ESL-57, Trouble Shooting
This article is for persons seeking assistance in ESL-57 troubleshooting. The first part of this memo outlines how an ESL-57 panel sounds and behaves when it is damaged. The second part explains how to test the ESL-57 to determine which panels are damaged.
NEVER touch any panel while either the high-tension supply or the amplifier is switched on. Take particular care to disconnect the amplifier. The ESL 57 multiplies the voltage output of the amplifier one-hundred-fold, from about 50 volts to about 5,000 volts. Consequently, both the high-tension supply and the amplifier deliver about the same high-voltage charge to the panels. However, while the high-tension charge is limited to several micro-amperes, the amplifier can deliver several amperes. Quad speakers contain fatal currents and voltages. Consequently, take particular care to switch off and to disconnect the amplifier.
Identifying the Problem
The following simple guidelines should help you to work out what your ESL problem is.
Quad ESL-57 bass panels have their conductive coating on the outside of the panel. Consequently bass panels are extremely difficult to arc, and are rarely damaged. If the bass panel is not working, or plays extremely quietly, it is probably because the EHT (high-voltage) rectifier unit is defective. If you want to have it repaired, send only the EHT unit itself, and not the entire metal frame and transformer assembly. I offer more details about EHT repair and replacement options later in this memo. Often, there is corrosion-effluorescence on the panel around the rivet. It is purely on the conductive paint on the surface of the panel, and has no impact on the panel's performance. See more later in this memo.
Quad ESL-57 treble panels have their conductive coating on the inside of the panel. Consequently they are easily arced.
- If the treble panel is not working at all, it is probably badly arced, and requires a rebuild
- If the treble panel works, but there is a buzzing sound
- if there are also sparks, then there is arcing damage and the panel requires a rebuild
- if there are no sparks, then the dustcover has probably lost its tension or is torn and must be replaced.
If you are not familiar with Quad ESL-57's, bear in mind that the speaker was designed in the mid-1950's to work with 15-watt amplifiers, and has no protective circuitry. Given this, an amplifier putting out as little as 30 watts may arc the treble panel. Even the Quad 303 amplifier can arc an ESL-57, although it was designed specifically for it. You will be surprised at how loudly an ESL-57 can play, but take care to play within the speaker's limits.
If you do decide to play loudly, do it at night, in darkness, so that you will be able to see if arcing is occuring. At the point of arcing small flashes of blue light come from various points on the treble panel, and the music acquires a rough edge (like a blues harmonica), as if the amplifier is overloading. Turn down the volume, or else you may damage the panel. The speakers play purely right up until the point of arcing, so the transition is abrupt. You can find the speaker's limit by gradually raising the volume until you see the point at which arcing begins. The speaker will probably be damaged if you raise the volume beyond this limit, or if you play near this limit for a long time.
The diaphragm on an ESL-57 bass panel is charged to 6,000 volts, while the diaphragm on an ESL-57 treble panel is charged to 1,500 volts. This would suggest that it is far easier to arc a bass panel than a treble panel, but in fact the opposite is true.
Bass Panel Arcing
ESL-57 bass panels rarely suffer arcing damage. There are several reasons for this. Firstly the bass panel's conductive coating is on the outside of the stator panel, secondly the stator-to-diaphragm gap is quite large, thirdly the stator panel is made of fairly thick plastic (which adds to the distance between the diaphragm and the bass panel's conductive coating), and fourthly the panel's diaphragm is made of fairly thick Saran.
Treble Panel Arcing
Unlike ESL-57 bass panels, ESL-57 treble panels arc relatively easily. Even Quad's own 303 amp, which was designed for the ESL-57, is capable of arcing the ESL-57's treble panel. The reasons why ESL-57 treble panels arc relatively easily are firstly that the treble panel's conductive coating is on the inside of the stator panel, and secondly that the stator-to-diaphragm gap is quite small. This permits the diaphragm to actually contact the conductive coating on the interior of the treble panel stators, or to approach close enough that an arc will jump between the diaphragm and the stator.
If the arc is minor, it will burn a small hole in the diaphragm. The hole will be the same size as the hole in the stator panel, about 25mm across. The arc will cauterize the edges of the hole, and the diaphragm will be as strong as ever, and function as well as ever, so long as the user is prudent. Obviously, if the diaphragm accumulates many minor arcs, their eventual cumulative effect will be to weaken the diaphragm and predispose it to major failure. Major failure can also occur if too powerful an amplifier is used. The diaphragm will be driven into close contact with the stator panel over a large area, thereby creating many individual arc burns simultaneously.
Whether major damage occurs immediately or over time, the way it occurs is generally the same. Individual arc burns will accumulate in a single location on the diaphragm, and weaken it at that point. This part of the diaphragm will displace more easily into a stator panel, which will melt the individual arc burns into a single elongated arc with ragged edges. The ragged edges of the diaphragm will easily contact the stator panel, and will arc even at low volumes.
What Arcing Sounds Like
If you slowly raise the volume to the point where the diaphragm is about to arc, the sound will acquire a rough edge like an over-amplified blues harmonica. There is evident audible distortion. If you listen in darkness, you will see small flashes of blue light coming from various points on the treble panel. These initial arcs are due to burning of charge-carrying particles in the diaphragm. At this stage, when the charge-carrying particles burn, they do not yet burn holes in the diaphragm. If you are prudent, you will now turn down the volume.
