Power amp - how much power do you *need*? [Archive]

Chippy_boy

14-10-2009, 10:47

I am considering upgrading my power amp (currently a Tag McLaren 100x5R) to something better.

The Tag is a 5 channel amp primarly aimed at the HT market but I am using it as a stereo setup with 4 channels driving 2 speakers bi-amped. It's rated at 120w/channel into 8 Ohms but delivers considerably more, and my speakers are nominally 6 Ohms. So I reckon I have about 400w/channel on tap.

In terms of a replacement, I have to say I am confused. On the one hand, I hear the likes of Musical Fidelity saying you need huge amounts of power to be able to realistically reproduce transients and musical crescendos. Similarly Chord and others produce amps with hundreds of watts per channel.

On the other hand I see valve amps around here with very low output and even reputable manufacturers such as Naim with outputs perhaps 1/4 of what I have at the moment.

So my questions are these:

(1) how many watts do I need for loud(ish) volume in a medium sized room (18 x 12) and 89db/w speaker?

(2) would I miss bi-amping, do you think? Does this offer a significant improvement? My speakers are Wilson Benesch Curves.

Thanks for your thoughts. I don't need ear splitting levels, but I do like a wide range of music from rock to classical and sometimes I do like to "turn the wick up" a bit. But the quality of the sound is by far the most important thing - it's no good going loud if it sounds rubbish!

twelvebears

14-10-2009, 11:10

Hi Chippy.

Oh my word, you really are opening a can of worms with a post like that aren't you?

Well aware of the MF view of how many watts you need, I'm a proud owner of a Nuvista M3 myself, however I REALLY wouldn't spend too much time worrying about it and would concentrate on the sound you're trying to achieve.

I've had massive power with inefficient speakers (my M3 driving ATC SCM20s at about 85db/w), but I've also had the same speakers driven with 'only' 100w from a Primare A30 integrated and still had excellent results.

Yes it does make a difference if you REALLY do want to hammer the SPLs a lot of the time, but most of the time it really won't.

Obviously you can't match a low output valve amp with difficult speakers, but Wilson's are cracking speakers which I imagine you want to keep? And I can't imagine that the Tag is short on power, so is your question because you're not happy with the sound or the volume levels?

Bi-amping can be beneficial, but then it depends on both the amp and speakers. I once owned a pair of Rotel RB980 power amps which I used in preference to an Audiolab setup in a bi-amp configuration. However I then found the amps worked better when running bridged and one speaker each, so go figure....

Personally, I'd say that 200w per speaker was plenty for pretty decent levels unless you are luck enough to own a concert hall for a house.

Perhaps it's a change of amp that you need to consider, but for the type of sound rather than the power?

I don't know how much Tag now differ from their Audiolab roots, but I'd expect the TAG amp to be quite analytical? i.e. very 'matter of fact' and possibly prone to hardening up when pushed? This was my experience with Audiolab power amps 'back in the day' and the reason I opted for the warmer Rotels.

This may be the same thing you're experiencing, but this isn't really about bi-amping or power, but the nature of the TAG amp itself.

Also, I believe the Wilson's tend to be quite 'warts 'n' all' too, which may be part of the problem, but not a reason to change them.

Out of curiosity, where are you based? Are you in the UK?

Chippy_boy

14-10-2009, 14:09

Hello mate - thanks for your thoughts.

Regards your question about why I am interested in changing, it's really just because I want it to sound "better". This isn't because of any particular shortcoming. In fact I fairly recently acquired the Wilson Benesch's and more recently still a Benchmark DAC1 HDR and I have to say the sound is completely wonderful and totally engrossing. I have never enjoyed listening to a hifi so much - it's totally addictive.

That being the case, maybe I should just leave it as it is and just enjoy the ride, but I am mindful of the fact that the Tag amp is perhaps the weakest part of the setup and I wondered just how much better it could be if the amplification was top notch as well. The power output from the Tag is seemingly limitless and it goes miles louder than I would ever need. Whether the sound becomes harsh towards the limit, I really can't say because I have never been near it. Certainly it's been fine at any volume I have tried.

I just wonder whether something like a Chord SPM 1200E would be a big step up. But then I am intrigued by things like the Avondale amps which are "only" 90w/channel and yet are talked of very highly. Do you think 90w would be enough? Or am I likely to think this is a big step backwards? Of course I could get 4 Avondale monoblocks and bi-amp them, but that would cost double and could be a complete waste of time. Or on the other hand it could cost be £4,000 and still might not be powerful enough and £4k is wasted.

I am really unsure what to do. I guess I need to borrow a high-quality, lower powered amp and see how I get on.

Cheers & thanks again

Chip

Barry

14-10-2009, 14:46

Chippy_boy

14-10-2009, 15:12

Hello Chip,

The theoretical answer to this is quite straightforward.

