Adjusting ELS hybrids by changing the voltage

I read the Final 0.3 modification list here, and got interested in adjusting the sound by purchasing an adjustable power adapter.

In the Final -case someone reduced the voltage from 12v->8v. I on the would like to do the opposite, and boost the electrostatic by increasing the voltage.

- Is it safe?
- How does the voltage affect the output - how big dB change can I expect per +1V ?
 

Tim Keenan

Member
2006-05-11 10:28 am
Final 0.3 hybrid esl

Hi ljpp, I've just bought a pair of second hand Final 0.3s and unlike all other posters here I find the panel o/p sublime and the cone unit awful when driven by a Bel Canto eVo2i.II (strangely they sounded fine on demo being driven by a Naim Nait! Which may just show that you can't hear properly on demo, or may show that Naims are the ideal amp for these). There is clearly too much energy going to the cone as it flaps and runs into its end stops on Jah Wobble's 'Mu' and sounds flabby and one-note on orchestral music. So how did your experiment go? Did you turn up the psu supply?

Tim
 
Voltage

Hi,

I don´t think thats a good idea. Every designer tries to have as much voltage for theyre panels as possible. The reason for this is, that the driving force on the diaphragm is app. proportional to the voltage and the voltage (i.e fieldstrength..is responsible for SPL. It was discovered however that SPL and voltage are not proportional (as formulas should indicate) but nearly constant in the useful voltage range.
When You rise the voltage from 0V on You get a rise in SPL that is proportional to the voltage but at a certain point (around 600V/mm) the SPL flattens out and there is nearly any rise in SPL till the point of flashover begins (1.6...2kV/mm) . Now voltage can change a bit simpley by room temperature, humidity and so on. So its good use to set the operating point at the highest voltage, so that there is no considerable loss in SPL when its humid for example.

Soundwise a high voltage is good too. Lowering voltage leads to a softer sound with darker timbre. Reaction on impulses sounds rounded. Reducing voltage below the ´knee-point´ of the SPL/voltage-curve will lead to a treble loss. So fudging with voltage will just provide for max. +-1dB within the useful range.

There are two good solutions to this problem.
a) take the highest possible voltage and construct a new bass
b) take the highest voltage and construct a active crossover

jauu
Calvin

ps: btw. talking about voltages doesn´t mean 8-12V but rather 1-5kV!!! You have to look at the panels voltage not the powersupplies´s! try to figure out which input voltage of the PS gives which figures of output voltage at the panel itself.
 

Tim Keenan

Member
2006-05-11 10:28 am
Final 0.3 bass output

Hi Calvin,

Thanks for that. I kind of suspected that would be the case, but it's not my field. I went for the opposite solution. I put a 15 ohm resistor in series with each positive bass o/p terminal. This has now stopped the overexcitation and made them listenable again. And boy do they have potential. The 15 ohm choice was no choice because it was the only suitable pair I had available, but it's made a helluva difference and a great starting point for futher experimentation. on first listening I think I will need 10 ohm for orchestral and 20 or more for dance CDs. It's by no means the ideal solution. I'd rather alter the values inside the xover. Any ideas guys? What does that 10ohm resistor do?

Tim
 
R

Hi,

the resistor does several things.
It forms a voltage divider together with the impedance of the crossover network of the bass and the bass itself.
It reduces current flow through the bass hence SPL of the bass
It rises the Q-factor of the bass a bit.

A value of 15Ohms is very high. If You think that now the sounds ok for You, I assume the whole concept of the speaker to be problematic, or You might have severe probs with our room acoustic, or somenthing inside the speaker is not working properly!!??

jauu
Calvin
 

Tim Keenan

Member
2006-05-11 10:28 am
Final 0.3 bass output

Hi Calvin,

Yes, I agree with you, it is high and of course it is not the solution. The point is the speakers are now listenable to. Before the cone o/p was overpowering. Actually I've now found a pair of 8 Ohm resistors and this is better, but I think if I was going with this solution I'd settle for a value nearer 10 Ohms. Unfortunately, this solution is also acting more like a bas tone control. With Jah Wobble (Mu) the bass o/p is still too much, for orchestral (Shostakovich) it is about right, but for a string Quartet (Schoenberg) it is too thin.

