief when the next renovation is required - this would tend to downgrade epoxy and possibly cyanoacrylate also.
2) Acquiring, stretching and appropriately tensioning the membrane
Most seem to recommend a Mylar or PET membrance of 3-6 microns in thickness. This is not easy to find in the UK or EU without buying in bulk or paying what seems like a very inflated price for small quantities from specialists. Best links I have so far (they both do 3 & 6 micron):
Electrostatic Speaker Membrane Dupont Mylar C 6um 20M | eBay
Winkel - ESL Plus
By the way, do people think the extra mass of 6 micron has too much penalty (i.e. earlier high frequency rolloff) for the extra strength and easier handling?
Once I have this, it needs to be stretched on a jig or frame of some kind. I have seen some talk of a pneumatic arrangement though I have not been able to find details of this. I believe that some heat treatment can also be beneficial with either a hair dryer or heat gun, but it's not clear whether this should be before or after it has been attached to the matrix and stator.
Then, how does one ensure that it is at the correct tension? I have heard talk of methods with coins, tension gauges and microphones. The only good panel I have disassembled so far had a fundamental resonance at 31-34Hz depending on how you tapped the matrix, with a strongish second peak at about 71Hz. Is this normal or has the unit stretched over the years? I have seen talk of 70-72Hz being the required resonance but I'm not sure if that is the acoustic resonance of the bare unit or the combined electro-acoustic resonance in circuit. I have both coins(!) and a good measuring microphone/system so reliable methods using one or both of those that match the original Quad spec are sought after.
3) Applying the conductive surface to the diaphragm
I have read a LOT about this, and many different opinions, including the mega-thread mentioned above which I finished at about 4am this morning!
To repeat my thoughts for requirements from the adhesives section. "What we want is something that:
- lasts a long time, in many realistic use cases (> 10 years?)
- is not too toxic to work with
- is relatively easy to apply
- is relatively easy to get hold of and not too expensive(in the UK or EU hopefully)"
I guess an extra couple of considerations here are "how well does it bond with the mylar without any surface finishing treatment?" and "how consistent is it over realistic temperature and humidity variations?"
There seems to be a number of options:
3.1) PVA adhesive with graphite powder or black ink
3.2) Commercial anti-static electronics products such as Licron Crystal, ACL Staticide 6300/6500, Statclear A50, others?
3.3) ESL-specific products from enthusiasts/specialists (e.g. EC/HTEC-coating from MJ Dikstra here, the coatings from ER audio in Perth Australia, Shackman in Germany, others?).
3.4) Weird and wonderful ideas like nylon fishing line melted in toxic solvents, anti-static floor polish, Turtle Wax F21, others?
I like the simplicity, cheapness and availability of (3.1) but there are some concerns about it being hygroscopic and perhaps therefore not lasting very well. Any brand names for recommended adhesives or inks are welcomed. Perhaps some PVAs are better than others in this respect? For example:
500ml PVA BOND ADHESIVE AND SEALER PRIMER & BONDING AGENT PLASTER CEMENT TIMBER 5060021361653 | eBay
and
Liquiwire™ 100 conductive graphite shielding paint paste glue ink pen | eBay.
I already have graphite powder. Was this the original Quad method? It seems to have lasted well on my units although the measured resistances given below vary widely across the membranes.
I'm happy to explore something in (3.2) as long as it's available in sensible quantities and not too expensive. I have read some concerns about how long Staticide stays conductive.
Also happy to go down the path of (3.3) and in fact am considering buying the entire renovation kit from ER audio. Any particular recommendations and if so with price and availability?
I like options in (3.4) if they are proven and not too toxic to work with. But for instance, how long do floor or car polishes stick and stay conductive on a thin mylar membrane with 5kV across it vibrating at 10kHz?
Then once you have the right thing, you have to apply it. There seems to be some consensus here that a foam pad or roller works well, possibly after diluting the coating medium with water or alcohol as appropriate.
