The Quad thread

[swt61]

Dammit Dinny! If word gets out that bigger is not necessarily always better, my already fading sex life is going to take an even greater hit. Man up and start singing the praises of the 12s!

[Dusty Chalk]

The 12L actives do, indeed, rock.

[deepak]

Any thoughts on the ESL-988? Sound quality, reliability, etc?

[spritzer]

Get some rebuilt ESL63's, same speaker but a rebuilt unit should be far better.

[deepak]

Get some rebuilt ESL63's, same speaker but a rebuilt unit should be far better.

Curse you

It's going to take a while to find some. I might get some MMGs for the time being while I wait for a nice ESL63.

[spritzer]

I aim to please... If you get a good deal on ESL988's then go for it but they are the same speaker as the 63...

[Craig Sawyers]

Completed the last part of the ESL57 odyssey. This involved collecting two remanufactured and matched treble panels and a bass panel re-dust-filming kit from One Thing Audio. Ron gave me a one hour practical demo of how to do the dust cover, and how to solder to the rivetted on tags to the stators without melting the stator plastic.

One "trick" is to use spray on carpet glue to stick the film to the wooden frame, and trim the edges flush with a soldering iron (quicker and easier than a craft knife). Then on the frames that have the tag board, first shrink the bottom few inches only and apply a strip of their thick edge banding tape as strengthening of the bottom inch or so of the film.

Preparing the naked bass panels involves checking them for electrical leakage. The trick is to feed 6kV from the supply via a neon flasher - 10meg series resistor, then a neon in parallel with a 47nF cap. A flash every 3 seconds or more indicates a good panel (I calculate that means a leakage resistance between the film and frame of >10G-ohms). There are various possible leakage paths - either a conductive track has formed from the stator metallisation to one of the many rivets. The rivets all connect to the film, and are at 6kV whereas the stator is at 0V when there is no signal - so tracks can slowly form from points of high potential gradient, like the sharp edges of the rivets. Mine only needed vacuuming and then a wipe over with a pad moistened with iso propyl alcohol to remove surface grot - at which point one flash every 6 seconds or so.

After taping on the dust cover frames and heat shrinking the film, all four bass panels gave one flash every 10 seconds or so (so 30G-ohm approx leakage R).

After complete speaker assembly minus rear cover, and as a final check, the neon circuit was used again. One flash every 10 seconds on both speakers.

The consequence of new treble panels and refurbished bass panels is that they charge up stonkingly quick. I left them for half and hour, and they were at full sensitivity - I could actually have put music through them before that. The reason that people on the web reckon that 24 hours is necessary I suspect is because of bass panel leakage (what One Thing term a "lazy panel"). The Quad manual actually reckons only 30 seconds is necessary, but they advise that 1 minute might be necessary to get full performance - for a new or completely refurbished and checked speaker of course.

One interesting thing, that Ron at One Thing warned about with refurbished treble panels, is that they sound a whole lot brighter. Reason is that the sensitivity of treble panels falls as the decades tick by, mainly through leakage - so old ESL57's always sound treble shy if they have old, original and not burnt out treble panels.

Anyway, imaging is needle sharp now, and balanced (ie not skewed towards one speaker).

All this was a 2-day operation, the majority of which was spent on the bass panels.

Craig

[postjack]

Craig, I love reading the details of your Quad journey, even if I don't understand all of it. Thanks for sharing.

As I said earlier in this thread, Kent swapped out a resistor when he repaired my speaker that he says will help the speaker last longer but will also lower the volume. Therefore the repaired speaker is quieter then the stock speaker. He sent some very detailed instructions and a new resistor for me to try to install in the stock Quad. I even went and bought an iron from Radio Shack and opened the speaker up to attempt the resistor swap, but I just ended up laughing at myself when I could understand even the most basic instructions (apparently a "jumper" isn't the same on a Quad ESL-63 as on a CD-ROM drive). So for fear of electrocuting myself or destroying the speaker, I closed it back up and am just using the balance controls. This is a bit of a bummer, since it takes away some from the midrange purity of the speakers, which is what I initially found so amazing about them.

Having said that the midrange is still pretty magical, just not AS magical. I had some amazing experiences this week listening to King Crimson's "Red" and "Starless and Bible Black". Those are some outstanding sounding records that the Quads just love.

[Craig Sawyers]

As I said earlier in this thread, Kent swapped out a resistor when he repaired my speaker that he says will help the speaker last longer but will also lower the volume.

That is interesting. I've been peering at the ESL63 schematic, and can't see where an obvious resistor mod could be carried out. More details please?

But what scrutinising the schematic did show is that the neon flasher is permanently installed in the ESL63 HT charging circuit - same values as One Thing suggested, which indicates that is where they got the idea (they also rebuild ESL63 panels). The service manual indicates that anything significantly faster than one flash per second is bad news.

[Craig Sawyers]

Correction to the leakage resistance of a panel in relation to neon flash rate.

The important thing is that in the time between flashes, the panel self-discharges. When the difference between the panel voltage and the charging voltage exceeds the difference between the striking and sustaining voltage of the neon, the neon fires and recharges the panel. A typical panel neon has a difference of 30V between strike and sustain.