I would also agree with you that this result suggests there is something wromg with the speakers, but both speakers produce identical results, so if there is a fault it is common to both, which seems unusual. There is also the possibilty that they have been modified, but this does not appear to be the case. (I bought them from Richard Lord the founder of REL subwoofers, and it is possible I suppose that they have been tweaked for him by the importer without his knowledge - he himself appears to have little esl engineering knowledge.) This leaves th epossibilty that they are not being driven properly by the amplifier.

I have tried three amplifiers and none have driven these speakers properly. My Papworth 50 W push-pull valve amp just grunts and farts, my Aura 50 W transistor amp does a better job but runs out of steam quite early, and finally a Bel Canto eVo 2 Gen 2, which is advertised at 200 W pc into 4 ohms, trips its protection circuits on moderately loud transients in the Shostakovich (symphony 8, 3rd movement). I'm going to try them with a Krell Kav-300il soon. Interestingly this amp has been pronounced an ideal match to the Finals in an American review, so we'll see.

We can rule out the room acoustic - cones flapping and hitting their end stops has nothing to do with the room. (Also I've dusted and hoovered the panels, so it's not dust, etc.)

Finally, thanks for the xover info. Might increasing the value of that 10 ohm resistor be a way to go?
 

Tim Keenan

Member
2006-05-11 10:28 am
Final 0.3 bass output

Postscript:

6.8 Ohms is sounding a fair compromise, but it rules out listening to dance music at 'normal' levels. Orchestral and piano are sounding rich without one-note blare in the bass. An LP of Yes, 'Close to the Edge' sounds analytical, though the strong bass guitar line is occasionally blurred and spread. However, reducing bass further by using a higer value resistor makes the mid-range hard and thin.

Equipment:
Dynavector 10x4
Project RPM 9X with SE speed box
Bottlehead Seduction phono amp
Marantz CD17 KI
Bel Canto eVo 2 Gen 2 amp
All cables by Chord
 
Amps

Hi,

if i may give a recommendation regarding amps.

I made good experience with Rotels.....980, 990-series.
They are very affordable and show construction details especially around the powersupply which even much more expensive amps often miss. They work relatively good into the complex load an ESL forms.
A imo even superior and still very affordable solution can be some digi-amp. I refer to these babies here: www.hifiakademie.de
They play absolutely stable, because by principle they are not effected by complex loads as classical Class AB feedback-amps are. And they sound very much better than their price tag might suggest! Imho the best You can do drive an ESL properly. Using a stereo module with optional crossover, You could even think of activating Your Hybrid-ESL, solving loudness problems of the drivers elegantely :)

jauu
Calvin
 

Tim Keenan

Member
2006-05-11 10:28 am
Final 0.3 bass output

Hi Calvin,

Yes an electronic xover would be an elegant solution. Maybe one day I'll get the time to research and build one. I had a look at that website and the kits look very nice, though not cheap. However, my one experience with digital amps is not that promising. The Bel Canto struggles to power the Finals and sometimes trips its protection circuits.

So, I'm looking again at the Final xover. Is there any way of either reducing bass output or increasing panel output by changing component values?

Tim
 
recomm

Hi,

well, You know cheap means cheap and not priceworthy!
The amps of HiFiAkademie work flawlessly and on a very high sonic niveau. The parts used are of high quality and the double sided circuit board is excellent. You must keep in mind, that the price for a complete Kit include a two channel amp plus power supply o one board. Too the support is very good and friendly. If You need different output filters for Your speaker, this won´t be a prob at all. IcePower of B&O e.g. specify their amps only for load impedances greater than 4Ohms, which might give probs when You want to drive an ESL!
Compare the HA´s to the UcD-modules for example or other digi-amps and You´ll find out that those won´t be cheaper at all and(!) You may even save on money by building that amp all by Yourself. Imo the HiFiAkademie-amp is superior to other Digi-amps, because it causes no trouble, runs stable and sounds good!