There is an entire debate about how high the resistance should be, with the higher the better in terms of charge migration/distortion up to the point where it takes too long to charge up and/or other very high resistance paths start to enter the picture. My own simple measurements in (apparent) MOhms taken with a basic multimeter that measures up to 20MOhms and 2x 2 pence coins sitting on the original Quad membranes produce widely varying values:
With the coin edges 2mm apart
============================
Bass panel: 3.3 - 7.3
MF panel: 2.7 - >20
With the coin edges 25mm apart
============================
Bass panel: 7.2 - 18
MF panel: 8.2 - >20
I doubt if these are accurate results but they at least give me something to aim for, presumably a bit higher would be good so that I can just about read something near 20MOhms with coins 2mm apart. Interestingly, I could not get any measurements <20 MOhms on the reverse side of the membranes so that suggests that they are not a uniform layer.
Please feel free to send links to discussions on diyAudio or elsewhere on the web if they address any of the questions that I have asked here with definitive answers, and thanks very much in advance for having had the patience to read all the way down to here!
I hope this thread becomes a useful resource for others in my position, of whom I'm sure there are many.
Happy to be emailed/CC'd directly if you would like to: hop333 A T gmail D O T com
TIA
Michael
After a house move, a pair of very old (SN 2086/7) Quad ESL63s are buzzing, humming and hissing more than I remember - and enough to stop me using them and find some answers. After a bit of research on possible causes (and in the hope that it was only electronics to blame) my first port of call was replacing caps and diodes in the EHT unit and the 1000µF 16V PS cap on the protection board.
Unfortunately, this made no apparent difference and so the most likely (and most feared) cause was now rearing its head - that I have the dreaded 'stator separating from matrix' problem. Stripping them down to the bare panels has confirmed this. So I now have two options: delivering them to a professional service to renovate the panels or having a go myself. The first is easier, far more expensive and potentially means me being without them for a long time. The second is cheaper, a lot more personally satisfying but also quite a challenge/learning curve. I say that as someone who in my 20s worked for both Rogers as QA on BBC monitors and designer of the Studio 1a, and then later as a senior engineer at KEF with the privilege of working on advanced loudspeakers with such pioneers as Laurie Fincham and Richard Small. So I'm not afraid of electroacoustics, soldering or circuit diagrams! Unfortunately, KEF went the same way as Quad but did so first i.e. bought out by a Chinese company that was more interested in the brand name than the R&D brains and so squandered a considerable talent pool and focused on marketing and moving as much production as possible to China to save money. 🙁
First of all, a couple of apologies. This is quite a long post and covers ground that has been discussed before. In particular I have read a number of detailed threads here (e.g. this monster ESL Diaphragm coating). The reason I think it's relevant are:
0.1) I'm sure details have changed over the years as people here gained more experience with important issues like adhesives and membrane coatings
and
0.2) many others who are in the the same position as me would I'm sure like to have the key 'difficult' renovation issues all collected in one place.
So here goes with what I think are the toughest things to get right:
1) Choice of adhesives
What we want is something that:
- lasts a long time, in many realistic use cases (> 10 years?)
- is not too toxic to work with
- is relatively easy to apply
- is relatively easy to get hold of and not too expensive (in the UK or EU hopefully)
And there are 3 places where it needs to be used, only the first 2 of which seem to widely discussed.
1.1) stator to matrix
1.2) membrane to matrix
1.3) nylon damping to stator
In my case, (1.3) is relevant as several of the panels have the nylon damping material peeling away from the stator and this obviously needs to be addressed carefully.
I have seen several adhesives discussed:
- polyurethane
- contact adhesive
- epoxy
- super glue/cyanoacrylate
- others?
I would like to canvas opinion here from those with experience and expertise about which adhesives (with brand names or type numbers if possible please) have stood the test of time and satisfy what I believe are the key requirements in each case. It would be nice to use something that can be removed without too much grief when the next renovation is required - this would tend to downgrade epoxy and possibly cyanoacrylate also.
2) Acquiring, stretching and appropriately tensioning the membrane
Most seem to recommend a Mylar or PET membrance of 3-6 microns in thickness. This is not easy to find in the UK or EU without buying in bulk or paying what seems like a very inflated price for small quantities from specialists. Best links I have so far (they both do 3 & 6 micron):
Electrostatic Speaker Membrane Dupont Mylar C 6um 20M | eBay
Winkel - ESL Plus
By the way, do people think the extra mass of 6 micron has too much penalty (i.e. earlier high frequency rolloff) for the extra strength and easier handling?
