I'm Looking to Upgrade My Current Sound System

[Gordon]

reply to #52 and 53: have a look at this book - ISBN: 978-0240515786 - edited by John Borwick especially the chapter by Martin Colloms that goes into this issue in some depth. It blows quite a few HiFi myths up and also presents some salutary advice. A warning though, the book is expensive - even on Kindle!!

If you Google the book you can get some extracts to read including Colloms' Chapter;

http://books.google.co.uk/books?id=lPd2OucRfu8C&pg=PT298&lpg=PT298&dq=dBA+loudspeaker&source=bl&ots=fY7sMjbNJg&sig=xRjNHJicsW0GhlPWqjAcufwdUww&hl=en&sa=X&ei=JkjhUZnPGsaZ0AWakYGgDw&ved=0CFMQ6AEwBA

PS: a distorting/clipping amplifier produces harmonics which tilts the power spectrum considerably to the HF and so challenges the tweeter whose power rating is typically quite low on the assumption that the spectrum up there is sparse. Moral: Don't let your amp distort. How do you know it isn't clipping transient peaks [read the book]?

[Dave2002]

PS: a distorting/clipping amplifier produces harmonics which tilts the power spectrum considerably to the HF and so challenges the tweeter whose power rating is typically quite low on the assumption that the spectrum up there is sparse. Moral: Don't let your amp distort. How do you know it isn't clipping transient peaks [read the book]?

Thanks Gordon,

In fact your PS also restates something which I have believed to be true, which goes against the peak vs average argument. Clipping is likely to produce very significant amounts of HF, and if, as you say, tweeters are not designed to cope with that, it's very plausible that they will be overloaded. Allegedly it is much harder to overload speakers with powerful amplifiers than low power ones.

[Gordon]

Allegedly it is much harder to overload speakers with powerful amplifiers than low power ones.

Quite so. The danger with a big amp that avoids clipping by having huge headroom is that one is tempted to drive the level hard and so load the speaker with a high level of average power as well as peak. That overheats the drivers and fatigues them faster. Also the transients that arise if one drops the stylus, or there is a bit of dirt, in the groove with full volume isn't a good thing ether!! All but a few % of the driving power ends up dissipated as heat in the speaker wiring. Speaker design is as much about getting that heat out as the sound.

The issue with moving coil speakers is not the voltage drive [this embodies the fidelity of the audio signal] being transferred with high precision to the sound pressure output it's the corresponding instantaneous current that has to be supplied by the amp and it's usually not in phase with the voltage. By Ohm's Law any drop in the speaker impedance from a nominal say 8 ohms doesn't improve matters. All this stresses the output stage devices and also places demands on the power supply and can upset the amp stability.

[jayne lee wilson]

Thanks Gordon,

In fact your PS also restates something which I have believed to be true, which goes against the peak vs average argument. Clipping is likely to produce very significant amounts of HF, and if, as you say, tweeters are not designed to cope with that, it's very plausible that they will be overloaded. Allegedly it is much harder to overload speakers with powerful amplifiers than low power ones.

Hmm... when I blew up the (right channel, where the brass were!) QED P300/Kelly KT2(95db), the volume setting was lower than it would usually have been for the Wharfedale 505s(87db), it wasn't being overdriven. There was, very briefly, overt distortion on that Beethoven 7 climax before it blew (no remote, my cat-startling dive to reduce the level - too late!). Given the HFN analysis, it seemed highly likely that the amp simply couldn't deliver enough into lower impedances to cope. Yes, it's fair enough to say that highpowered amps might be driven too hard for a speaker to cope, but despite running 200 watt amps into 87db Wharfedales and later 86db Harbeths for many years (nominal power handling 125/150) I never seem to have approached this point (despite playing mainly large-scale orchestral music). But surely clipping IS the limits of available power being reached? Which can happen suddenly if the musical peak is beyond the amp's reserves. And as I said, you tend to find more peak current available in the higherpowered amps anyway; and the two nominally 200 watt models I've used both nearly doubled their power into lower impedances.

The longer-term overload of an amp run at consistently high volume will not usually be a problem for classical listeners, as there's so much dynamic relaxation in the music; which is why I concentrated on the current ability earlier.