If you continue to raise the volume, you will see that at music peaks there will be crackling explosive sounds from the diaphragm, accompanied by sparks. The music will be badly distorted, and the crackling sounds will be loud enough to drown out the music. At this stage, holes are being burned in the diaphragm. If you turn down the volume immediately, your diaphragm may only have one or two holes, and the speaker will remain workable.
If you do not turn down the volume, the fascinating crackling explosive sounds and pretty sparks will continue until virtually every part of the diaphragm able to contact a stator panel has melted. The speaker will eventually fall silent - permanently.
If you do turn down the volume, but do it too late, you will subsequently experience crackling explosive sounds and sparks even at modest volumes, because almost any amplifier input will drive the diaphragm's ragged edges into contact with the stator panels. When a badly damaged speaker is plugged in but not in use, it may hum or it may be silent. It depends whether the damaged diaphragm sticks to either of the two stator panels or whether it oscillates between them.
What an Arced Panel Looks Like
When the diaphragm of an ESL-57 has an arc-burn hole in it, there is also a corresponding arc-burn on the stator panel. This leaves a black sooty deposit on the stator panel (similar to candle-smoke). The deposit is primarily on the interior of the panel, but it is also visible on the edge of the hole in the stator panel where the burn occurred. Treble panels with minor damage usually have only a single arc-burn. If you hold the panel up to the light and then move it, you will see that one stator panel hole is blackened vis- a -vis all the others. You will notice that you can see light reflecting off the diaphragm through the holes in the stator panel, but if you look closely at the hole where the arc-burn is, you will see no reflection from the diaphragm through that hole, because the diaphragm has been burnt away there.
Unless they have been torn, bass panel dustcovers rarely need to be replaced. However, the dustcover film Quad uses on the treble panel loses its tension over time, and the film eventually sags onto the convex front of the panel, and then buzzes whenever music is played. Quad bass and treble panel dustcovers behave differently because they use different material. To check whether the dustcover is the problem, simply remove the front grill, check the tension of the dustcover, and play music to determine where the buzzing comes from. At the same time, you'll be able to see whether there are any arc burns on the treble panel. If the problem is just a torn or buzzing dustcover, I suggest you fix it yourself. Unless you are a perfectionist, it's not worth the time and money to send panels to me just to have dustcovers changed. I have seen speakers which had been used with torn dustcovers for long periods of time, and was surprised to find that nonetheless the panels were in excellent condition. If they are not in a dusty environment, the panels seem to remain quite clean. Even if there is dust accumulation inside the panel, you may be able to remedy it by removing the dustcovers and vacuuming the "naked" panel. Don't worry that you may damage the diaphragm, it's pretty sturdy, and is sufficiently tensioned that no vacuum cleaner is likely to suck it up against the stator panel and rip it (if you actually do rip it, it's because it was already damaged). Just make sure that you disconnect the speaker and let it completely discharge, so that any dust stuck on the diaphragm loses its static attraction to the diaphragm and can be sucked off.
Occasionally, you will see what appears to be rust on the panels. It is typically on the bass panel, in the vicinity of the panel's rivets. My understanding is that this is not rust but rather effluorescence. This is a benign chemical reaction between the rivet, the panel and the paint, which occurs as chemicals leach out of the paint over time. The panel's rivet heads, each carrying a 6,000 volt charge, project from the surface of the stator panel. The charge each rivet is carrying is apparently sufficient, over time, to stimulate a chemical reaction which does not occur elsewhere on the bass panel, or anywhere on the treble panel.
Effluorescence does not affect the integrity of the panel itself. The reaction is on the surface of the panel, while the panel interior and diaphragm remain in pristine state. I found it impossible to remove the effluorescence from my own panels without causing damage, and so left the panels as they were. Because rivets holding the two stator panels together pierce the charged diaphragm, Quad had to mask both bass and treble stator panels so that there is no conductive paint in the vicinity of any rivet, so as to avoid a short-circuit. Consequently, the area around each rivet is inert as far as producing sound is concerned.
The specified voltage is 1,500 V DC for the treble panel, and 6,000 V DC for the bass panels. Quad high-voltage supplies tend to weaken after about twenty years. Low volume is a sign that the EHT supply is defective. To confirm, you have to test with a high-voltage probe. Take note however, that a general-purpose high-voltage probe will not give the correct reading for the bass output of the EHT unit. The reason is that the probe has an impedance of 1 gigOhm, while an impedance of at least 10 gig Ohm is required if the probe is not to load the circuit. This loading the circuit is not a problem under normal circumstances, as high voltages are typically associated with currents of several amps or more. Relative to a current flow of several amps, the drain of the probe is insignificant. It is only when the current is almost nonexistent (as is the case in an EHT unit), that the current drain of the probe becomes significant, to the point that it even corrupts the reading. In other words, when the probe tries to take a reading of the EHT's voltage output, it does so by demanding a greater current-flow than the EHT unit can deliver, and so actually reduces the voltage it is trying to read. If an Enquist EHT unit gives a reading of 5,400 volts on a standard high-voltage probe and a Quad unit gives a reading of 4,900 volts, that corresponds to a true 6,000 volt output.
Pictures of the EHT supplies mentioned above