Once I have this, it needs to be stretched on a jig or frame of some kind. I have seen some talk of a pneumatic arrangement though I have not been able to find details of this. I believe that some heat treatment can also be beneficial with either a hair dryer or heat gun, but it's not clear whether this should be before or after it has been attached to the matrix and stator.
Then, how does one ensure that it is at the correct tension? I have heard talk of methods with coins, tension gauges and microphones. The only good panel I have disassembled so far had a fundamental resonance at 31-34Hz depending on how you tapped the matrix, with a strongish second peak at about 71Hz. Is this normal or has the unit stretched over the years? I have seen talk of 70-72Hz being the required resonance but I'm not sure if that is the acoustic resonance of the bare unit or the combined electro-acoustic resonance in circuit. I have both coins(!) and a good measuring microphone/system so reliable methods using one or both of those that match the original Quad spec are sought after.
3) Applying the conductive surface to the diaphragm
I have read a LOT about this, and many different opinions, including the mega-thread mentioned above which I finished at about 4am this morning!
To repeat my thoughts for requirements from the adhesives section. "What we want is something that:
- lasts a long time, in many realistic use cases (> 10 years?)
- is not too toxic to work with
- is relatively easy to apply
- is relatively easy to get hold of and not too expensive(in the UK or EU hopefully)"
I guess an extra couple of considerations here are "how well does it bond with the mylar without any surface finishing treatment?" and "how consistent is it over realistic temperature and humidity variations?"
There seems to be a number of options:
3.1) PVA adhesive with graphite powder or black ink
3.2) Commercial anti-static electronics products such as Licron Crystal, ACL Staticide 6300/6500, Statclear A50, others?
3.3) ESL-specific products from enthusiasts/specialists (e.g. EC/HTEC-coating from MJ Dikstra here, the coatings from ER audio in Perth Australia, Shackman in Germany, others?).
3.4) Weird and wonderful ideas like nylon fishing line melted in toxic solvents, anti-static floor polish, Turtle Wax F21, others?
I like the simplicity, cheapness and availability of (3.1) but there are some concerns about it being hygroscopic and perhaps therefore not lasting very well. Any brand names for recommended adhesives or inks are welcomed. Perhaps some PVAs are better than others in this respect? For example:
500ml PVA BOND ADHESIVE AND SEALER PRIMER & BONDING AGENT PLASTER CEMENT TIMBER 5060021361653 | eBay
and
Liquiwire™ 100 conductive graphite shielding paint paste glue ink pen | eBay.
I already have graphite powder. Was this the original Quad method? It seems to have lasted well on my units although the measured resistances given below vary widely across the membranes.
I'm happy to explore something in (3.2) as long as it's available in sensible quantities and not too expensive. I have read some concerns about how long Staticide stays conductive.
Also happy to go down the path of (3.3) and in fact am considering buying the entire renovation kit from ER audio. Any particular recommendations and if so with price and availability?
I like options in (3.4) if they are proven and not too toxic to work with. But for instance, how long do floor or car polishes stick and stay conductive on a thin mylar membrane with 5kV across it vibrating at 10kHz?
Then once you have the right thing, you have to apply it. There seems to be some consensus here that a foam pad or roller works well, possibly after diluting the coating medium with water or alcohol as appropriate.
There is an entire debate about how high the resistance should be, with the higher the better in terms of charge migration/distortion up to the point where it takes too long to charge up and/or other very high resistance paths start to enter the picture. My own simple measurements in (apparent) MOhms taken with a basic multimeter that measures up to 20MOhms and 2x 2 pence coins sitting on the original Quad membranes produce widely varying values:
With the coin edges 2mm apart
============================
Bass panel: 3.3 - 7.3
MF panel: 2.7 - >20
With the coin edges 25mm apart
============================
Bass panel: 7.2 - 18
MF panel: 8.2 - >20
I doubt if these are accurate results but they at least give me something to aim for, presumably a bit higher would be good so that I can just about read something near 20MOhms with coins 2mm apart. Interestingly, I could not get any measurements <20 MOhms on the reverse side of the membranes so that suggests that they are not a uniform layer.
Please feel free to send links to discussions on diyAudio or elsewhere on the web if they address any of the questions that I have asked here with definitive answers, and thanks very much in advance for having had the patience to read all the way down to here!