[Dave2002]

Hmm... when I blew up the (right channel, where the brass were!) QED P300/Kelly KT2(95db), the volume setting was lower than it would usually have been for the Wharfedale 505s(87db), it wasn't being overdriven. There was, very briefly, overt distortion on that Beethoven 7 climax before it blew (no remote, my cat-startling dive to reduce the level - too late!). Given the HFN analysis, it seemed highly likely that the amp simply couldn't deliver enough into lower impedances to cope. Yes, it's fair enough to say that highpowered amps might be driven too hard for a speaker to cope, but despite running 200 watt amps into 87db Wharfedales and later 86db Harbeths for many years (nominal power handling 125/150) I never seem to have approached this point (despite playing mainly large-scale orchestral music). But surely clipping IS the limits of available power being reached? Which can happen suddenly if the musical peak is beyond the amp's reserves. And as I said, you tend to find more peak current available in the higherpowered amps anyway; and the two nominally 200 watt models I've used both nearly doubled their power into lower impedances.

The longer-term overload of an amp run at consistently high volume will not usually be a problem for classical listeners, as there's so much dynamic relaxation in the music; which is why I concentrated on the current ability earlier.

Sounds like an unfortunate event to write off to experience.

If, as you say, there was a significant reduction in the impedance of the speakers in the range under consideration, then currents could have been high. Re clipping having a limiting effect - well no - not exactly. What clipping does is generate a lot of high frequency noise, and actually moves the power into a range where the loudspeaker (in this case the tweeter) can't cope. There may have been other factors in your case though.

Seems a bit odd that you blew out the amp as well as the speaker. Presumably it didn't have any protection, or if it did, it didn't cut in fast enough.

[jayne lee wilson]

Sounds like an unfortunate event to write off to experience.

If, as you say, there was a significant reduction in the impedance of the speakers in the range under consideration, then currents could have been high. Re clipping having a limiting effect - well no - not exactly. What clipping does is generate a lot of high frequency noise, and actually moves the power into a range where the loudspeaker (in this case the tweeter) can't cope. There may have been other factors in your case though.

Seems a bit odd that you blew out the amp as well as the speaker. Presumably it didn't have any protection, or if it did, it didn't cut in fast enough.

But clipping occurs in both the power domain and the frequency domain - those distorted HF signals are created precisely because it's running out of power. A greater dynamic ability can prevent that - as I found out in practice, when I moved from amps with 15+ amps available (not enough) to those with 25 or 30 (more than enough).

As Wiki has it, "clipping is a form of waveform distortion that occurs when an amplifier is overdriven and attempts to deliver a voltage or current beyond its maximum output capability." It's only after that condition of overload is reached that HF distortion can follow. Which is why it is indeed easier to blow up a speaker with an underpowered amp rather than a cruising powerhouse. Excess is best!

http://en.Wikipedia.org/wiki/Clipping_(audio)

Oh, yes, when I rang the dealer (after a very uncomfortable night)...
I told my sad tale...

A broad Lancastrian accent which I hadn't encountered there before said. "Oh love, if you knew some of the things I've done, and I'm in the trade..." I could have kissed him - would have done if he'd been there!
The smartsuited guy who collected them said simply "only you know how loud you were playing them..." But they asked for no redress.
I was too meek (and too puzzled at the time) to even tell him that the volume was at a very low setting when it happened...

[Dave2002]

Bit of a tragedy really. I suspect something wasn't quite right with either the amp or the speaker, or perhaps the wiring.
If the volume setting was low this seems improbable. Presumably wasn't a problem with the CD or LP either? Didn't have atomic bomb sounds in the middle of the Beethoven trio did it? Have you played that track since?

Were the speakers bi-wired? I'm not sure what impact bi-wiring would have - but I doubt that it would have made things any worse.**

An aside comment or observation might be that some of such problems might be avoidable with active speakers, about which I know relatively little. I had assumed that most active speakers are sold on convenience factors, such as not having to couple them to a high power amp, and some newer ones may even manage to make do with only a power connection if they use wireless communications for the signal. However active speakers can also have active cross-over units, so that matching the different drive units might become easier. Instead of having to attenuate the signals to some drive units to match levels, the levels for the other units are raised. Presumably active speakers can be driven by relatively low power amps. Can they even work effectively with pre-amps?