An orchestra going full belt, ie. fortissimo, produces a sound level 120dB above the threshold of hearing (10^-16 W/square cm). Your speakers have a sensitivity of 89dB/W, you therefore require 31dbW = 1.26kW of power!

In practice of course, this amount of power is required so infrequently that one can get away with considerably less. 200W/channel ought to be sufficient.

Regards

It's not that straightforward though, is it Barry.

1. the 120dB level you cite seems incredibly high. I have seen 109~112 as being more credible.

2. You have 2 speakers, therefore the sound output for 2 is +3dB higher than for 1. I think you omit that from your calcs.

3. Sensitivity of 89db/W is at 1m. At 5m, it's 14db lower.

So overall, your estimate is too low!

I am sure in the real world, 200w is sufficient. But that's really not my question. Is 90w sufficient? What about 30w?

Barry

14-10-2009, 18:43

Ok,

So we'll take a figure of 110dB above the threshold of hearing for orchestral tutti. Over two speakers that's 107dB from each speaker. Each speaker has a sensitivity of 89dB/W at one metre, thus you need 18dBW per speaker. At a distance of 5m, you'll need an extra 10*log(5^2) = 14dB of power, that is 32dBW = 1.58kW/channel.

Reducing the upper limit to 100dB, then the power requirement is 158W/channel. This is still pretty dammed loud and will not be required all the time, so 100W, or 90W will probably be quite sufficient.

I am satisfied with 50W/channel (deliberately limited from a capability of 100W/channel; to protect my speakers), as I don't play music loudly and deep bass is not important to me.

How loudly do you play music? Is it heavy metal, rock, opera or symphonic orchestral music? If your tastes are for smaller scale music: folk, small jazz ensemble or chamber music, then you can get away with far less. I expect that many here on AoS use amplifiers of around 25 - 30 watts.

Regards

chrism

14-10-2009, 18:56

Never even considered watts when selecting amps I have always judged them on sound quality and ability to make the speakers disappear. I suspect that selecting on watts alone may not help you get the sound you want.

I use an Avondale A260 and it goes louder than I will ever need into normal speaker loads. I like the "grunt" it can produce with deep bass notes and percussion as well.

Regards

Macca

14-10-2009, 19:49

IMHO its not just a question of watts>sensitivity>loudness - some amps will simply not drive some 'speakers in the bass regardless of 'spec' figures. I have 89db sensitive speakers also - but a £600 50W per channel Rotel integrated did not manage to do it. (Yes, I was surprised too :doh:)

Its replacement, an identically rated ( in terms of watts) Linn Lk100 does it with no bother at all.

Martin

jandl100

15-10-2009, 07:17

Hmmm ... yeah, well, there's more to this than amp power and speaker sensitivity.

I reckon you can calculate dB's and Watts 'til the cows come home but it still won't tell you how well or how loudly any given amp will drive any given speakers.

I think it comes down to the details of the speaker load and how it interacts with the details of an amp's power supply and circuit.

I've played a 89dB/W speaker (Kharma Ceramique 3) witha 6wpc T-amp at 'leasebusting levels' on a SciFi blockbuster DVD with no problems, but a Beard 30wpc integrated valve amp ran out of juice way before the alien mothership had even entered orbit! :)

And my old Cary valve monoblocks at 15wpc easily (and amazingly) drove my current Infinity ribbon hybrids far better, with more apparent power and bass tautness and slam than a Krell 400xi with 400wpc into the nominal 4 ohm load. :scratch:

At the end of the day, you just have to try it and see!
.... but feel free to carry on calculating dB's if it pleases you. :lol:

Chippy_boy

15-10-2009, 07:46

.... but feel free to carry on calculating dB's if it pleases you. :lol:

It gives me no pleasure at all. And in fact I didn't start it.

The "problem" is simply that there are perhaps 100 different amps I could possibly choose and I can audition maybe 5. I have to start somewhere in homing in on possible candidates and it would have been useful to know what sort of power range I should be aiming at.

For example, if I could rule out anything with less than (say) 100w as being unlikely to be suitable, and anything more than (say) 300w to be of no sonic benefit, that would have helped.

One of the issues is, without a spl meter it's very difficult to know what my "normal" listening volume is and how many watts my amp is pushing out. The volume I listen at on my amp varies hugely according to the CD in question (given the massive difference in recorded levels on commercial CD's) but I would say in general it's around -20 dB on the volume control. This was with a DAC with a nominal 2V maximum output and my amps need 1V input for 200W of output.

If my calcs are right (and I do get confused), that would mean a peaks at 2V input would equate to 800w output (assuming the amps could produce it). If I am -20db down from that, that shoud mean peaks at 8w/channel.

Seems incredibly low, but maybe that is right?

@Chrism "I suspect that selecting on watts alone may not help you get the sound you want."