I hope this thread becomes a useful resource for others in my position, of whom I'm sure there are many.
Happy to be emailed/CC'd directly if you would like to: hop333 A T gmail D O T com
TIA
Michael
Last edited:
I was thinking... oh no... not another thread about QUAD (since there is so many already..)
But I was soo wrong!
You got most of it right, and seems to have done the best compilation of knowledge there is really.
Some comments:
1: Polyurethane etylacetate glue diluted with acetone is good.
https://media-pms2.schoenox.net/casco/docs/cascouniversalkraftblister_se_se_msds.pdf
I really recommend it, you can dilute it in acetone, giving you extra time to mess around before its dry.
Dilute it 1:1 with acetone as a start.
You will have time to apply it on stator two rounds with a brush before assembly.
Then apply a lot of pressure!
Be sure that you have some rubber like material that is flexible that presses down the mylar. I use approx 300-400 kg of pressure, (a car jack, in my specially welded combined strech jig.
Sounds much, but calculated over the total area it´s really a minimum.
2: Coating, good coverage.
3: Measuring resistance.
I recommend using a DC supply and measure the voltage over a shunt resistor. 1-10 Meg ohm shunt is ok and also protects you from lethal currents.
Be aware!! Voltage above 60 Volts requires knowledge and is lethal!! I do not recommend anyone to try higher voltage than this at home!! (I use approx 350Volt, and it works, never tried lower voltage and i do not know for sure that 60volts is good enough, sorry)
If you use two straight plates, place the plates at the same distance as the plate width.
As an example, 3cm wide plates spaced 3cm from each other.
Difference between Anti-Static, Dissipative, Conductive, and Insulative
Here is a video of some compiled measurements:
Quad ESL63 Repair Part02C Resistivity conclusion - YouTube
My goal is to have 200Meg - 500Meg Ohm/square on new panels, but if you are using an active crossover cutting between 100-200 hz, 50Meg Ohm is fine. 200Meg Ohm can be done with some practice and a repeatable process, ideally a highly automated process is preferable. But not even QUAD seems to have implemented that.. which is a shame really.
Pre coated (painted, sputtered) film with right properties is something I have been looking for. I got an offer some years ago but i did not have enough cash to buy 20000 square meter, which was minimum quantity.
A redesign of the panels needs to be done to have this to work as well, to avoid leakage currents. But that is another story..
But I was soo wrong!
You got most of it right, and seems to have done the best compilation of knowledge there is really.
Some comments:
1: Polyurethane etylacetate glue diluted with acetone is good.
https://media-pms2.schoenox.net/casco/docs/cascouniversalkraftblister_se_se_msds.pdf
I really recommend it, you can dilute it in acetone, giving you extra time to mess around before its dry.
Dilute it 1:1 with acetone as a start.
You will have time to apply it on stator two rounds with a brush before assembly.
Then apply a lot of pressure!
Be sure that you have some rubber like material that is flexible that presses down the mylar. I use approx 300-400 kg of pressure, (a car jack, in my specially welded combined strech jig.
Sounds much, but calculated over the total area it´s really a minimum.
2: Coating, good coverage.
3: Measuring resistance.
I recommend using a DC supply and measure the voltage over a shunt resistor. 1-10 Meg ohm shunt is ok and also protects you from lethal currents.
Be aware!! Voltage above 60 Volts requires knowledge and is lethal!! I do not recommend anyone to try higher voltage than this at home!! (I use approx 350Volt, and it works, never tried lower voltage and i do not know for sure that 60volts is good enough, sorry)
If you use two straight plates, place the plates at the same distance as the plate width.
As an example, 3cm wide plates spaced 3cm from each other.
Difference between Anti-Static, Dissipative, Conductive, and Insulative
Here is a video of some compiled measurements:
Quad ESL63 Repair Part02C Resistivity conclusion - YouTube
My goal is to have 200Meg - 500Meg Ohm/square on new panels, but if you are using an active crossover cutting between 100-200 hz, 50Meg Ohm is fine. 200Meg Ohm can be done with some practice and a repeatable process, ideally a highly automated process is preferable. But not even QUAD seems to have implemented that.. which is a shame really.
Pre coated (painted, sputtered) film with right properties is something I have been looking for. I got an offer some years ago but i did not have enough cash to buy 20000 square meter, which was minimum quantity.