OTOH it could be that I've got this wrong, and active speakers may not actually make things any better re matching/clipping etc. I don't know in practice, as I've never used any and don't know how they are designed or intended to be used. There may be significant differences between different makes/models of active speakers.

** It might have allowed more current to flow.
Note also that there is a difference between bi-wiring and bi-amping, the first arguably being pointless.

[Phileas]

An aside comment or observation might be that some of such problems might be avoidable with active speakers, about which I know relatively little.

There's a section on active speakers in the John Borwick book linked to in Gordon's post above:

http://books.google.co.uk/books?id=lPd2OucRfu8C&pg=PT298&lpg=PT298&dq=dBA+loudspeaker&source=bl&ots=fY7sMjbNJg&sig=xRjNHJicsW0GhlPWqjAcufwdUww&hl=en&sa=X&ei=JkjhUZnPGsaZ0AWakYGgDw&ved=0CFMQ6AEwBA

Among the many advantages of active speakers is the fact that a clipping woofer amp can't damage the tweeter because it's not connected to it.

Active speakers are becoming more popular now and several companies make "domestically acceptable" models (as opposed to the "industrial" designs common in recording studios). I have a couple of pairs from AVI and I can highly recommend them. :)

[Dave2002]

phileas

Borwick's book is now relatively old - which doesn't invalidate most of the theoretical and other ideas, but it may not give appropriate guidance for the current market for active speakers.

Do you drive your speakers from a conventional amp? Preamp? How do you do it?

PS: There's a review of some AVI speakers here - http://www.gramophone.co.uk/editoria...em-alternative
Looks like you have to have wire between the speakers. I thought maybe two mains plugs would do, but that doesn't seem to be how this equipment works.

[Phileas]

Dave

My AVI speakers include a DAC and (remote controlled) pre-amp in the left-hand/master and there's a signal cable running to the right-hand speaker. Most active speakers require a pre-amp and one signal cable to each speaker. There's also a power cable to each speaker.

Some "domestic" active speaker makers:
AVI, Dynaudio, ATC, PMC, Acoustic Energy (coming soon).

There are many "pro" active speaker makers including (off the top of my head):
Adam, ATC, PMC, Mackie, Yamaha, Fostex, Neumann...

[Beef Oven]

Dave

My AVI speakers include a DAC and (remote controlled) pre-amp in the left-hand/master and there's a signal cable running to the right-hand speaker. Most active speakers require a pre-amp and one signal cable to each speaker. There's also a power cable to each speaker.

Some "domestic" active speaker makers:
AVI, Dynaudio, ATC, PMC, Acoustic Energy (coming soon).

There are many "pro" active speaker makers including (off the top of my head):
Adam, ATC, PMC, Mackie, Yamaha, Fostex, Neumann...

I'm planning on getting a pair of Genelec active speakers when I return from my holiday in September.

http://www.genelec.com/home-audio/

[Phileas]

I'm planning on getting a pair of Genelec active speakers when I return from my holiday in September.

I forgot about Genelec. :)

[Gordon]

Hmm... when I blew up the (right channel, where the brass were!) QED P300/Kelly KT2(95db), the volume setting was lower than it would usually have been for the Wharfedale 505s(87db), it wasn't being overdriven. There was, very briefly, overt distortion on that Beethoven 7 climax before it blew (no remote, my cat-startling dive to reduce the level - too late!). Given the HFN analysis, it seemed highly likely that the amp simply couldn't deliver enough into lower impedances to cope.

Yes, it's fair enough to say that high powered amps might be driven too hard for a speaker to cope, but despite running 200 watt amps into 87db Wharfedales and later 86db Harbeths for many years (nominal power handling 125/150) I never seem to have approached this point (despite playing mainly large-scale orchestral music).

Hmm indeed. The sensitivities you quote are not unduly low so the demise of both speaker and amp channel – the one seemingly delivering the brass – is odd. There could have been incipient issues that pre-disposed this to happen but it seems unlikely but not impossible. The “overt distortion” is a clue – that distortion could have generated HF that contributed to the problem. Otherwise no obvious explanation. Unless there was a transient fault in the content supply? Again unlikely but not impossible.