Well of course. No-one would be stupid enough to buy an amp because they calculated they needed 143w/channel and amp XYZ has that power output. I only care about what it sounds like and will it go loud enough without starting to sound strained, or god forbid - clipping.

twelvebears

15-10-2009, 07:50

Hi Chippy.

To be honest, from the way you describe the sound, you seem to be worrying about a problem that doesn't actually exist. I'd recommend you just kick-back, relax and enjoy the music.....

If, on the other hand, you've got some cash burning a hole, and you're just thinking 'what next?', rather than 'I'm not quite happy with that' (which seems to be the case), then your decision over what to 'fix' next is a bit more difficult. I mean you're assuming your TAG amp is the weak link. Why? Because it's a coming up for 10 years old and it was designed as an AV amp? I wouldn't be so quick to point the finger is the sound is so great. You're next improvement could come somewhere else and the TAG could be a great match for your speakers.

If you are thinking amps however, why not try the opportunity to give something with a tubes a try and see how it fits. If it doesn't work for you and your speakers and solid state is more your bag, then I'd have a shot at Chord. They're powerful, look fab and don't sound too bad either. ;)

technobear

15-10-2009, 07:53

Chippy_Boy, you are sick! :)

You are suffering from audiophilia nervosa :mental:

Your TAG amp is a very good one, home theatre or not.

Bi-amped it is even better.

You will need to spend many thousands of pounds to get something that "sounds better" (BTW you haven't defined what you mean by that :scratch:)

You don't need a better amp. Spend the money on music!

jandl100

15-10-2009, 07:59

Oh dear. You come across as a little offended in your reply, Chippy Boy. :( My somewhat light-hearted response was intended to be informative, my apologies are hereby tendered if I caused offense.

But I really don't think that you can predict in advance what an amp will do into a particular speaker load. Unless someone here has specific experience of how your speakers, which I have heard may not be that easy a load, respond to particular amps any advice will be necessarily generic. What are the speaker maker's recommendations?

As you say, there must be at least 100 or more contenders. Probably more. Too much choice these days!

I really do think that, as 12 Bears suggests, you should try a few things and see /hear how they do for you and your choice in music and musical presentation, including some valve contenders. Where are you located? If you are near me, I'd be only too happy to come round with an amp or three and try them out to see. :)

John

15-10-2009, 08:00

The best advice as always is to hear a few amps in your system and trust your ears
Try a mixiture of valves and transitors
I think firat narrow it done around costs and then try a few that fit within your price range

Chippy_boy

15-10-2009, 08:33

Thanks boys for your considered comments.

(Jerry - no offense taken at all, and certainly no need for any apology from you. I was just meaning that all this talk of dB's is making my head spin!!!! and certainly not a route I am enjoying going down.)

I guess the only thing to do is to try some options and see how I get on. But I do take twelvebears and others' points: if it aint broke, why fix it. I certainly don't sit back and think ("that doesn't sound right"; "if only it had xyz"; "I don't like abc") etc. It sounds really wonderful as-is, to my highly untrained ears.

My desire to upgrade is really just that I have got the bug again and wondered if the sound could be even better for perhaps not too much outlay. I am very intrigued by the Avondale amps and a couple of M130's minus the sale of my 100x5R wouldn't cost a fortune. But if it's not an improvement, why bother.

Cheers all

Chippy

@Jerry: forgot to answer your question about my speakers. The speakers are 3 way with just a single blocking capacitor on the tweeter and the mid range driver is directly connected to the amp - no filtering at all! So the crossovers are incredibly simple and therefore no horrible phase angles nor impedance dips. Minimum impedance is never lower than 4 Ohms and sensintivity is 89db/w, so they should be a doddle to drive. WB don't give any minimum amplifier recommendations; they merely say peak power handling 200w unclipped and maximum spl of 109 dB.

Barry

15-10-2009, 08:40

Chippy,

If you type "ATC SCM20" into Google, you will find that ATC themselves state that a suitable amplifier to drive the speakers will have a power output in the range 50 - 300W.

The active version of the SCM20 has built in amplifiers with a power capability of 200W for the LF and 50W for the HF.

Looking at reviews written by users, the amplifiers being used mostly have outputs 100 -200W/channel. One user drives his speakers with a Bryston 14B SST amplifier; having an output of 600W into 8 ohm and 900W into 4 ohm.

Trust this is of some help.

Regards

Chippy_boy

15-10-2009, 08:55

jandl100

15-10-2009, 09:10

Thanks Barry, but I am confused. Why are you citing the ATC's? I have Wilson Benesch Curves. Or do you just mean this as an illustration?

Thanks again

I think bdh's brain is like Elvis - and has temporarily left the building! :eyebrows:

Barry

15-10-2009, 10:59

Oops!