jauu
Calvin
 

Tim Keenan

Member
2006-05-11 10:28 am
hybrid ESL balance

TimA,

Well there's still life in this thread then! Since originally posting I discovered that both panels were significantly down in o/p. Not having much experience of ESLs I didn't realise that they'll go on outputting even with significant degradation of film. I've just fitted brand new rebuilt panels and the balance problem immediately reversed - too much panel o/p. I corrected that by reducing the AC o/p of the supplied psus from near 15V! to much nearer the spec of 12V. I've also replaced the HF caps with higher quality polyprop ones. They now sound pretty good (dreadful with cheap speaker wire), though I'm planning upgrades of internal wiring, maybe externalling the crossover (and upgrading the power trannie which is underspecced and liable to overheating), and improving bass-bin damping. They could be keepers as they seem to be superbly transparent mid range up, though as a consequence very inforgiving of poorly miked recordings and badly mastered transfers.

I'd welcome any comments on readers' experience of these interesting speakers.

Tim
 

TimA

Member
2002-11-17 5:39 pm
UK
Tim,

Glad to hear you got it sorted. Did you by any chance do the rebuild yourself? I have been asked to re-diaphragm one belonging to a friend and wonder if there may be complications. I have fitted new diaphragms to a number of Quad 57s and 63s so ought to be able to do this, however the Final is an "unknown" as I haven't opened it yet. I'd welcome any tips you may have!

TimA
 

Tim Keenan

Member
2006-05-11 10:28 am
Calvin,

The panels are brand new, Final made a new batch recently, they cost me £400. They take 50-100 hours to break in, and in my experience are very sensitive to speaker cable. I initially used what I had around, a twisted TNT CAT 5 job (http://www.tnt-audio.com/clinica/diycables.html), and it was horrible (though had sounded good with cones and ribbons). I then switched to Chord Rumour 4 (bi-wire) I bought off ebay and the improvement was gobsmacking.

Do you (or anyone) know about uprating the tiny power transformer used in the Final 0.3? I haven't researched a replacement yet, but if you have any suggestions I'd be grateful for the info.

Tim
 

TimA

Member
2002-11-17 5:39 pm
UK
Tim,

Thanks for the link. I'll post the results of my rebuild here.

Incidentally, Final appear to have a made a considerable change to the step-up transformer at some point during production, from a large heavy one (S/N 2000 30617), to a much smaller and lighter encapsulated one (S.N 2000 30684).

Perhaps we can narrow down the point at which the change was made with the more serial numbers from other forum members who own 0.3s. I have attached a picture of the heavier (earlier) one.

Tim.
 

Attachments

  • final 0.3 xover.jpg
    final 0.3 xover.jpg
    92.1 KB · Views: 208

TimA

Member
2002-11-17 5:39 pm
UK
That's the same trannie as mine.

What serial number, please?


I assume you are asking about the cross-over values.

The bass section consists of a 2.2mH inductor followed by a conjugate filter which consists of a 1.25Ohm (IIRC) resistor and a 220uF cap (I have reworked this part of the cross-over on a different pair for a more even response up to the cross-over point).

The treble section consists of 33uF (bipolar) bypassed with 1uF, seen in the front right of the pic.

TimA.
 

TimA

Member
2002-11-17 5:39 pm
UK
I have disassembled a faulty panel. The mylar appears to be 12 microns thick and is affixed with double-sided tape. This all comes off fairly easily leaving a nice clean surface on which to attached new film.

As for the electrical connections, the "charging ring" consists of a thin piece of wire going round the perimeter of the panel and stopping at the top of each side. The electrical connections to the stators are not easily accessible however, probably best left well alone.

The two foam strips running the down the middle will probably need to be replaced as they don't want to come off in one piece. A better alternative to these can probably be found in any case.

TimA.