Advanced Repair Notes


The Quad ESL 57 operates with internal lethal voltages. DO NOT attempt any servicing work unless you are totally confident about what you are doing.

After many years repairing the ESL 57 I have decided to make as many of the little tips and tricks that I have picked up along the way available to everyone. You probably wont find these written anywhere else - they are just little things that I look for when repairing speakers. I have done so many that now I know what to look for.

I have broken this page down to these sections :

Testing a treble panel prior to fitting a new diaphragm

So, picture the scene - you have stripped a treble panel down so you now have 2 stators and 2 dust covers. What do you check before you fit a new diaphragm ? It may sound obvious but you ought to check that the audio and high voltage are actually going into the panel and that they are ending up where you want them. There is an immediate problem trying to do this. The inside of the panel is painted with grey paint and its not obvious where the high voltage and audio signal are located. To get round the grey paint problem I use pointed probes on my DVM. These are pointed enough to be able to go through the paint and get to the carbon paint that is located underneath (the audio signal is carried by carbon paint on the inside surface of the stator where all the holes have been cut, the high voltage is carried round the periphery of each cavity). You need to test all 3 sections of each panel to make sure the audio signal is getting onto the inside surface of each section of both stators. Expect to see some resistance. The carbon paint is acting like a large very flat resistor so its not surprising. The high voltage is a little more difficult to test. The high voltage comes into the panel on 2 rivets per stator. It is then sent to 4 vertical strips. The position of these vertical strips are easily seen as they have white tape (masking tape to dampen any noise if theres a gap between the stators) stuck over them. However, at the top and bottom of the stator between the tape strips there are areas that do not have tape over them. They actually connect the inner 2 strips to the outer 2 strips. These can be tested in the same way as the audio - just make sure that both rivet connections are checked. Finally when you have checked all electrical connections into each stator, re-insulate anywhere that you have gone through the grey paint (I use clear nail varnish).

Coating the diaphragm

Coating the diaphragm is the single most important thing that needs to be done correctly. Otherwise some of the faults that I will explain below will happen. As you will see elsewhere on this website, I use as close to the original coating that Quad used. This is soluble nylon. This information is widely available. What is not clear is in what proportions did Quad make their coating ? From my experience of doing many treble panels, if you get the proportions wrong the following will happen :

1. Too little dissolved nylon - this will cause a quiet panel as it will not charge up properly. Also as there is so little nylon on the diaphragm it may cause the diaphragm to come unstuck from the stator. As an experiment, I coated a piece of diaphragm with pure Methanol (pure in the sense that it has come straight out the bottle and did not have any nylon in it). It left a white deposit on the diaphragm. This is probably impurities in the alcohol but it leaves a coating which comes off easily (stick some sticky tape to it and pull !!) and obvioulsy wont charge up.

2. Too much dissolved nylon - this will cause an unstable diaphragm. In my early days of panel repair I used powdered graphite. This is a very conductive coating. When voltage is applied to it the charge builds up very quickly. I imagine the charge as a cloud of electrons sitting on the surface of the diaphragm. As the coating is low in resistance the cloud will be large. Some of this charge from this large cloud of electrons will be induced into the carbon paint that usually carries the audio signal. Now, if you remember electrostatic theory it states that like charges repel. So you have a massive cloud of electrons on the diaphragm and the same electrons being induced onto the carbon paint that carries the audio. The innevitable will happen. One stator will have more induced charge than the other and the diaphragm will be pushed hard onto the inside face of the opposite side. Even though the carbon paint that carries the audio is effectively earthed through the audio transformer it still manages to push the diaphragm. This is a problem I have seen many times in my early days and the result is a treble panel that sounds horribly bright. It still makes noise but as I said its very bright (distorted ??) and much louder than it should be. So if it does this with graphite coatings, does it happen with nylon ? Well you'd better believe it does. Too much nylon will cause exactly the same symptoms and the only cure is another diaphragm. The other ways to see if you have an unstable diaphragm is to look carefully inside the panel with a torch (when its connected and charged). Does it look like the diaphragm is closer to one stator than the other ? The final symptom of an unstable diaphragm is to connect the panel to an audio transformer and power supply, leave it charging up for a few hours (maybe even 24 hours), turn the mains off and see if you get any strange noises out of the panel. If, after a few minutes you hear a pop (and maybe a few more) this is the diaphragm returning to its original position inside the middle of the panel. To speed things up when I do this I kill the high voltage quickly by using a resistor and earthing the high voltage. If I hear any pops from the panel I know its a dud (this pop should not be confused with the sound that the spark generates if you kill the high voltage with a resistor).