My apologies - I misread your initial post and got it muddled up with Twelvebears' reply, which mentions ATC SCM20 speakers. (My excuse is I had a very late night with little sleep!)

OK, Wilson Benesch Curves:

http://www.wilson-benesch.com/curve/curve.html

Looks like 200W/channel is the most you will need.

Trust this is of some help; now I really will leave the building and look for my brain! :mental:

Regards

Barry

15-10-2009, 11:11

Chippy_boy

15-10-2009, 11:24

At the risk of doing the unthinkable, that is, refer you to another forum; a similar question to yours is discussed inter alia on the ZeroGain forum:

http://www.zerogain.com/forum/showthread.php?t=20161

Might give you some amplifiers to think about.

Regards

Thanks Barry.

An interesting read, but I swear to god, I expect 80% of the people posting on there have never heard the kit on which they pass their learned judgement.

The perceived wisdom is that the WB Curves are bright and analytical, and everyone jumps on that bandwaggon. But this is far from the truth. When I first got them (and they were already run in) they sounded so dull and flat I thought I had made a huge mistake. It's only after you get used to the smooth sound that you realise that all the detail is there and you realise just how fabulous they are.

But, no. One magazine says they are analytical (I doubt that reviewer had heard them either) and then a bunch of idiots start spouting the same. Makes me wonder why I read any of this tripe in the first place.

twelvebears

15-10-2009, 11:57

On the subject of amp power to speaker matching, you are far more likely to kill a speaker with a small amp pushed too far, than by using an overkill amp on small speakers.

I've seen a nasty 30w solid state amp fry hefty speakers, and I've personally used 400w amps with budget stand-mounts without issues. All that happens is they go louder and louder completely cleanly until there's a 'clack, clack' noise of the drivers hitting their end stops, but never with any damage.

Electrostatics can be a bit more touchy however.

Rare Bird

15-10-2009, 12:32

You know i've never ever ever had a problem with power amps/speaker mismatching.Saying that i've never played music loud in all the time i've been into hi-fi, i don't understand for the life of me why people want hundreds of watts of power in a house, true it is wise to pick a more powerful amp over underpowered in some instances, it's not a PA System remember it's an household hifi...I'd rather opt for highish efficiency speakers with transistor amps, tube amps for speakers with difficult loads.Never believed in high power tube amps either.20 ish watts is ample

Saying this i'm not really an hi-fi man as such, just have an unhealthy obsession with the period late '60's/ early '70's hence my fettish for vintage audio around this period..I have dabbled with so called high end modern equipment etc, but it's not for me, i find the whole hobby can obscure your love & devoted time spent with what an hifi system was in the first place invented for! Lifes too short & money is something you should spend wisely regardless of any claims that you have plenty.

Themis

15-10-2009, 13:56

On the subject of amp power to speaker matching, you are far more likely to kill a speaker with a small amp pushed too far,
DC output, no ?

Chippy_boy

15-10-2009, 14:05

DC output, no ?

Not really.

When the speaker cone is moving normally, most of the energy is converted to sound waves and dissipated away.

When you look at the output waveform for a clipped amp, it's a smooth sine wave with the tops and bottoms flattened off. The flat (clipped) part of the output dumps all the electrical energy as heat in the speakers' voice coils. So they overheat and burn out.

Also the point in the wave where it starts to go flat is a nasty sharp angle. This means lots of high order harmonics. At high volumes this means much more energy than normal going to the tweeters, so they burn out too.

anthonyTD

15-10-2009, 14:07

On the subject of amp power to speaker matching,
you are far more likely to kill a speaker with a small amp pushed too far, than by using an overkill amp on small speakers.

I've seen a nasty 30w solid state amp fry hefty speakers, and I've personally used 400w amps with budget stand-mounts without issues. All that happens is they go louder and louder completely cleanly until there's a 'clack, clack' noise of the drivers hitting their end stops, but never with any damage.

Electrostatics can be a bit more touchy however.
totally agree!!:)
A...

Themis

15-10-2009, 14:08

Not really.

When the speaker cone is moving normally, most of the energy is converted to sound waves and dissipated away.

When you look at the output waveform for a clipped amp, it's a smooth sine wave with the tops and bottoms flattened off. The flat (clipped) part of the output dumps all the electrical energy as heat in the speakers' voice coils. So they overheat and burn out.

Also the point in the wave where it starts to go flat is a nasty sharp angle. This means lots of high order harmonics. At high volumes this means much more energy than normal going to the tweeters, so they burn out too.
Thanks for explaining. :)

Chippy_boy

15-10-2009, 15:06

Thanks for explaining. :)

No problem. I have often heard it said that when a speaker clips, the lack of airflow over the voicecoil causes it to heat up and burn out. The lack of airflow part is true, but it's not the reason.

When the cone is stationary, all of the electrical energy has nowhere to go other than into the voicecoil as heat. The heat output in the voicecoil can be *miles* higher with a clipped signal.