A redesign of the panels needs to be done to have this to work as well, to avoid leakage currents. But that is another story..
Last edited:
Thanks for that input esl63, very interesting and it sounds like you have a lot of experience in this area.
So would you recommend "Polyurethane etylacetate glue diluted with acetone" for all 3 of the bonds that I ask about? It's difficult to find information about the 3rd one but I really need to do it without clogging either the stator or the nylon damping and it must stick securely or else the membrane will start hitting it.
I'm still puzzling about how to tension the mylar without a specialist jig. Current idea is stretching across glass with many strips of gaffa tape all tensioned to 2.5Kg until it settles but I also worry that this won't be enough to get the right resonant freq.
So would you recommend "Polyurethane etylacetate glue diluted with acetone" for all 3 of the bonds that I ask about? It's difficult to find information about the 3rd one but I really need to do it without clogging either the stator or the nylon damping and it must stick securely or else the membrane will start hitting it.
I'm still puzzling about how to tension the mylar without a specialist jig. Current idea is stretching across glass with many strips of gaffa tape all tensioned to 2.5Kg until it settles but I also worry that this won't be enough to get the right resonant freq.
Gluing the net i would try to avoid acetone or other solvents, so you have a point.
I have used it but one has to be careful to not "melt" the net and clogging it.
Apply the glue as is to the stator and let dry slightly before attaching the net.
I have used it but one has to be careful to not "melt" the net and clogging it.
Apply the glue as is to the stator and let dry slightly before attaching the net.
May I make a humble suggestion:
Quad 63 Repair Kit
This will provide you with all the raw materials and the instructions to rebuild all your panels. It will be FAR cheaper and quicker than fighting down that road yourself.
That said, I use slightly different materials, but my techniques and materials are close enough to this kit that I'd just buy the kit.
To answer your questions sort of:
Choice of adhesives:
For the stators to support structure. You want something with a good lifespan and not brittle. There's a bunch of choices but sparingly applied polyurathane glues work well here.
For the diaphragm, there are a lot of choices. I prefer a fairly fast setting adhesive so that the diaphragm doesn't have a chance to creep and lose tension in my tensioning setup. I use 5 minute epoxy, but 5 minutes is a short pot life and you need to know what you are doing and can't dawdle.
For the damping screen: I've never needed to replace one, but if I did, I'd probably use 3M 77 spray adhesive.
Stretching and tensioning membrane. This will requitre some experimenting. I'd use the assembled panel's resonant frequency as a guide. Keep in mind the Fs will shift downward as the mylar relaxes over time as the panel breaks in. so you want to aim for a Fs that is higher than you want with the knowledge of how far it decreases as it breaks in. The way you tension and how much is a function of the material you choose and the geometry of the sheet you are tensioning.
As for 6um film, that's pretty thick and will start rolling off about about 16KHz. I use 3.2um or 3.5um.
The conductive coating is actually pretty easy these days, but you have to become a part-time chemist. There are a ton of liquids and powders aimed at making static dissipative surfaces or even the LCD screen manufacture. I use a proprietary one for my rebuilds that is based on a metal oxide nanoparticle suspended in a water based polyester resin with some dispersant and an additive to lower the surface tension for good sheeting and flow-out. It's a wipe-on coating.
Hope that helps,
Sheldon
Quad 63 Repair Kit
This will provide you with all the raw materials and the instructions to rebuild all your panels. It will be FAR cheaper and quicker than fighting down that road yourself.
That said, I use slightly different materials, but my techniques and materials are close enough to this kit that I'd just buy the kit.
To answer your questions sort of:
Choice of adhesives:
For the stators to support structure. You want something with a good lifespan and not brittle. There's a bunch of choices but sparingly applied polyurathane glues work well here.
For the diaphragm, there are a lot of choices. I prefer a fairly fast setting adhesive so that the diaphragm doesn't have a chance to creep and lose tension in my tensioning setup. I use 5 minute epoxy, but 5 minutes is a short pot life and you need to know what you are doing and can't dawdle.
For the damping screen: I've never needed to replace one, but if I did, I'd probably use 3M 77 spray adhesive.