During the droop time, the panel is essentially constant current discharging into the leakage resistance. Since the voltage droop is 30V in 3 seconds for a good panel (and longer for an excellent one), that enables the leakage resistance to be directly calculated.

That gives a rather astonishing 3000G-ohms. The 47nF cap in parallel is there to ensure there is enough stored energy stored to actually strike the neon.

That implies a CR time constant of 600 seconds. Note that this discharge rate means that the panels hold their charge for a very long time indeed - to get down to 50V takes 48 minutes. Which is why I guess that Quad recommended about an hour after switch off before servicing.

[spritzer]

As I said earlier in this thread, Kent swapped out a resistor when he repaired my speaker that he says will help the speaker last longer but will also lower the volume. Therefore the repaired speaker is quieter then the stock speaker. He sent some very detailed instructions and a new resistor for me to try to install in the stock Quad. I even went and bought an iron from Radio Shack and opened the speaker up to attempt the resistor swap, but I just ended up laughing at myself when I could understand even the most basic instructions (apparently a "jumper" isn't the same on a Quad ESL-63 as on a CD-ROM drive). So for fear of electrocuting myself or destroying the speaker, I closed it back up and am just using the balance controls. This is a bit of a bummer, since it takes away some from the midrange purity of the speakers, which is what I initially found so amazing about them.

Having said that the midrange is still pretty magical, just not AS magical. I had some amazing experiences this week listening to King Crimson's "Red" and "Starless and Bible Black". Those are some outstanding sounding records that the Quads just love.

You could post the instructions here along with some internal pics and we can perhaps guide you along...

Correction to the leakage resistance of a panel in relation to neon flash rate.

The important thing is that in the time between flashes, the panel self-discharges. When the difference between the panel voltage and the charging voltage exceeds the difference between the striking and sustaining voltage of the neon, the neon fires and recharges the panel. A typical panel neon has a difference of 30V between strike and sustain.

During the droop time, the panel is essentially constant current discharging into the leakage resistance. Since the voltage droop is 30V in 3 seconds for a good panel (and longer for an excellent one), that enables the leakage resistance to be directly calculated.

That gives a rather astonishing 3000G-ohms. The 47nF cap in parallel is there to ensure there is enough stored energy stored to actually strike the neon.

That implies a CR time constant of 600 seconds. Note that this discharge rate means that the panels hold their charge for a very long time indeed - to get down to 50V takes 48 minutes. Which is why I guess that Quad recommended about an hour after switch off before servicing.

Thanks for sharing that and the Quad design is just too cool. It's a shame PJW retired and didn't carry on like his contemporary in Japan.

[Craig Sawyers]

Thanks for sharing that and the Quad design is just too cool. It's a shame PJW retired and didn't carry on like his contemporary in Japan.

The really neat thing is that this is a trivially simple circuit for measuring resistances in the T-ohm range using a watch to measure the time between flashes . Costs pennies, or built out of the spares box. Not as accurate as a proper meter, but around five or six orders of magnitude cheaper!

The problem with PJW retiring was that he handed the reins to his idiot son Ross. Ross has a degree in economics and doesn't have a technical bone in his body. Under his stewardship he piloted Quad into financial ruin. In Kessler's book about Quad, he interviewed Ross Walker, and it is clear that he blamed the engineering team for telling him porkies about the maturity of the new remote handset controlled range - and on that basis, like a fool he launched the product. He missed the point entirely that it was his responsibility to know - and if he had done what all good small company CEO's do and walk around the shop floor on a daily basis he would have known what was going on. It was what Hewlett (of HP) used to call "management by walking around". PJW for sure would have known precisely what was happening on an hourly basis.

[spritzer]

Yeah, any CEO should know exactly what is going on at any time let alone when you are going to release something which can make or break the company. Now if PJW had just stayed on for a couple of years and finished those sphere speakers...

[Duggeh]

PJW, as I understand it, was not in the robust health that his Japanese contemporaries were in at his age.

[Hens]

Going to go listen to some quad esl 57's today, should be great

[postjack]

That is interesting. I've been peering at the ESL63 schematic, and can't see where an obvious resistor mod could be carried out. More details please?

Kent included a pic of the PCB, circling the resistor that needs to be replaced. He also sent me the necessary resistor. I used a camera to take a picture of his pics, so they are not super clear but hopefully this helps:

Also, here is exactly what Kent said regarding the resistor in an email to me before he proceeded with repairs:

At some point the EHT power supply board has been replaced with a Quad

produced board. I imagine the other speaker has had the board replaced as

well. The EHT board/supply work correctly but these replacement boards

produced by Quad have the EHT bias voltage set at maximum. This is a much

higher bias voltage than was originally delivered with the speakers.

Quad raised the bias voltages over the years so the speakers would play

'louder', i.e. more sensitivity. Early production speakers were about

10-15% lower in bias voltage. Raising the voltage does make the speaker

play louder but at the price of decreased element life. I do not think

this is a good performance trade off and routinely set bias voltage at the

original factory spec.