Barry

15-10-2009, 17:34

smithy

15-10-2009, 18:13

Ali Tait

15-10-2009, 18:41

Agree 100% with all that you have said about speaker and tweeter damage caused by an under-powered amplifier running into overload and clipping.

It has also been said, for reasons I don't understand, that a lower powered valve amplifier can produce a similar perceived loudness level as that of a higher powered solid state amplifier. Perhaps Antony TD or Ali Tait can explain?

Regards

Well I think Anthony will know much more than me,but in my limited understanding,this is for two main reasons-

1.Valve amps generally cope better than solid-state with a nasty speaker load

2.Valve amps usually ascend quite gently into clipping,so you can run such amps very near,or occasionally into, clipping and it will still sound quite nice.Sand amps just hit a brick wall,so you notice it far more easily.I'm sure Anthony will be along to correct me presently! :lolsign:

NRG

16-10-2009, 07:29

So how much power is really needed?

Some of the figures posted above are a bit off ;) Noel Keywood wrote a good article in the November 2005 issue of HFW.

It reads like this: You can get a ball park figure with a bit of simple maths, by measurement small loudspeakers average out at around 86db at 1m for 1watt of power.

That means they will be tolerably loud if you sit 1m in front of them at 1watt. If you double the distance the SPL drops at approx 6db, so if we sit 4m away (not untypical) the SPL drops to 74db, about the same level as a quiet conversation….however there are two channel doubling power so add on another 3db to make 77db. (The SPL doesn’t quite drop at 6db with doubling of distance because we have 4 walls a floor and ceiling that reinforces the sound).

If we now want to play loud at 100db at the listening position (an increase of 23db) we would need approx. 200w from the amplifier (100w per channel)…..that’s an increase of power by x200!

Power levels jump rapidly for any extra 3db of SPL, Concorde was measured at 120db so with our example 1300w would be required to hit this SPL. However, if we relax our requirement for high SPL or sit closer then the power requirements drop rapidly.

So what is loud? In a typical domestic living room most of us are constrained by neighbors and as NK puts it ASBO’s! Plus the need to prevent ear damage…. so very loud under these conditions is 90~100db. Therefore we need an amplifier of 100~300w / channel using our 86db ‘speaker.

If we uses nice sensitive 90db ‘speakers then a 20watt amplifier get us to a loud 90db SPL and a 100watt to a very loud 100db.

Going in the other direction to an insensitive 82db ‘speaker drives the requirement up to 100watts for 90db and well over 300w for 100db.

IMHO, sensitive ‘speakers are the way to go, if domestically acceptable as they tend to be larger and produce much better bass…. they are not so demanding of the amplifier easing the current and voltage demands placed upon it. Also it means the designer has an easier time and the amplifier can be simpler and run in Class A….preferably via a nice set of glowing valves ;)

Chippy_boy

16-10-2009, 09:59

I think he probably has enough loudness as it is,the question is whether he would get better bass definition and impact with a bigger amp ,maybe borrow a MC2 1250 if it makes little difference then stay put.I know it has a little bit of fan noise but the MC2 amps are highly regarded in the bass department.

To be honest, it's really the other way around.

I have an amp with stupidly high output and tremendous grip. I am worried I might lose than if I buy something new that has less power. I like the look of the Avondale power amps but potentially I would be going for something with perhaps 1/4 of the power output of what I have already and that concerns me (maybe needlessly?)

SPS

16-10-2009, 11:05

there are other aspects to the db/watts calculations..

So.. assuming all low and high eff speakers where equally well constucted.....

the higher efficency speaker has a more powerfull motor, it goes lounder because the motor 'powers' the cone more, so its 'in and out' movements are closer to the musical sine wave from the amp

a lower eff speaker does not follow the sine wave as accuratly, if it did it would go louder...
so, if it follows the sine wave wave more accurately... its not just more volume you get?

All speakers tend to loose some efficiency the more power you put in, the lower efficiency types even more so than one with a powerfull motor,
thats is why most high eff drivers do scale much better...

that is called power compression

also the db calculations are based on live music, we listen to recorded music which is all compressed to some degree, just some more than others,
most modern pop music is well compressed.. it has to be to sound resonable on a mp3...
whilst some live music may have an 100db+ 'envelope'... you are lucky to get more than a 30-40 db envelope from most recordings..

but even this 40 db range will produce quite alot of 'power compression' with some speakers at high-ish volume levels

also.. a speakers ability to take loads of watts is linked to its constuction ( heavy voice coils, heat disapation etc)
rather than its ability to reproduce musical peaks...

and this thread is about the playing of musical peaks..

all this high watts stuff is crap really...there are good and bad in all amps

so i would guess if one thinks a more powerfull amp will do the trick.. you could spend a lot of money finding it not to be so

2-10 watts though 95+ db, 40 watt speakers is more than most recordings need..

steve

Chippy_boy

16-10-2009, 14:05

a lower eff speaker does not follow the sine wave as accuratly, if it did it would go louder...