But surely clipping IS the limits of available power being reached? Which can happen suddenly if the musical peak is beyond the amp's reserves. And as I said, you tend to find more peak current available in the higher powered amps anyway; and the two nominally 200 watt models I've used both nearly doubled their power into lower impedances.

That is part of the problem, yes. Power is the product of voltage and current and when there is a phase shift between them there needs to be some compensation for that involving an increased current. If the amp runs out of volts that is only half the story and it can do so for a number of reasons one being a prolonged but sustainable power demand [sustained organ pedals do this] that reduces the power rail voltages [ie the power supply sags] which renders the amp more susceptible to clipping when a large transient turns up. The situation is like positive feedback. An amp can also run out of current even if the volts are OK. Current limiting – the amp’s inability to deliver what the LS impedance requires – also causes a distortion as a consequence.

The LS impedance typically varies with frequency and it may dip to low values in part of the spectrum – but only there. That in itself will not be a problem because it only affects the narrow band of frequencies in that region.

High power amps do tend to produce max power into low loads like 4 ohms but usually this makes assumptions about the phase angle being sensible. Give an amp a high peak power demand where the current demand is near quadrature with the volts and there will be trouble – once again the power demand may reduce the rail voltages and even big smoothing capacitors will run out of charge if it is removed fast enough.. The output devices will be sorely tried and if a large current is delivered it might be large because the odd impedance demands it and the amp allows it. Electrostatics are an example of where their natural capacitance is presented to the amp and this kind of load is difficult for some amps.

Biwiring has the single benefit of putting more copper in the feed to the speaker which is why I use it [and because the speakers have the facility]. I can’ say that I hear any significant difference. The amp sees the same load but at least the low/mid range drive is separated from the top and the crossover MAY be slightly simpler.

The longer-term overload of an amp run at consistently high volume will not usually be a problem for classical listeners, as there's so much dynamic relaxation in the music; which is why I concentrated on the current ability earlier.

Agreed, which makes the demise of your amp/speaker all the more puzzling given your description of usage. However modern digital sources [and some analogue ones] have significant dynamic range and this sets the peak to mean requirement. This means that a transient is more likely to hit the clippers than before.

Active speakers avoid many of these issues or at least simplify the interfaces between the drivers and their amps. By subsuming the amps into the speaker the unit cost goes up of course but overall they are not that much more expensive. When the time comes to renew my system I will look seriously at active speakers and also look at integrating a DAC with USB and Wifi. One drawback is putting all one’s eggs into the same basket.

[Dave2002]

I forgot about Genelec. :)

From the review I linked to the AVIs seem quite good, and might be of interest to our OP. However it did seem to me that they are not so good on extreme bass. Arguably few speakers are, but I have heard some (traditional, and old designs - not necessarily "state of the art modern" - there were some good speakers 30 or 40 years ago!) which do/did a significantly better job than most with the low frequencies. Some traditional designs did manage to get down to around 30 Hz with a significant output. Doing that without boom or cabinet resonances is hard, and in any case is likely to require larger speakers - which might not fit in with room and aesthetic constraints.

The AVIs are fairly small I think, though they can be coupled with a sub woofer, which might improve things at the low end. I once heard some LS3/5A speakers coupled with a sub woofer. The subjective SQ went from very good to excellent when the extra bass unit kicked in. However, I know from experience with cheapish AV set ups that a badly integrated sub woofer can render the SQ significantly worse, and even with blockbuster movies, where this might be expected, the results can be undesirable and overwhelmingly horrible. On classical music such sub woofers could be a disaster if not carefully matched to the rest of the system.

[Dave2002]

Hmm indeed. The “overt distortion” is a clue – that distortion could have generated HF that contributed to the problem. Otherwise no obvious explanation. Unless there was a transient fault in the content supply? Again unlikely but not impossible.

Another "long shot" possibility is a loose wire in the speaker, which somehow managed to short the amp. It would give distortion before the catastrophe. It does seem very unlikely though. That at least could explain how both the speaker and the amp were wiped out.