Stretching and tensioning membrane. This will requitre some experimenting. I'd use the assembled panel's resonant frequency as a guide. Keep in mind the Fs will shift downward as the mylar relaxes over time as the panel breaks in. so you want to aim for a Fs that is higher than you want with the knowledge of how far it decreases as it breaks in. The way you tension and how much is a function of the material you choose and the geometry of the sheet you are tensioning.
As for 6um film, that's pretty thick and will start rolling off about about 16KHz. I use 3.2um or 3.5um.
The conductive coating is actually pretty easy these days, but you have to become a part-time chemist. There are a ton of liquids and powders aimed at making static dissipative surfaces or even the LCD screen manufacture. I use a proprietary one for my rebuilds that is based on a metal oxide nanoparticle suspended in a water based polyester resin with some dispersant and an additive to lower the surface tension for good sheeting and flow-out. It's a wipe-on coating.
Hope that helps,
Sheldon
@hop333
stokessd is one the top re-builder/restorer of Quad speakers. He really knows what he is about with these speakers (ESL and '63s.)
stokessd is one the top re-builder/restorer of Quad speakers. He really knows what he is about with these speakers (ESL and '63s.)
The original panel uses contact glue to the mylar, which breaks down over time, (ozone mostly).
Cleaning the panels from glue/mylar is easy done with a sharp knife and acetone.
Panels that has been refurbished with Epoxy is hard to clean!
Worse is that epoxy is stiff and gives sharp edges which can damage the membrane.
When I glue the membrane I "paint" the membrane with the urethane glue some mm extra into the membrane area, giving the membrane some damping at the edges.
This gives two benefits:
1: Damping of resonance reflection at the edges.
2: Smoother transition between the moving membrane and the stiff frame.
Many Epoxy glued membranes seems to more easy rip the membrane at the edges in a way that I seldom see on original panels.
Stator to frame can be glued with epoxy. And in all gluing process cleanliness and priming is of great importance.
Cleaning the panels from glue/mylar is easy done with a sharp knife and acetone.
Panels that has been refurbished with Epoxy is hard to clean!
Worse is that epoxy is stiff and gives sharp edges which can damage the membrane.
When I glue the membrane I "paint" the membrane with the urethane glue some mm extra into the membrane area, giving the membrane some damping at the edges.
This gives two benefits:
1: Damping of resonance reflection at the edges.
2: Smoother transition between the moving membrane and the stiff frame.
Many Epoxy glued membranes seems to more easy rip the membrane at the edges in a way that I seldom see on original panels.
Stator to frame can be glued with epoxy. And in all gluing process cleanliness and priming is of great importance.
Thanks @stokessd for these helpful tips from a known expert on the matter. I have already contacted Rob at ERAudio and he has been very helpful in terms of what I need. I'll probably just go with his kit as you suggest for this first attempt which will I'm sure generate plenty of other headaches!
I am keen to become a bit more independent eventually in case I need to do it again or want to do this for others, hence the fundamental questions about adhesives, coating material etc. For example, I like the idea of using a bicycle inner tube to apply the tension to the mylar evenly and repeatably and have fat bike inner tubes 26" x 4" that will do the job very nicely I think with an appropriate jig. One could even measure the panel resonance on the jig and adjust precisely with more or less PSI in the tube. I have heard about this type of jig and can imagine how it would work but have not seen actual photos or designs anywhere yet.
I am keen to become a bit more independent eventually in case I need to do it again or want to do this for others, hence the fundamental questions about adhesives, coating material etc. For example, I like the idea of using a bicycle inner tube to apply the tension to the mylar evenly and repeatably and have fat bike inner tubes 26" x 4" that will do the job very nicely I think with an appropriate jig. One could even measure the panel resonance on the jig and adjust precisely with more or less PSI in the tube. I have heard about this type of jig and can imagine how it would work but have not seen actual photos or designs anywhere yet.
Back 15 or so years ago, over time I rebuilt, or tried to, at least 50 panels. Used Rob's kits which mostly worked out fine. Made every mistake you could during the process. Did like his method of supporting the mylar on a big piece of glass and tensioning it using the supplied force guage.
Pain in the butt was removing all the old glue from the rim of the "egg crate".
Had an application problem with the "special sauce" he provided to add conductivity to the mylar. A number of times it seemed to "clot" and not provide an even coating. May have been the humid conditions I was working under.