The efficiency of the speakers is influenced by lots and lots of things: the size the cones and their mass; the size of the cabinet; the type of loading (reflex, infinite baffle, transmission line etc), the crossover. Probably more factors I can't think of right now.

It's perfectly possible to have very ineffecient speakers that are highly accurate in that sense, it's just they don't produce much volume. Conversely, you can have some horrible disco 14" paper coned thing that will do 95db/w and be all over the place in terms of fidelity.

So it's just not right to imply that efficient speakers are more faithful to the original recording. In fact it's often a struggle for designers to keep the colouration as low as possible whilst at the same time keeping the sensitivity up to make them easy to drive.

SPS

16-10-2009, 18:25

The efficiency of the speakers is influenced by lots and lots of things: the size the cones and their mass; the size of the cabinet; the type of loading (reflex, infinite baffle, transmission line etc), the crossover. Probably more factors I can't think of right now.

It's perfectly possible to have very ineffecient speakers that are highly accurate in that sense, it's just they don't produce much volume. Conversely, you can have some horrible disco 14" paper coned thing that will do 95db/w and be all over the place in terms of fidelity.

So it's just not right to imply that efficient speakers are more faithful to the original recording. In fact it's often a struggle for designers to keep the colouration as low as possible whilst at the same time keeping the sensitivity up to make them easy to drive.

i did start with a get out clause...

So.. assuming all low and high eff speakers where equally well constucted.....

i never said they where where more faithfull to the original recording though.. I stated that they should follow the musical sine wave more accurately.. and in doing so they go louder.. my experence is that there generally is more fine detail exposed too...

i have a couple of pairs of 100 db speakers if you would like to compare ?

It's perfectly possible to have very ineffecient speakers that are highly accurate in that sense

i'd like to hear them.. we are talking cone speakers though..

Chippy_boy

16-10-2009, 18:37

I stated that they should follow the musical sine wave more accurately..

Yes but that's the bit that is not correct.

How well a cone "follows the musical sine wave" has nothing to do with how loud it plays. Sorry! An inefficient speaker just outputs less. That doesn't mean it adds distortion. As I said, paradoxically it's easier to make an accurate *inefficient* speaker than an efficient one - because small, heavy, stiff cones produce less colouration, but are inefficient. Big, light, wobbly cones are very efficient - they move a lot of air very easily - but they are not well controlled and tend to produce lots of colouration. So actually the opposite of what you say tends to be true.

Regards very accurate, very inefficient speakers, how about the BBC LS3/5A? Incredibly low sensitivity and loved the world over for their incredibly low colouration.

QED.

Marco

16-10-2009, 19:24

Hi Chips,

Big, light, wobbly cones are very efficient - they move a lot of air very easily - but they are not well controlled and tend to produce lots of colouration.

Yes......sometimes, but that's far from written in stone. I can't say that the above is the case with my Tannoy 15" Monitor Golds. Since I've upgraded the original crossovers, they're amongst the least coloured speakers I've ever heard :)

Hi-fi is about choosing your compromises. As far as speakers are concerned, experience tells me that a good big 'un is always better than a good little 'un, providing that the partnering amplifier is suited to the task in hand.

You can't get away from the fact that generating realistic, believeable, sounding music involves shifting some air...

LS35As may have "incredibly low coloration", but only in the midrange because there's no bass to speak of! ;)

Marco.

SPS

16-10-2009, 20:29

Yes but that's the bit that is not correct.

How well a cone "follows the musical sine wave" has nothing to do with how loud it plays. Sorry! An inefficient speaker just outputs less. That doesn't mean it adds distortion. As I said, paradoxically it's easier to make an accurate *inefficient* speaker than an efficient one - because small, heavy, stiff cones produce less colouration, but are inefficient. Big, light, wobbly cones are very efficient - they move a lot of air very easily - but they are not well controlled and tend to produce lots of colouration. So actually the opposite of what you say tends to be true.

Regards very accurate, very inefficient speakers, how about the BBC LS3/5A? Incredibly low sensitivity and loved the world over for their incredibly low colouration.

QED.

i have a pair of kef concerto's they have all the same drivers as the ls3/5 as far as i know..

they are nice speakers, but detailed, dynamic and lifelike.. not in my view , more hi fi like.. i know the drivers have quite a following

so what makes a speaker less or more efficient.. ?
its not floppy cones, that would make them less efficient, thats actually what low efficiency speakers have.. but not in the way you mean...
what is pysically happening to get that extra volume or not happening to get less volume when the musical input is the same? please explain... you may change my views..

thanks
steve

Chippy_boy

17-10-2009, 10:30

so what makes a speaker less or more efficient.. ?
its not floppy cones, that would make them less efficient, thats actually what low efficiency speakers have.. but not in the way you mean...
what is pysically happening to get that extra volume or not happening to get less volume when the musical input is the same? please explain... you may change my views..
thanks
steve

No problem.