3. Just the right amount of dissolved nylon - This section wouldnt be complete without stating what I use. I use Elvamide 8023R in a 4% solution in Methanol. In other words 4 gramms of Elvamide in 100 mls of Methanol (I know that any alcohol has a lower specific gravity then water but hey, its all relative). I have found this to be a reliable proportion. Too high like 7% I find I get the problems detailed in point 2 above. Its far too easy to apply too much coating and get an unstable diaphragm. So I do the 4% mix. When applying the coating I coat both sides of the diaphragm and I start at opposite ends. So one side I coat from top to bottom, the other side I do bottom to top. I always coat the diaphragm when its on the horizontal so I dont get any runs in the coating. I start at opposite ends because I get a more even coating this way. I use a ball of cotton wool dipped in the coating. Innevitably it will apply more coating per side when you start coating as its wetter than when you finish. This means when you start applying the coating it will have more applied. To even this up, on the other side of the diaphragm I start at the opposite end. Its a simple trick but one that I find gives consistent results.

Testing the new diaphragm

So, you are now in the position of having coated the diaphragm and assembled it back into the panel. When I do this I leave it all still suspended in the diaphragm stretching frame and begin testing the panel. Really I only do 2 tests at this point. They are usefull nonetheless. Firstly, I connect the high voltage but leave all the audio connections free. I make sure none of the audio cables are touching anything or each other and I turn the high voltage on. I leave the panel for a short time to charge up (say 1 minute) and I carefully measure to see if there's any voltage on the audio cables. Now, as noted above you should expect to see some voltage as the charge on the diaphragm will be inducing a voltage on the carbon paint that carries the audio. However what should happen is you should see an initally high voltage (I usually see 100 volts on the brown wires and twice this on the blue wires) but it should die back to near zero after a few seconds. If it doesnt you either have a short from the high voltage to an audio connection somewhere inside the panel (unlikely) or there is too much coating on the diaphragm (more likely). If you dont see any voltage at all even upon initial connection the chances are you have no charge on the diaphragm due to a lack of coating and therefore it wont work. Either way you may be in for problems. Secondly, I connect the audio wires. I dont connect all the wires at the same time. First I do the brown wires, then disconnect them and do the blue wires. I do this to make sure that both sets of wires are producing sound from the panel. Lastly I connect them all up just to be sure.

Audio transformer testing

By this point you may have succesfuly refurbished a treble panel, fitted it back in the speaker, wired everything up only to find its not sounding right. Either it crackles or its overbright compared to the other speaker or something else that you just cant put your finger on.

An often overlooked part of the speaker to check is the audio transformer and its crossover. All thats needed to check it is a volt meter that can measure resistance. This bit has been written assuming the newer style transformer is being measured although the same holds true for earlier types (just that the pin numbers are not correct). Checking the transformer is simply a matter of measuring the resistance accross all the resistors and capacitors. No wires need to be disconnected as it should all be open circuit. So, refering to the pin numbers in the service manual the following should be seen :

Pin number Pin number Reading Notes
1 6 180Kohms Resistor check
2 7 180Kohms Resistor check
3 8 272Kohms Resistor check - this should be 2 or 3K higher than pins 8 to 13 as its measuring the resistor and the windings. If this only measures a few Kohms one or both the 560pf capacitors between pins 3 and 8 has shorted (usually because of arcing).
4 9 180Kohms Resistor check
5 10 180Kohms Resistor check
6 11 180Kohms Resistor check
7 12 150Kohms Resistor check. If this only measures a few Kohms the 560pf capacitor between pins 7 and 12 has shorted (usually because of arcing).
8 13 270Kohms Resistor check
9 14 150Kohms Resistor check
10 15 180Kohms Resistor check
3 13 1.7 to 2.0Kohms Windings check
7 13 1.7 to 2.0Kohms Windings check
1 2 1.7 to 2.0Kohms Windings check
4 5 1.7 to 2.0Kohms Windings check

Its a good idea when checking the resistors to try to hold the probes in one hand and just exert a little force on the side of the resistor with the other. This is checking for resistors with hairline cracks that will cause problems. If the measured resistance changes then the chances are you have a damaged resistor. Also it is highly unlikely that any of the resistors will measure what they should be. Its not unusual to find they are some way out. For example, the 180k resistors quite often measure 210k. I dont see this as a problem as the other speaker most likely measures the same. I would not advocate changing resistors just because they are some way off what they should be.

NO web creating software was used in the production of these pages !!!
Its more efficient that way.

This site is optimised for Netscape V2.0