Anyway, found that most the rebuilt panels didn't last too long - eventually gave up and sold the speakers. Even rebuilt panels from W Piquet only lasted a few years.
Did find that they didn't like a humid environment. Never had air conditioning, which I now feel is a must if you have electrostats. Also, originally left them on all the time which in my situation probably wasn't a good ides.
Did build up a center channel "woofer" using extra Quad parts.
Charles
Pain in the butt was removing all the old glue from the rim of the "egg crate".
Had an application problem with the "special sauce" he provided to add conductivity to the mylar. A number of times it seemed to "clot" and not provide an even coating. May have been the humid conditions I was working under.
Anyway, found that most the rebuilt panels didn't last too long - eventually gave up and sold the speakers. Even rebuilt panels from W Piquet only lasted a few years.
Did find that they didn't like a humid environment. Never had air conditioning, which I now feel is a must if you have electrostats. Also, originally left them on all the time which in my situation probably wasn't a good ides.
Did build up a center channel "woofer" using extra Quad parts.
Charles
Last edited:
That's a bit dispiriting @stellavox. I am hoping for this repair to last for several years.
When you say they didn't last too long, what was it that failed? The stator to egg crate bond, the mylar to egg crate bond or did the mylar itself tear or burn? If it was the last one I'm wondering if using 6 micron mylar and losing 1-2dB in the top octave might be a worthwhile compromise.
When you say they didn't last too long, what was it that failed? The stator to egg crate bond, the mylar to egg crate bond or did the mylar itself tear or burn? If it was the last one I'm wondering if using 6 micron mylar and losing 1-2dB in the top octave might be a worthwhile compromise.
I had a thought today. One could use 6 micron mylar for the bass panels and gain benefits like ease of handling and extra strength because the those panels are producing almost nothing at 10kHz anyway. Comments?
Hi hop333,
It happens that I just completed the restoration of my very early ESL63s (serial 6603/4) with ER Audio repair kit (no affiliation). They were in a very sorry state, many stators separated from their matrix and some teared diaphragms due to arcing.
All I have to say is that although it has been a long and tedious work I am just amazed with the results. Scrupulously following the very detailed instructions provided with the kit made the restoration process a trouble free do it yourself job.
As I thought it might help others, I made a photo album starting from speakers disassembly to complete reassembly, and two short videos right after first speaker was completed :
- Quad ESL 63 reassembled - YouTube
- Quad ESL 63 - YouTube
After thorough thinking on how to re-glue the front stators to their matrixes without touching the intact diaphragms, I had the idea to use fishing line that I delicately sneaked through the stator holes. That allowed me, after tensioning the line, to hold the stators firmly against the matrix during the 24 hours glue curing time. That saved me the hassle of redoing good diaphragms.
After a few days for the coating to cure, flash rate of the neon bulb indicating how often the system needs to be "recharged" reveals an ideal situation : whereas a normal rate is about 1 flash every 30 seconds for a brand new ESL63, or in perfect condition, mine flashes every 50 seconds 🙂 This means that there's not the smallest EHT leak anywhere. Sound is just as fanstastic as those ESL are able to deliver, truly amazing.
Other than that, I replaced the original 220µF electrolytic cap and 1,5R wirewound resistors with MKP and ceramic non inductive resistors from Bourns. In the improvement department, rather than redoing the dust covers with shrinkable polypropylene, I used the same 3µ mylar as used for the diaphragms restoration. This improves high frequencies transparency thanks to lesser mass and inertia.
While I was at it, I also replaced all of the foams Quad used everywhere in the speaker structure, all of them had fallen into dust.
Feel free to ask questions, I understand how helpless one can feel before starting this job, I have been there. Honestly, if you are patient, have some time on your hands and carefully follow instructions provided by Rob you will definitely succeed.
I hope this helps.
It happens that I just completed the restoration of my very early ESL63s (serial 6603/4) with ER Audio repair kit (no affiliation). They were in a very sorry state, many stators separated from their matrix and some teared diaphragms due to arcing.
All I have to say is that although it has been a long and tedious work I am just amazed with the results. Scrupulously following the very detailed instructions provided with the kit made the restoration process a trouble free do it yourself job.