I perform this change on all Quads in my shop at no charge and would like

to lower the bias voltage on your speaker as well. This will present a

problem in that your existing speaker will play a bit louder than this

speaker. You would need to adjust your preamp balance control to make up

the difference.

I keep records on each speaker I service, i.e. bias voltage levels,

sensitivity, etc. If the other speaker is sent in for service at some

point I can make the same bias voltage adjustment so you would have a

matched pair again.

Or I can send the necessary resistor and instructions for lowering the

voltage on your speaker if you can do the work or know someone that can.

It is easy just replacing a single resistor.

Lowering the bias voltage and turning off the speakers when not in use

will yield the longest panel life. And they should be used in a dry

environment, i.e. humidity levels no higher than 50%.

Edited May 23, 2010 by postjack

[postjack]

You could post the instructions here along with some internal pics and we can perhaps guide you along...

a very kind offer, I'll type up all his instructions soon, but really this is the one I am having the most trouble with:

3. Drain the stored charge in the EHT supply by placing one end of the jumper to ground and the other end to the orange wire connected to the EHT supply board. Wait a few moments for the supply to drain.

Basically I don't know what a jumper is in this context, but I know this step has something to do with me not electrocuting myself so I want to understand it completely before proceeding.

[cetoole]

Basically short the output of the bias supply to ground, to discharge any voltage stored in the capacitors. Do this with the power disconnected, obviously.

[Dusty Chalk]

And while not grounded yourself, nor touching anything else for that matter. In fact, probably safest to just don a body condom.

[spritzer]

In electronics, the term jumper is often used to mean a short length of wire. For instance if you are going to run a balanced amp with RCA inputs you have to install a jumper in the XLR input to short pin 1 and pin 3. The jumper is just a short length of wire.

While the EHT supply stores very little energy the high voltage means you would get a nasty shock from it should you come into contact with the circuit even after the speaker has been turned off. The fix does make perfect sense and shouldn't be too hard with a soldering iron and some soldering wick/desoldering pump to clear off the old solder from the holes in the board.

[Hens]

So I ended up buying the quad 57's I say today. Quite a nice sound, I feel though that it lacks a bit of resolution and rolls off a bit early...

Can someone explain to me what the little jumper on the back above the power LED's are? is that a voltage selecter switch?

[Craig Sawyers]

Kent included a pic of the PCB, circling the resistor that needs to be replaced. He also sent me the necessary resistor. I used a camera to take a picture of his pics, so they are not super clear but hopefully this helps:

Also, here is exactly what Kent said regarding the resistor in an email to me before he proceeded with repairs:

That is indeed interesting. Issue 1 and Issue 2 EHT boards seem to be identical, with the final Issue being changed. I guess that what Kent has done is replace the 56k resistor with the originally fitted value of 180k.

15% on EHT voltage should change the sensitivity by around 1.2dB. It is odd why Quad did this, because the loudness is a function of the electrode spacing only - over a level defined by that the film hits the stator and burns out (hence the many protection circuits in there). So changing the EHT only changes the sensitivity, not how loud they will play.

But Kent is certainly right that a higher voltage may well limit lifetime. These sort of things tend to go as a high power law, like fourth or thereabouts - so increasing the voltage by 1.15 may well reduce the lifetime by close to a factor of two.

Is there anyone close to 'jack who is skilled with a hot stick and can do the resistor retrofit? If I wasn't 6000 miles away, I'd voluteer

[Craig Sawyers]

So I ended up buying the quad 57's I say today. Quite a nice sound, I feel though that it lacks a bit of resolution and rolls off a bit early...

Can someone explain to me what the little jumper on the back above the power LED's are? is that a voltage selecter switch?

Well done! You might have a bit of work to do to get them back on song, but it is seriously worth it. A properly sorted pair are most definitely not treble shy.

Check my posts for the odyssey of fixing mine, which now sound absolutely awesome. I suspect from your description that one or both of your treble panels is burnt out - most are. This not only clobbers the high frequency response, it also slugs the EHT voltage to the bass panels too - so the bass is quiet as well. There are plenty of places that will supply replacement treble panels if that is indeed the problem. You can easily check this - take the rear grille off (screws around the edge and under the plinth). Now lift up the felt that is behind the centre treble panel - you will easily see if it is burnt out (carbon charring, perforated dust cover etc).

If you do replace the treble panels, be sure to retrofit the protection clamp board that prevents too much voltage getting to the treble panels and burning them out.

The jumper is a voltage setting - select whatever the nominal local mains voltage is (110V, 240V etc).

Let us know how you get on.

[Hens]

My bass is definately there, lol, I think they are in quite good nick all things considered, when I get some time though, Ill take the cover off and check it out.

I am lucky, there is a well respected quad reconditioner in Perth, worst comes to worst I'll let him have a look at it see what he thinks.

All in all I am quite happy, I managed to pick them up for a very reasonable amount of money (cheaper than a 2nd hand pair of o2s lol...)

[n3rdling]

The treble is extremely beamy. Point the panels directly at your ears in normal seating position and see if that fixes your treble roll off problem.