As I said above, how efficient a speaker is can be down to many things. I am not an expert on this, so I can't give you a definitive list, but here's a few examples:

1. If you want a speaker cone to accurately reproduce in the incoming signal, then clearly you don't want the cone itself distorting and wobbling as this will introduce harmonics that are not in the original signal. However, if you make the cone stiffer, then generally it will have to be heavier. Of course you can use carbon fibre or other exotic materials but it is true that for a given material, the thicker (and heavier) it is, the stiffer it will be. When you want to accelerate and decelerate the cone (i.e. drive a signal through it) much more energy is wasted moving the cone around if the cone is heavy, compared to a same-sized cone that is lighter. They will both shift the same amount of air if they are the same size and have the same displacement, but the heavier one will need more juice to get it moving (and stopping).

2. If you have a large cone, it can easily shift a lot of air relative to a smaller one. An 8" driver moves more twice as much air as a 6" driver given the same amplitude of movement. (The radiating area is more than 2x the size when you take the rubber surrounding the cone into account.) This means the cone has to move less to produce the same SPL, which is much more efficient.

3. If you have a sealed (infinite baffle) enclosure, then if the enclosure is small, the pressure created inside the cabinet is greater when the cone moves, compared to if the same cone was in a larger sealed cabinet. This pressure acts as a dampener on the cones movement. For each cycle, as the pressure increases, it's harder and harder to move the cone inwards and you need more and more power to get a given amout of cone movement. With bigger cabinets, this effect is reduced and you need less power for the same amount of cone movement.

4. The back pressure on the cone will be less still if you have a transmission-line arrangement where the cone is effectively vibrating freely on the end of an open pipe.

5. The strength of the magnetic field in the speaker, and how the voicecoil is wound will determine how much of the electrical energy is converted to kinetic energy - i.e cone movement - and how much of it is converted to heat in the voicecoil.

These are just some examples. As I say, there are probably many more.

Hope this helps?

SPS

17-10-2009, 11:44

No problem.

As I said above, how efficient a speaker is can be down to many things. I am not an expert on this, so I can't give you a definitive list, but here's a few examples:

1. If you want a speaker cone to accurately reproduce in the incoming signal, then clearly you don't want the cone itself distorting and wobbling as this will introduce harmonics that are not in the original signal. However, if you make the cone stiffer, then generally it will have to be heavier. Of course you can use carbon fibre or other exotic materials but it is true that for a given material, the thicker (and heavier) it is, the stiffer it will be. When you want to accelerate and decelerate the cone (i.e. drive a signal through it) much more energy is wasted moving the cone around if the cone is heavy, compared to a same-sized cone that is lighter. They will both shift the same amount of air if they are the same size and have the same displacement, but the heavier one will need more juice to get it moving (and stopping).

2. If you have a large cone, it can easily shift a lot of air relative to a smaller one. An 8" driver moves more twice as much air as a 6" driver given the same amplitude of movement. (The radiating area is more than 2x the size when you take the rubber surrounding the cone into account.) This means the cone has to move less to produce the same SPL, which is much more efficient.

3. If you have a sealed (infinite baffle) enclosure, then if the enclosure is small, the pressure created inside the cabinet is greater when the cone moves, compared to if the same cone was in a larger sealed cabinet. This pressure acts as a dampener on the cones movement. For each cycle, as the pressure increases, it's harder and harder to move the cone inwards and you need more and more power to get a given amout of cone movement. With bigger cabinets, this effect is reduced and you need less power for the same amount of cone movement.

4. The back pressure on the cone will be less still if you have a transmission-line arrangement where the cone is effectively vibrating freely on the end of an open pipe.

5. The strength of the magnetic field in the speaker, and how the voicecoil is wound will determine how much of the electrical energy is converted to kinetic energy - i.e cone movement - and how much of it is converted to heat in the voicecoil.

These are just some examples. As I say, there are probably many more.

Hope this helps?

that is quite interesting , thanks you...

point 1

the flexing cone.. this is called cone break up, all drivers do this they have to otherwise they will cause peaks, cone are designed very carefully to flex, some use rings which you will see molded in to the cone, others use different methods

point 2,
so are you saying larger speakers are more efficent,
they may be in some cases and it may have a relevance, but there are some very efficient 5,6 and 7 inch cones, and some very efficient tweeters

point 3, in my view is very relative to cabinet design, but very liitle to do with what makes a driver more efficient, yes cabinets have an effect on the driver, but i was talking about driver efficiency

point 4, as point 3

now point 5, is sort of my point.. if you have a powerful motor driving the cone, it starts and stops quicker.. and is more able to keep up with the musical sine wave

think of it as a car, and the signal moving the gas pedal and brake.. motor size counts, the cone will travel further up the sinewave before the sinewave changes direction the other way

the modern hifi way, is to use low efficiency speakers
which are just fast enough to do the low notes, and in many cases over run the signal, adding a bit of extra bass, but they hardly get up the sine wave in the mids and top, this gives an even frequency response from cheap drivers

and that is why most hi fi does not sound lifelike...

hi efficiency drivers are much more difficult to get level, and much more expensive.. due the the cost of the magnet and limited production

one could write a book on the subject and not cover it all..

if i remember correctly, an acoustic watt is 113 dbs, so if you had no losses which is impossible, the music sine wave at 2.83 volts would produce that volume, any thing less = a loss
the musical signal does not slow down to keep up with the driver, the driver just plays/moves as fast as it can.. and it does miss bits out..

but to get all that in perspective,there are some very good performing 90dbs drivers and some not so good 93 db drivers and thats where it gets cloudy... and the hi fi mystic sets in...
but i did start with ...
.. assuming all low and high eff speakers where equally well constucted.....

thanks
steve

Chippy_boy

17-10-2009, 12:38

Steve, I could go on, but honestly mate you are confused about lots of things.

You asked me "so what makes a speaker less or more efficient.. ? " and I answered. You didn't ask me only about differing driver sensitivity, although I gave you reasons for that too. I explained why large cones are more efficient than small, yet you just dismiss it. I see no point in me posting further for you to just argue with it.

The bottom line is, there is no simple correlation between speaker fidelity and sensitivity. Speaker sensitivity is influenced by many factors and to assume that "sensitive speakers follow the sine wave better and therefore sound better" is a gross oversimplifiction, even if it were true and I am not sure it is. In an ideal world a speaker cone would have zero mass and infinite stiffness and as no such material exists, the designer is inevitably balancing mass with stiffness.

"assuming all low and high eff speakers where equally well constucted..... ", you are still mistaken.

Also, 1W does not equal 113dbs. This is like saying 1 mile per hour = 2 oranges. The two do not even measure the same thing, let alone the right numbers. 113 db is simply 2 x 10^^11 times louder than 1 db - which is considered the quietest thing a human can hear and is defined as being 2×10^^−5 Pascals. So 113 db just means 4 million times louder than the quietest thing. I has nothing to do with watts.

There's lots of sources to read up about this stuff if you are interested.

Cheers

Chip

Hypnotoad

17-10-2009, 13:25

Joe

17-10-2009, 13:30

I recommend 'Come On Baby, Let's Go Downtown' by Neil Young & Crazy Horse.

SPS

17-10-2009, 14:08

I think your right chippy ...we'll leave it there..

we are all entitled to our opinions based on our own experence and understanding,

i'm quite aware i've simplified things, other wise it gets too complex,

i've probably read/own more speaker books than many one here, and quite understand how speaker cabinets and drivers work, but thanks for the pointers

cheers
steve

anthonyTD

17-10-2009, 17:13

Well I think Anthony will know much more than me,but in my limited understanding,this is for two main reasons-

1.Valve amps generally cope better than solid-state with a nasty speaker load

2.Valve amps usually ascend quite gently into clipping,so you can run such amps very near,or occasionally into, clipping and it will still sound quite nice.Sand amps just hit a brick wall,so you notice it far more easily.I'm sure Anthony will be along to correct me presently! :lolsign:
hi all,[ali]
well, i sort of agree with two points above, :eyebrows:but as for a valve amp coping better with nasty loads, ermmm,:scratch: they dont get all huffy about it if thats what you mean, but as i have stated many times on this subject, a valve output stage needs to see a constant impedance to produce an even power transfer at all audio frequencies, fortunetly our ears are log and not linear, so they are not bothered too much about small changes in output, but when that change occurs at frequencies our ears are paticularly sensitive to, then thats when the problems as far as lineararity come into play. it is far more dificult to make acurate high sensitivity drivers than it is to make high power transistor amplifiers, so on that score, low sensitivity drive units have become the norm for many years now, and therefore because of their higher damping properties stronger magnets and stiffer cones their probably more acurate to a point than most of the larger high sensitivity types.
as for the reason why valve amplifiers watt for watt can sound louder,,, or to put it another way have more presence than most equal powered solid state designs,well,,,this is a complicated one to answer, but one of the main reasons [IMHO] is to do with the complicated feedback paths needed in solid state designs [too much feedback can suppress perceived dynamics] versus the simple over-all type found in most valve amps, why???quite simply because valves are very linear devices to start with, hence they dont need or use such complications to work properly.
hope this helps.
regards,anthony,TD...