As I thought it might help others, I made a photo album starting from speakers disassembly to complete reassembly, and two short videos right after first speaker was completed :
- Quad ESL 63 reassembled - YouTube
- Quad ESL 63 - YouTube
After thorough thinking on how to re-glue the front stators to their matrixes without touching the intact diaphragms, I had the idea to use fishing line that I delicately sneaked through the stator holes. That allowed me, after tensioning the line, to hold the stators firmly against the matrix during the 24 hours glue curing time. That saved me the hassle of redoing good diaphragms.
After a few days for the coating to cure, flash rate of the neon bulb indicating how often the system needs to be "recharged" reveals an ideal situation : whereas a normal rate is about 1 flash every 30 seconds for a brand new ESL63, or in perfect condition, mine flashes every 50 seconds 🙂 This means that there's not the smallest EHT leak anywhere. Sound is just as fanstastic as those ESL are able to deliver, truly amazing.
Other than that, I replaced the original 220µF electrolytic cap and 1,5R wirewound resistors with MKP and ceramic non inductive resistors from Bourns. In the improvement department, rather than redoing the dust covers with shrinkable polypropylene, I used the same 3µ mylar as used for the diaphragms restoration. This improves high frequencies transparency thanks to lesser mass and inertia.
While I was at it, I also replaced all of the foams Quad used everywhere in the speaker structure, all of them had fallen into dust.
Feel free to ask questions, I understand how helpless one can feel before starting this job, I have been there. Honestly, if you are patient, have some time on your hands and carefully follow instructions provided by Rob you will definitely succeed.
I hope this helps.
Last edited:
Thanks very much @Thmartin. That gives me courage! Mine are even older and in some cases the nylon damping material is even separating from the rear stator but I think this will be easy to fix. I may PM you if I have any questions. Actually, I have one question already: what resonant frequency did you aim for with the panels and how did you achieve it?
I have decided to completely rebuild all the panels anyway, even though 3 out of the 4 that I have taken apart so far have had intact diaphragms. Amazing that such delicate things are still perfect! Shame about the stator-to-matrix bonds though... 🙁 Apart from wanting to make everything consistently new, I believe that the coating used by ER Audio gives a significantly higher resistance than the original Quad and so should allow less charge migration and hence lower bass distortion.
When I first got these in the early 1990s I was working in the loudspeaker industry and had access to very large plastic film capacitors, so I have always used them with 2x 100 microfarad + 1x 20 microfarad as the input capacitor. At the time, I was convinced that the sound was better than the original Alcap bipolar electrolytic. The other modification that I think is worthwhile is a more open and damped grille material - perhaps a strong polypropylene plastic or a plastic coated metal?
I have decided to completely rebuild all the panels anyway, even though 3 out of the 4 that I have taken apart so far have had intact diaphragms. Amazing that such delicate things are still perfect! Shame about the stator-to-matrix bonds though... 🙁 Apart from wanting to make everything consistently new, I believe that the coating used by ER Audio gives a significantly higher resistance than the original Quad and so should allow less charge migration and hence lower bass distortion.
When I first got these in the early 1990s I was working in the loudspeaker industry and had access to very large plastic film capacitors, so I have always used them with 2x 100 microfarad + 1x 20 microfarad as the input capacitor. At the time, I was convinced that the sound was better than the original Alcap bipolar electrolytic. The other modification that I think is worthwhile is a more open and damped grille material - perhaps a strong polypropylene plastic or a plastic coated metal?
I didn't aim at anything specific other than restore them identical to new condition. The 3µ mylar has to be tensioned evenly at 2.6KgF (a dynamometer is supplied in the kit) and that's it. I can tell you one thing : the amount of bass these speakers are able to deliver without audible distorsion is something I wasn't expecting.
As for the capacitors I walk away from so called "audiophile" products, whose only difference with normal capacitors is a ridiculous price. I choosed these (available from Mouser EZP-V60117MTC Panasonic | Mouser France) because :
- they are MKP, which I wanted
- their price is reasonable, although not cheap since you need to buy four items
- and because they are a perfect fit to available space in the speaker base without any modification. 57,5mm is the exact width you want. Measure twice, cut once.
I have seen so many "surgeries" done in butchered ESL63s to install soda can sized audiophile capacitors, hilarious
Last edited: