Why do watts matter?

good Question. not sure if I can answer it very well but

89db is at 1 M.
for approximately a 2 M distance increase you lose around 10 dB in loudness

The Input power Vs Ouptput is lograthimic it goes approximately this way
0-50 W = 17 db increase
50-75 = 2 db more
75 to 100= around 2 dB more etc

additionally to increase the sound by 3 dB you need to double the power.

Also the room has an impact.. the bigger the air column in the listening area, the more the power required.


Hence a 87db speaker in a 10 X12 room may need mless power than say a 94 db Sepaker in a 25 X 20 room etc etc

Not to take into account the speaker desin in itself which is not really well represented by the sensitivity figure as 2 speaker with exactly the same sensitivity may need 2 different power levels..

more info below

http://www.sabre-int...aker-technology.html

Neutral schrieb:

The average sensitivity of speakers (as claimed by manufacturers) is 89dB for a signal power of 1 Watt at a distance of 1 metre. 89dB is pretty loud - the safe limit for human hearing is 80dB as per WHO health guidelines. If the manufacturers' claims are true, an amplifier that delivers just 1 watt per channel is all that is required for fairly loud sound. Then why do we end up buying amps that deliver over 50W per channel.

Arj already started down the correct path & so let me elaborate along those lines.
Indeed, 89dB SPL @ the listening position 1m away with 1W of input.
However, who listens to speakers @ 1m (3.3ft) away? At that close a distance, the drivers don't even integrate & one will get a separate-tweeter-separare-woofer kind of sound & not any sort of music!
Most speakers (esp floor standers) are meant to be listened to at 8-9 ft away.
Forevery doubling of distance, we lose 6dB in SPL.
So, if you are sitting at, say, 9'(2.7m), the anticipated SPL from 1 speaker is 89dB - 6dB (1m->2m) - approx 3dB (2m->2.7m) = approx 80dB/1W.
This is still fairly loud - kind of SPL you'd want if you are listening while all the other family members are sleeping @ night.

If you want to take 80dB upto 83dB then you'll need 2W
80dB -> 86dB, you need 4W
80dB -> 90dB you need 10W.

90dB is very loud @ 9' away & 10W/ch is probably enough continuous power.
The remaining juice left in the power amp is for dynamics - cymbal crash, drum thumps, cresendo of an instrument, chorus lines in vocals, etc. Some of these transients reach pretty loud levels even in our home listening particularly in classical music. It is not uncommon to require reproducing 100dB SPL for these transients so that you can feel the air & hear the spatial nature of the recording.
So, from 80dB -> 100dB SPL, you will need 100W/ch.
When the amp is reproducing these transients, you don't want to power amp to have exactly the needed power, you want it to have a bit more i.e. some reserve so that it doesn't get out of breath & doesn't compress. Thus, tacking on another, say, 25W, to make it a 125W/ch amp would be wise.
Am I making sense?

Neutral schrieb:

Recently, a member on this forum complained that a Marantz AV receiver was not loud enough for his 86dB sensitivity Wharfedale speakers. Can the technically knowledgeable forum members (Amp Nut, Arjun, Sachi, Mohan, Manek) please explain why we really need powerful amps. Do only watts matter or is current the real requirement?

Well, voltage & current are inextricably linked i.e. where there is one, the other also shows up automatically. They call is Ohms Law.
Also. power = voltage X current = voltage squared/resistance = current squared X resistance

>> Do only watts matter or is current the real requirement?
the answer to this is "yes"!
there are some speakers that require high watts & lower current. The large number of watts produces a large voltage swing at the speaker inputs for it to work effectively.
there are other speakers that require high watts but also high current.
I believe that it has do w/ the cross-over design.

WOW.....so spot on Bombaywalla

Here’s one more simple article explaining the wattage-amp/speaker relationship in simple layman terms
http://www.whathifi....D=34&newssectionID=3

Arj and bombaywalla has already dwelled on the distance/power relationship. Let me add, that I have had some funny experiences with some weak amps with the Sheffield drum test track cd. Most amps with weak power supplies and power ratings struggle with those tracks as the transients are very sharp and the dynamic range very high. They clip so hard that this cd has become my fav for checking speaker/amp compatibility.

Here is an interesting thing however.Recently a freind of mine measured a piano quartet that we attended.From half way in the concet hall the reading measured 70-75dbA,from row D the variation was about 4dbs i.e 74-79dbA.We never felt for a moment that we were hearing it too softly in either position, so I have doubts regarding loudness perception with regards to live sounds versus reproduced sounds,because we always listen to reproduced sounds louder than live.I don't have a theory why but this was just an observation I thought I should share with the forum members.
Regards Deaf.

square_wave schrieb:

Here’s one more simple article explaining the wattage-amp/speaker relationship in simple layman terms http://www.whathifi....D=34&newssectionID=3 Arj and bombaywalla has already dwelled on the distance/power relationship. Let me add, that I have had some funny experiences with some weak amps with the Sheffield drum test track cd. Most amps with weak power supplies and power ratings struggle with those tracks as the transients are very sharp and the dynamic range very high. They clip so hard that this cd has become my fav for checking speaker/amp compatibility.

that comes from the regular and dynamic opwer which bombaywalla touched upon.

in the end power is not static.. at each frequency the speaker (can be considered an RLC circuit) has different impedences and hence reduires different power. hence the capability of the amp to source that power is more important.

eg a very Robust 20W amp with capability to act as a high current source will be perceived as more powerful than a 50W amp with lower capability. thats the reason why amps like NAD 320BEe/Rotel RA01 are so highly regarded, and can take hgher loads.

Yup. I remember the day when prithvi asked me to try out the 320bee at home with my old diamond 8.4’s. I was my wits end trying to breathe some life into those power hoggers without clipping. I had tried out a few Japanese mass-market 90/100 watters without any improvement. I didn’t believe him asking me to try out this flimsy looking 50watter. I couldn’t stop smiling when I hooked it up and the wharfies started to sing effortlessly for the first time………

deaf,

Your observation about loudness in concert halls and home listening is very correct. I too have noticed this
Off the top of my head a couple of reasons for this could be...
1) concert halls are usually very well insulated from the outside noise so what you are hearing is 75db of music which is being played on stage and not 75db of sound which is a combo of your speakers + external noise
2)Acoustically the hall being superior, maybe the decay of amplitude of sound waves may be slower than in a home for whatever reason(just thinking aloud, no proof here...)
3)In a concert hall subconciously there is a higher level of concentration on a persons part to hear every note.

Maybe we should try the same measurement in an insulated recording studio. That may draw similar readings ?

Manek.

Manek schrieb:

deaf, Your observation about loudness in concert halls and home listening is very correct. I too have noticed this Off the top of my head a couple of reasons for this could be... 1) concert halls are usually very well insulated from the outside noise so what you are hearing is 75db of music which is being played on stage and not 75db of sound which is a combo of your speakers + external noise 2)Acoustically the hall being superior, maybe the decay of amplitude of sound waves may be slower than in a home for whatever reason(just thinking aloud, no proof here...) 3)In a concert hall subconciously there is a higher level of concentration on a persons part to hear every note. Maybe we should try the same measurement in an insulated recording studio. That may draw similar readings ? Manek.

Dear Manek,
The controlled enviroment reading has also been done by us,the results are the same,we end up hearing reproduced sounds louder than live.With regards to point No 3, I listen with concentration and casually, a lot,both live and reproduced,still the observation remains the same.
The only real thing I can think of is noise floor in equipment versus none in instruments,which seems to be the only logical reason for cranking up the volume.I could ofcourse be very wrong though.
Regards Deaf.

Wow! So many replies and all so intelligent.
Bombaywalla wrote:
[Most speakers (esp floor standers) are meant to be listened to at 8-9 ft away. Forevery doubling of distance, we lose 6dB in SPL.
If you want to take 80dB upto 83dB then you'll need 2W
80dB -> 86dB, you need 4W
80dB -> 90dB you need 10W.

90dB is very loud @ 9' away & 10W/ch is probably enough continuous power. The remaining juice left in the power amp is for dynamics - cymbal crash, drum thumps, cresendo of an instrument, chorus lines in vocals, etc. Some of these transients reach pretty loud levels even in our home listening particularly in classical music. It is not uncommon to require reproducing 100dB SPL for these transients so that you can feel the air & hear the spatial nature of the recording.
So, from 80dB -> 100dB SPL, you will need 100W/ch.]

Thanks Arjun and Bombaywalla, I understand you line of reasoning. 1W is great if you are doing near-field monitoring. But given real room conditions and sound-absorbing surfaces, it pays to have more watts. And transients would require more power - maybe as much as 20dB more for classical music?


Bombaywalla wrote:
[there are some speakers that require high watts & lower current. The large number of watts produces a large voltage swing at the speaker inputs for it to work effectively.
there are other speakers that require high watts but also high current. I believe that it has do w/ the cross-over design]

Bombaywalla and Abhi, do elaborate on the cross-over design aspect. Would be interesting to know even for us non-designers.


Arjun wrote:
[in the end power is not static.. at each frequency the speaker (can be considered an RLC circuit) has different impedences and hence reduires different power. hence the capability of the amp to source that power is more important.

eg a very Robust 20W amp with capability to act as a high current source will be perceived as more powerful than a 50W amp with lower capability.]

Arjun, That's one of the major challenges in speaker design. Making a speaker that won't suddenly change its impedance as the frequencies vary. This is also the principal reason why just a good amp + a good speaker does not equal good sound. The two must pair together well. The amp needs to accommodate the speaker's weaknesses. A technical question: Does a higher slew rate help an amp to cope with these changing impedances?


Deaf wrote:
[Recently a freind of mine measured a piano quartet that we
attended.From half way in the concet hall the reading measured 70-75dbA,from row D the variation was about 4dbs i.e 74-79dbA.We never felt for a moment that we were hearing it too softly in either position, so I have doubts regarding loudness perception with regards to live sounds versus reproduced sounds,because we always listen to reproduced sounds louder than live.]

Deaf and Manek, My take is that it's an issue of psycho-acoustics. In a real concert room environment, you have visual cues. At home, you don't. That is why you compensate by raising the volume. You are forcing your mind into believing that what you are hearing is real even though you can't see any of it.
Incidentally, I usually listen to recorded music at lower volumes - 60 to 70dB only - because I focus on the emotion rather than what I hear. As Morpheus said in the Matrix - your mind makes it real. Please try listening to music that stirs your emotions and check if you really need to play it loud - 80 to 100dB.

Neutral schrieb:

A technical question: Does a higher slew rate help an amp to cope with these changing impedances?

I really am not very technical .. but my simple understanding is that moment we talk about figures like Power/slew rate etc etc.. they all relate to numbers on one particular frequecy. but when you play audio you play so many frequencies simultaneously. some amps do not do well at certain frequecies but do others very well, and each amp has its quirks.

ditto about the response of speakers..

Thats the reason why i would rather not get too technical on this and leave it to the DYI folks. but its important that they sound well together !

Arj schrieb:

Neutral schrieb: A technical question: Does a higher slew rate help an amp to cope with these changing impedances? I really am not very technical .. but my simple understanding is that moment we talk about figures like Power/slew rate etc etc.. they all relate to numbers on one particular frequecy. but when you play audio you play so many frequencies simultaneously. some amps do not do well at certain frequecies but do others very well, and each amp has its quirks. ditto about the response of speakers.. Thats the reason why i would rather not get too technical on this and leave it to the DYI folks. but its important that they sound well together !

Yes……….This is the reason why good audio is a black art. Attaining perfect synergy with the room and the active components like amp, source and speakers is the key to good sound that will stir your emotions. This is never something that one can learn from books or websites. I once asked a senior audiophile how I can replicate similar sound like his setup in my house. He asked me to buy the same equipment including all the interconnects and power cables and put it in a similar sized room with the same furniture…………..I could only smile……..If you just go by the obvious specs and put together a setup which theoretically should work fine, it could end up as a total disaster. It is here that very professional and knowledgeable audio consultants with decades of experience with different brands comes into the picture. I wish we had such people who could charge a decent fee for their services. People who could put together a system for someone depending on his music tastes and budget. We usually have people who want to sell you stuff that they have in stock. Most are sales people who have very little knowledge outside the brands which they stock and most do not understand what “ good sound” is.

Dear Square Wave,
You are right 'most' is the word, but not all of us are just traders.There are some who genuinely are in the pursuit of audio excellence, and can acheive fabulous results if the client is willing to invest in our products and knowledge.It is finally up to the client how to discerne a knowledable dealer from a mere trader.Thank you for bringing this point up, it will go a long way in raising the level of audio performance in our country. Infact if you deem it fit, please start a new thread with this topic.
Regards Deaf.

Deaf Said: Recently a freind of mine measured a piano quartet that we attended.From half way in the concet hall the reading measured 70-75dbA,from row D the variation was about 4dbs i.e 74-79dbA.We never felt for a moment that we were hearing it too softly in either position, so I have doubts regarding loudness perception with regards to live sounds versus reproduced sounds,because we always listen to reproduced sounds louder than live.I don't have a theory why but this was just an observation I thought I should share with the forum members.

I would like to provide my 2 cents worth of speculation, on why most of us listen to our stereos louder than "live."

Our 'listning' to live music includes various sensations and frequencies. Some of these include:
1. Different frequencies
2. Spatial clues such as location of the sound, within the soundstage ( Visual re-inforcement Does help a lot, in a live performance )
3. Textures and decay of the notes ( including local ambience)
4. Physical stimulation such as the impact of a kick drum played at close quaters.

A stereo system ideally needs to reproduce all this, and that too, in a different local environment.... more often than not, a Tall order ...

Hence we as listners, crank up the volume, to try to reproduce aspects ( 1 to 4 and Im sure there are many more )
that we do not experience. Often this results in one of the aspects being exagerated, and substituting the missing aspects.... eg Loud volume and strong, physical bass impact creates to some, a pleasurable experience that makes them not miss so much, the texture of the notes, and their decay.

Generally, I think that better systems do not typically need to be played loud to satisfy...

Neutral wrote: A technical question: Does a higher slew rate help an amp to cope with these changing impedances?

Slew rate is a measure of How Fast the amplifier can respond to a sudden change of the input signal.

It is measured ( for audio amps ) in Volts per micro-second.

A High Slew rate will improve stuff like the definition and decay in the sizzle of a cymbal.

It will also play a part in the Leading edge of the sound of the kick drum.

However, slew rate is not primarily responsible for the visceral impact of the drum.

The visceral impact is largely dependent on the Brute force capability of the Power amp.


A touch more of tech .....

Overall feedback does not improve the slew rate of an amplifier. Infact poor slew rate can make a High Feedback amplifier sound Miserable.

Poor slew rate results in TIM ... a distortion that muddies up the sound when fast transients are played thru the amplifier.

deaf schrieb:

Dear Square Wave, You are right 'most' is the word, but not all of us are just traders.There are some who genuinely are in the pursuit of audio excellence, and can acheive fabulous results if the client is willing to invest in our products and knowledge.It is finally up to the client how to discerne a knowledable dealer from a mere trader.Thank you for bringing this point up, it will go a long way in raising the level of audio performance in our country. Infact if you deem it fit, please start a new thread with this topic. Regards Deaf.

“Most” is definitely the keyword. I know at least a couple of dealers who are very knowledgeable and willing to invest their time for a fee. I guess a commercial member would be the right person to start a thread on this.

Neutral schrieb:

Bombaywalla[/b] wrote: [there are some speakers that require high watts & lower current. The large number of watts produces a large voltage swing at the speaker inputs for it to work effectively. there are other speakers that require high watts but also high current. I believe that it has do w/ the cross-over design] Bombaywalla and Abhi, do elaborate on the cross-over design aspect. Would be interesting to know even for us non-designers.

OK, let me try.
You might remember that there was a thread not too long ago called something like "Ampilifier design considerations - output stage"? I think that Amp_nut started it. Seek that thread out & read 1 of Amp_nut's posts which explains why it takes more power to drive reactive components L & C.

Basically, the L & C components have a phase angle attached to their frequency dependent impedance. The phase angle part manifests itself by the fact that if power is delivered into the L or C input terminals (speaker binding posts), it takes a short while before this power shows up on the output terminals (connected to the driver inside the speaker box). So, if a music dynamic pulse occurs right now, the L or C cannot deliver the power right now but a short while later by which time the music dynamic pulse is gone!
So, the instantaneous power in a reactive component is less than in a resistive component.
Having a large wattage in the power amp helps as it (amp)can deliver a large chunk of power to the x-over ckt in the speaker. Then, even if the instantaneous power falls short, there is so much of it delivered that whatever makes it to the driver terminals is more likely to be audibly sufficient.

Also, discussed in that thread is one of my posts that mentions that the best way to make a xover (really a filter) even today is to use R, L & C components. Search the internet & you'll see several examples of 1st, 2nd, 4th order xovers & you'll notice the extensive use of L & C components.

Further, several speaker manuf such as B&W (which I used to own) design hi(4th) order xovers. So, 1stly, a hi order xover uses more components. Then, many manuf also use the xover to compensate for driver SPL mismatches, lobing due to baffle reflections & impedance variations over freq. Thus, it is not uncommon to see sub networks within the main xover network to perform all these compensations. Obviously, a large number of R, L & C components are needed for this. This creates more havoc in the music signal because, don't forget, the phase angle in the L & C components xlates to phase shift thru the component. As the music signal passes thru the entire, complex xover, the phase angle goes thru some very complex chnages. You can see these in the speaker impedance & phase plots (published in each Stereophile review of every speaker they review).
You will notice that at 1 or more points the speaker impedance falls to some real low value often < 4 Ohms. In such cases/frequencies, the power amp has to dump current into a virtual short-circuit! If the power amp is not a brute, it'll run short on current & the music will sound compressed.

So, it's easy to see now how speaker xover greatly affects the amp's ability to deliver power into the speaker thus directly affecting the quality of reproduced music.

Some of the solutions used by users is to buy very high efficiency speakers (> 95dB/1W/1m SPL) that use relatively benign xover ckts like 1st xovers OR buy hi eff single driver speakers that use no xover components & then pair these speakers with flea watt power amps.
Others use active Linkwitz-Riley xovers where the active cktry does a much better job of controlling the overall phase shift & can do a dual job of amplifing the signal from the power amp to the speaker.
Still others buy speakers designed by people who know what the hell they are doing & those designers have shown that they can make a simple 1st order xover reproduce music effectively without making the power amp tear its hair out.

Dear Bombaywalla
Much as I agree with you that complex crossovers can cause complex problems for many a amp, it actually holds true if the crossover is usually in the 80-150 cycles region due to the large crossover component values.Let us not forget that one of the greatest amp busters of all time used a first order crossover, the Theil CS5i.Let me however say, I myself use a DIY speaker using a first order too as I belive in them.However the best speakers that I have heard in my life so far, all use extremely complex crossovers, obviously designed by people who really know what they are doing,with the exception of two speakers which used 1st order crossovers, also designed by people who 'really' knew what they were doing. I would love to pursue this topic as it is great fun.
Regards Deaf.

Phew!..a lot of foood for thought..gotto collect my thoughts...interesting..keep it going guys..

deaf schrieb:

Dear Bombaywalla Much as I agree with you that complex crossovers can cause complex problems for many a amp, it actually holds true if the crossover is usually in the 80-150 cycles region due to the large crossover component values.

agree! I don't think that I said anything different.
at low freq, the impedance of the xover falls to a very low value & the amp has to supply a lot of current to create a respectable voltage voltage in order to generate the user wanted SPL. I think that I wrote this in my post.
At higher freq, you do need much power at all but an ill designed xover will create a lot of additional phase shift that'll mess up the original music signal & you'll hear it!

deaf schrieb:

Let us not forget that one of the greatest amp busters of all time used a first order crossover, the Theil CS5i.

I'm not much of a Thiel fan. Just a personal choice. However, be careful at accepting the manuf words at face value: when a manuf says he's using a 1st order xover, he is probably correct BUT ask him if the xover is 1st order electrically PLUS acoustically. If it is electrically + acoustically 1st order, it is *true* 1st order system. Otherwise, the mechanical roll-off of the drivers (1st order) + the 1st order electrical xover add to create an effective 2nd order system.
I will not be surprised 1 bit to know that the Thiel was effectively a 2nd order (or higher) system.

deaf schrieb:

Let me however say, I myself use a DIY speaker using a first order too as I belive in them.However the best speakers that I have heard in my life so far, all use extremely complex crossovers, obviously designed by people who really know what they are doing,with the exception of two speakers which used 1st order crossovers, also designed by people who 'really' knew what they were doing. I would love to pursue this topic as it is great fun. Regards Deaf.

well, my "solutions" were not exhaustive & never meant to be. They were meant to highlight how people address this issue. Fine by me if you like higher order xover speakers.

Dear bombaywalla you are spot on with regards to 1st order crossovers,and regards to acoustic+ electrical rolloffs.It is therefore necessary to have the driver extend at least 2 octaves on either side of its proposed x-over point, to have a true 1st order.The other option is to have a mechanical rolloff like the way I have opted for in case of the midbass driver,and have a electrical one for the tweeter.Of course one has to physically offset the tweeter to lock the phase and to achieve the required time domain response.I know you like 1st order speakers,especially ones which say Green.
Takes one Marian to recognise another.
Deepest Regards Deaf.

Great posts by Bombaywalla & Deaf...

I'm soaking it in.

Do keep it coming.

Cheers

deaf schrieb:

Dear bombaywalla you are spot on with regards to 1st order crossovers,and regards to acoustic+ electrical rolloffs.It is therefore necessary to have the driver extend at least 2 octaves on either side of its proposed x-over point, to have a true 1st order.The other option is to have a mechanical rolloff like the way I have opted for in case of the midbass driver,and have a electrical one for the tweeter.Of course one has to physically offset the tweeter to lock the phase and to achieve the required time domain response.I know you like 1st order speakers,especially ones which say Green. Takes one Marian to recognise another. Deepest Regards Deaf.

Cool! we are on the same page then!
Yes, I like speakers that say "green..." If you have ever heard them then you could also tell why!
SOG! You are a Marian too? very cool!
I probably know you off-line by your real name but "deaf" has me guessing!

bombaywalla schrieb:

deaf schrieb: Dear bombaywalla you are spot on with regards to 1st order crossovers,and regards to acoustic+ electrical rolloffs.It is therefore necessary to have the driver extend at least 2 octaves on either side of its proposed x-over point, to have a true 1st order.The other option is to have a mechanical rolloff like the way I have opted for in case of the midbass driver,and have a electrical one for the tweeter.Of course one has to physically offset the tweeter to lock the phase and to achieve the required time domain response.I know you like 1st order speakers,especially ones which say Green. Takes one Marian to recognise another. Deepest Regards Deaf. Cool! we are on the same page then! Yes, I like speakers that say "green..." If you have ever heard them then you could also tell why! SOG! You are a Marian too? very cool! I probably know you off-line by your real name but "deaf" has me guessing!

Yup Deaf is confusing,but we know each other.Let us just say I am Marian,love a particular star shaped driver, and I am the keenest ear in the whole audiophile cirus.The last bit sound pompous,but what the hell,all audiophiles are allowed that much.
Deepest regards Deaf.

Can we know whats deaf's real name.. hmm deaf and keen ears are clear cut anatonyms??

deaf schrieb:

love a particular star shaped driver

pardon the "evesdropping" but does than happen to be a Manger perchance ?

Arj schrieb:

deaf schrieb: love a particular star shaped driver pardon the "evesdropping" but does than happen to be a Manger perchance ?

Me too think it is a manger. I always wanted to hear one. I wonder how the zerobox 109 with a REL sub would sound..

[quote="deaf
Yup Deaf is confusing,but we know each other.Let us just say I am Marian,love a particular star shaped driver, and I am the keenest ear in the whole audiophile cirus.The last bit sound pompous,but what the hell,all audiophiles are allowed that much.
Deepest regards Deaf.[/quote]

Got it!!
Now all you gotta do is to be a hermit & use those keen ears to listen to some music!

Bombaywalla,
Was deaf's identity ever in doubt ? A few guys have guessed, and may I add guessed right !

Manek.

Manek schrieb:

Bombaywalla, Was deaf's identity ever in doubt ? A few guys have guessed, and may I add guessed right ! Manek.

Manek, I don't know how you guys know but I certainly had no idea until now.
I think some of you might be early members & thus might have insight into which moniker belongs to which person.

so back to watts.....
There is another theory which I read somewhere but dont pay much heed to, a huge discussion on some forum.........lower powered amps generally have better timing, are more nimble and hence more musical....what do you guys say ?

Manek.

Hi Maneck,

I generally agree....

I am one of those who subscribe to it, although do not really have the technical base as some of the other members :). this may also be the time to get these layman views clarified/corrected

From my view, when you are playing dynamic content the variation in demand of current is very high. with a high sensitivity speaker the difference between the maximum and min current requirement is usually lower.. while with a lower sensitivity speaker this difference is very high.

hence the amp may have an easier time servicing the first than the the second.

What I am not clear about. is it equally difficult for
-A 10 W amplifier to move from say .1 A to .5 A (for a 100dB speaker) as it would be for
- A 200 W amp to move up from 1 A to 5 A for a (82 dB speaker)

of course these Numbers may not be mathematically relevant but are only for illustrative purposes

Manek schrieb:

so back to watts..... There is another theory which I read somewhere but dont pay much heed to, a huge discussion on some forum.........lower powered amps generally have better timing, are more nimble and hence more musical....what do you guys say ? Manek.

oh, watt did you say? Can't hear you properly 'cuz my megawatt power amp is playing on full power into an 82dB loudspeaker!!

Re. lower power amps: I believe that the design is usually targetted towards higher eff speakers so that about 1W of the amp is required to produce sufficient SPL. Hence the amp can be designed w/ minimal output devices implying no need to pay attention to output device matching. The distortion (that comes from multiple output devices) also goes away. Hence the power supply can also be designed for the minimal # of output devices & distortions from the p.s are also reduced. All these reduced distortions cummulate towards producing a purer sound, which xlates to better timing & better musicality.
Nimbleness: I would still attribute that to the slew rate of the amp, its power bandwidth, the power supply & the speaker load.

Arj schrieb:

I am one of those who subscribe to it, although do not really have the technical base as some of the other members :). this may also be the time to get these layman views clarified/corrected From my view, when you are playing dynamic content the variation in demand of current is very high. with a high sensitivity speaker the difference between the maximum and min current requirement is usually lower.. while with a lower sensitivity speaker this difference is very high. hence the amp may have an easier time servicing the first than the the second. What I am not clear about. is it equally difficult for -A 10 W amplifier to move from say .1 A to .5 A (for a 100dB speaker) as it would be for - A 200 W amp to move up from 1 A to 5 A for a (82 dB speaker) of course these Numbers may not be mathematically relevant but are only for illustrative purposes ;)

re. the load driving ability of a 10W/ch vs. a 200W/ch really depends on the speaker load i.e. x-over ckt & how the drivers themselves interact w/ the x-over. The reason is that the power amp is really delivering its power into the x-over ckt (its load).

You could pair up a relatively efficient speaker, say a Tannoy DMT10 Mk2, which is 94dB SPL/1W/1m, to a 10W/ch amp. This particular speaker is a flat 8 Ohm load. The 10W/ch will play music nicely thru this speaker but the dynamics will be compressed particularly in the bass region where enough power is required to activate the dual front ports.
However, the 200W/ch will be much better in driving this benign load.
So, in this case, the load is easily driven by both amps but the smaller one is limited by its absolute output wattage.
Then consider the eg that Amp_nut gave a few posts back: The Thiel CS5i, which broke many a high power amps! So, in the on-going tradition of Thiel, the CS5i is a difficult load even tho it's just a 1st-order x-over (naaaaaaah! ) & neither the 200W/ch nor the 10W/ch can drive it effectively. One probably needs a 600W/ch or a 1KW/ch power amp which can drive a a near dead short in the bass region & deal w/ the roller-coaster impedance curve in the mid-bass & upper bass.

What I am not clear about, Is it equally difficult for -A 10 W amplifier to move from say .1 A to .5 A (for a 100dB speaker) as it would be for - A 200 W amp to move up from 1 A to 5 A for a (82 dB speaker)

IMHO the absolute values rather than % increases are what is difficult to cater.

An ant, I am told is the strongest creature on Earth... capable of lifting 50 times its weight. However I wonder if that would have held, if the ant weighed 1 Kg

Bombaywalla has made some valid points about the difficulkty of loads...

Another inportant fact to consider is that lower power electronics usually have Much better performance that High power electronics.

As an example, a medium power transistor, ( say 5 Watts ) can easily have a 100 MHz bandwidth, but dont expect this from a 150 Watt Power Transistor....

Hence a 200 Watt Power amplifier has to make do with those slow devices, even when delivering its First Watt....

It only has an advantage, when called on to deliver what the little one just cant handle ...


Cheers

P.S: Incidentally, it was not me that mentioned about the Thiel. I know little of Loudspeakers, and would love to learn more. The Thiel sound is not really up my street...

Thanks Bombaywalla,
For your very enlightening reply on impact of xover design on watt requirements. I am also a "Bombaywalla" and would like to meet you sometime. We could have an interesting discussion that would be a lot easier than typing into this forum. Please PM me if you can spare the time.

This is what I understood from your posts:
1. A complex crossover adds a phase angle to the signal before it reaches the actual driver. This reduces the impulse response and lowers the sharpness of the transients.

2. Could you give us the full forms of R, L, and C

3. A question: Is phase shift audible. Some experts say that music recordings are often not phase correct themselves. So does loudspeaker phase distortion seriously affect sound quality?

4. It takes a high-current power amp to handle reactive loads, especially at lower frequencies.

5. An active Linkwitz-Riley xover handles phase shift better. Will I find this type of xover also in a passive speaker (which most of us have).

6. I agree that 1st order xovers maintain phase better. But a higher order xover would minimise driver resonances outside its useful bandwith. eg. a woofer peaking at say 4KHz.

7. When you say "mechanical roll-over" are you talking of some kind of a band-pass design? eg. the Lithos Terra sub-woofer.

8. A high slew rate will improve the definition of transients and lower TIM.

9. Some good amps like the Bryson 60W have a single output device. This lowers the distortion arising from output device mismatch.

10. If complex crossovers are a problem, is that why bookshelves often sound better? They have just 1 xover, compared to 2 in the average flrs.

11. Are single-driver loudspeakers a viable option? Can suitable drivers be obtained at affordable prices?

This has been one of the most interesting topics that I have created. Great participation by all members. Please tell me if I got the above points right.

P.S. : What is the latest version of WinIsd? My version is a few years old. I am using it to understand the basics of speaker terminology and design. Are there any other high-quality freewares available. PM me if you come across something.

Hi Neutral,

That is probably the most comprehensive summary of any thread that I have seen in a looong time. GREAT !

Now that it is all clear, let me confuse things a bit. All ( most ) of your summarys are correct at a priliaminary level, but if you dig deeper, in a quest for perfection, many audio component designers have turned most of these on their head, and have Sometimes .... created better products :

. A question: Is phase shift audible.

Yes, it is, and speaker designers often even physically stagger the tweeter w.r.t. the woofer, so that the signals from the tweeter and woofer arrive at the ear in the correct phase. It improves Imaging, significantly. I think Tanoy was the 1st to do this on their 15inch woofer which had the tweeter embedde in the woofer, near the woofer voice coil.

Could you give us the full forms of R, L, and C

R: Resistance

C: capacitance

L: Inductance

I agree that 1st order xovers maintain phase better.

Actually, even a SINGLE inductor or capacitor shifts phase.

Even a speaker driver in a closed box acts like an Inductor and / or a capacitor, often changing its beghaviour from Inductive to capacitive, at different frequencies.

Some good amps like the Bryson 60W have a single output device. This lowers the distortion arising from output device mismatch.

On the other hand, multiple output devices have lower output impedance ( better speaker control ), higher current delivery capacity, and better heat dissipation characteristics..

Are single-driver loudspeakers a viable option? Can suitable drivers be obtained at affordable prices?

Many horn loaded loudspeakers use a single driver, and have Very High efficiency ( require low amplifier power ).

They have their quirks, but make music in some of the most critical audio octaves... :-)


Im sure Bombaywalla ( and others ? ) can add some more info for all of us, on the points summarised by you.
Cheers

Neutral schrieb:

Thanks Bombaywalla, For your very enlightening reply on impact of xover design on watt requirements. I am also a "Bombaywalla" and would like to meet you sometime. We could have an interesting discussion that would be a lot easier than typing into this forum. Please PM me if you can spare the time.

Welcome!
I'll get your contact details & we can meet next time I'm back in India/BOM.

Neutral schrieb:

This is what I understood from your posts: 1. A complex crossover adds a phase angle to the signal before it reaches the actual driver. This reduces the impulse response and lowers the sharpness of the transients.

only the 1st part is correct. phase shift has nothing to do w/ reducing the impulse response.
the addition of phase shift manifests itself as altered harmonics of the music: guitars sound more metallic, piano notes sound sharper, cymbal crashes make you shield your ear w/ your hands, etc

2. Could you give us the full forms of R, L, and C

Neutral schrieb:

3. A question: Is phase shift audible. Some experts say that music recordings are often not phase correct themselves. So does loudspeaker phase distortion seriously affect sound quality?

phase shift is very audible! The music recording can be screwed up by the engineers doing the job. Often they have a good excuse for what they did. We tend to blame them but we really do not know under what conditions they recorded, mixed, mastered etc.
There is a saying in English: if the truth is known, all can be forgiven.
If we knew the "truth" i.e. conditions of recording we could understand & forgive the end recording.
However, the music on the CD/LP/8-track, etc is our reference for better or for worse.
Any system that DOES NOT corrupt it further is the better system. And, yes, loudspeakers can distort the distortion!

4. It takes a high-current power amp to handle reactive loads, especially at lower frequencies.

Neutral schrieb:

5. An active Linkwitz-Riley xover handles phase shift better. Will I find this type of xover also in a passive speaker (which most of us have).

no, a L-R active filter doesn't handle phase shifts better. A L-R active filter is a 4th or 8th order xover (both are available) that uses certain assumptions (based on Linkwitz's 1972 paper in the AES journal) that creates a -6dB attenuation at the cut-off points such that the overall response at the x-over point is flat. If his filter had only -3dB at the cut-off points, the overall response would have had a bump at the x-over point.
I don't know whether people use passive L-R x-overs - I haven't bothered to find out. Sorry!

Neutral schrieb:

6. I agree that 1st order xovers maintain phase better. But a higher order xover would minimise driver resonances outside its useful bandwith. eg. a woofer peaking at say 4KHz.

yes, higher order x-overs are better at curtailing the power from one driver into leaking into the other driver(s) vs. 1st order x-overs. So, one cannot play 1st order speakers *AS* loud but that doesn't mean that you cannot play a 1st order speaker loud. I hope that you understand this difference!
woofer peaking at 4KHz - that's very far away from a woofer max freq. If your woofer is peaking @ 4KHz & you can measure it, that speaker designer should close shop & go home!
Did you mean midrange peaking @ 4KHz?
Anyway, a higher order x-over solves that issue by aligning the phases of the various sections at the x-over point to avoid driver resonances from being heard.

Neutral schrieb:

7. When you say "mechanical roll-over" are you talking of some kind of a band-pass design? eg. the Lithos Terra sub-woofer.

no! don't know why it gave you an idea of bandpass design! mechanical roll-off implies the freq at which the physical driver begins to roll-off owing to its driver size. if you measure the driver as-is (no x-over attached) you'll find that every driver has a freq range. Above/below that freq range, the response rolls off.

Neutral schrieb:

8. A high slew rate will improve the definition of transients and lower TIM.

yes, definition of transients improves. Don't know whether TIM improves or not.

Neutral schrieb:

9. Some good amps like the Bryson 60W have a single output device. This lowers the distortion arising from output device mismatch.

yes, Bryston 60W/ch doesn't have to worry about distortion from multiple devices turning on to supply the load. However, this amp might not have enough juice to do the music justice so it might sound compressed.
So, what would you rather have: a low distortion amp that doesn't have the balls OR a higher distortion amp that does?
You might say: I want to have the best of both worlds! Of course!
Get out your wallet! How much money you got?
So, you can see why it's hard to make a good amp & those that do, charge a lot for it. Plus, you can count on 1 hand who those designers are world-wide.

Neutral schrieb:

10. If complex crossovers are a problem, is that why bookshelves often sound better? They have just 1 xover, compared to 2 in the average flrs.

LOL! the number of x-overs is not in question. Just because you have 1 or 2 or 3 or more xovers doesn't have much to do w/ the sound. That 1 x-over the monitors have can be a 4th order x-over!
just 'cuz it's 1 x-over doesn't mean it's a 1st order!!

Neutral schrieb:

11. Are single-driver loudspeakers a viable option? Can suitable drivers be obtained at affordable prices?

oh yes! I wouldn't know where as I'm not into single-driver but lots of DIY folks are designing w/ Lowthers, Jordan & Fostex drivers.

This has been one of the most interesting topics that I have created. Great participation by all members. Please tell me if I got the above points right.

Neutral schrieb:

P.S. : What is the latest version of WinIsd? My version is a few years old. I am using it to understand the basics of speaker terminology and design. Are there any other high-quality freewares available. PM me if you come across something.

no idea! I leave art of speaker design (there is a science to it, of course, but the best guys know how to xlate that science into a manuf product & therein lies the art) to those who know what they are doing & I ain't one those (who knows speaker design)!

In regards to watts and loudness lets get a few apects in place.Loudspeakers are one of the most inefficient transducers known to exist.Their efficiency is 0.1%( direct radiators) to 5%(hifi horns).A piano at its loudest produces 0.5 acoustic watts,therefore a direct radiating loudspeaker with average an efficiency will require about 500-1000 watts to replicate the event.But we all measure in SPL don't we, and the chart says with average efficiency we need only a couple of watts, so why is a 1000watts required? Distortion reduction of all forms during playback is the key I think.
All the components in a system should be at 10% of their full power to reproduce 110db,which means at 90db the system is operating at 0.1% of its abilities. Hence no distortion when 0.5 acoustic watts(guess work is about 110db) has to be reproduced momentarily;the system is still at 10% of full power.I didn't says this, some real gurus say this.
Puk studios in Denmark has a custom monitoring system designed with 10000watts of amplification.Apparently it is one of the only systems in the world which comes close to the real thing.Before I get baulked at or laughed at by everybody else let me say this, that this setup was designed in partnership with the chief engineer of B&K, the worlds foremost microphone manufacturer, and noise and vibration measurement manufacturer.Therfore there must be something in this high wattage concept which we do not fully understand.
Obviously, I think all of us including me, need to go and further study, wattage and their requirements in accurate playback a lot more, before we say something which may be inaccurate.
I hope I have not offended anyone by this post,and if I have said something wrong please forgive my lack of knowledge.
Regards Deaf.

Good post, Deaf, and an excellent point raised.

I however wonder if the loudspeaker, when fed the Gazillion Watts can actually put out 110dB acoustic output, undistorted....

But I certainly agree, their is certainly a role for High Power... I am struggling on that count, with a Highly in-efficient speaker, and a 45 Watt tube amp ...

Compromises, compromises !

Amp_Nut schrieb:

Good post, Deaf, and an excellent point raised. I however wonder if the loudspeaker, when fed the Gazillion Watts can actually put out 110dB acoustic output, undistorted.... But I certainly agree, their is certainly a role for High Power... I am struggling on that count, with a Highly in-efficient speaker, and a 45 Watt tube amp ... Compromises, compromises ! :.

Dear Ampnut
I did not say anything about 110 acoustic watts,I said a piano does 0.5 acoustic watts when it is struck all out, which I guesstimate to be about 110db.
However your query about the gadzillion watts is more suited on a military weapons forum.

Regards Deaf.

deaf schrieb:

In regards to watts and loudness lets get a few apects in place.Loudspeakers are one of the most inefficient transducers known to exist.Their efficiency is 0.1%( direct radiators) to 5%(hifi horns).A piano at its loudest produces 0.5 acoustic watts,therefore a direct radiating loudspeaker with average an efficiency will require about 500-1000 watts to replicate the event.But we all measure in SPL don't we, and the chart says with average efficiency we need only a couple of watts, so why is a 1000watts required? Distortion reduction of all forms during playback is the key I think. All the components in a system should be at 10% of their full power to reproduce 110db,which means at 90db the system is operating at 0.1% of its abilities. Hence no distortion when 0.5 acoustic watts(guess work is about 110db) has to be reproduced momentarily;the system is still at 10% of full power.I didn't says this, some real gurus say this. Puk studios in Denmark has a custom monitoring system designed with 10000watts of amplification.Apparently it is one of the only systems in the world which comes close to the real thing.Before I get baulked at or laughed at by everybody else let me say this, that this setup was designed in partnership with the chief engineer of B&K, the worlds foremost microphone manufacturer, and noise and vibration measurement manufacturer.Therfore there must be something in this high wattage concept which we do not fully understand. Obviously, I think all of us including me, need to go and further study, wattage and their requirements in accurate playback a lot more, before we say something which may be inaccurate. I hope I have not offended anyone by this post,and if I have said something wrong please forgive my lack of knowledge. Regards Deaf.

don't know who this post was directed at specifically? Since it's right after my post, could it be directed at me?

anyway, Deaf, you are right on! The loudspeaker system is the most inefficient transducers to exist in the audio chain. This is confirmed when manuf come up w/ new & innovative ways to cool the tweeter - these days ferro fluid cooling is standard. A lot of work is being done on loudspeaker spider networks to play a double role in cooling + how many times have we read that the phase plug is made metallic so that it can play a double role of cooling the mid/woofer voice coil.
Also agree w/ the concept that one cannot have too much amplification power. I heard this concept put forth by Anthony Michaelson of Musical Fidelity. To that effect he is one of the few manuf offering 500W/ch & higher amplification units.
For me the issue is that I have not yet found a high power amp that sounds good on all kinds of music. Well, maybe 1: the Krell 600FPB mono monster at 600W/ch driving Martin-Logan Statement E2 speakers. For all the limited music demo'd that evening, everything sounded simply fantastic!

I have also noticed that some well designed & manuf studio speakers (Tannoy, Genelec, JBL, Yamaha to name just a few) perform better than any audiophile brand in the home setting. I also noticed that these speakers are rated from some serious wattage even tho their form-factor is small. This seems to further Deaf's (who is anything but deaf! ) point about 100s of watts being used in the studio environment for playback during engineering & mastering.

I wasn't aware that B&K was the premier microphone manuf in the world!
When you write B&K, you mean that company up in Buffalo, NY (I think that they got taken over by Klipsch recently?) that makes B&K HT receivers?

Dear Bombaywalla,
You are too sweet a chap to direct anything at,hence, no I wasn't directing at you, and never will.You are right with regards to new loudspeaker driver technologies,all directed towards higher power handling and also reducing distortions.
By B&K I meant BRUEL AND KJAER from Denmark. www.bksv.com.
However checkout www.pukstudio.com and get serious convulsions .
Deepest Regards Deaf.

deaf schrieb:

Dear Bombaywalla, You are too sweet a chap to direct anything at,hence, no I wasn't directing at you, and never will.

Thank you!

deaf schrieb:

You are right with regards to new loudspeaker driver technologies,all directed towards higher power handling and also reducing distortions. By B&K I meant BRUEL AND KJAER from Denmark. www.bksv.com. However checkout www.pukstudio.com and get serious convulsions . Deepest Regards Deaf.

thanks for the feedback - I knew that I had the incorrect company in mind!

My best Regards Deaf,
Bombaywalla

deaf schrieb:

In regards to watts and loudness lets get a few apects in place.Loudspeakers are one of the most inefficient transducers known to exist.Their efficiency is 0.1%( direct radiators) to 5%(hifi horns).A piano at its loudest produces 0.5 acoustic watts,therefore a direct radiating loudspeaker with average an efficiency will require about 500-1000 watts to replicate the event.But we all measure in SPL don't we, and the chart says with average efficiency we need only a couple of watts, so why is a 1000watts required? Distortion reduction of all forms during playback is the key I think. All the components in a system should be at 10% of their full power to reproduce 110db,which means at 90db the system is operating at 0.1% of its abilities. Hence no distortion when 0.5 acoustic watts(guess work is about 110db) has to be reproduced momentarily;the system is still at 10% of full power.I didn't says this, some real gurus say this. Puk studios in Denmark has a custom monitoring system designed with 10000watts of amplification.Apparently it is one of the only systems in the world which comes close to the real thing.Before I get baulked at or laughed at by everybody else let me say this, that this setup was designed in partnership with the chief engineer of B&K, the worlds foremost microphone manufacturer, and noise and vibration measurement manufacturer.Therfore there must be something in this high wattage concept which we do not fully understand. Obviously, I think all of us including me, need to go and further study, wattage and their requirements in accurate playback a lot more, before we say something which may be inaccurate. I hope I have not offended anyone by this post,and if I have said something wrong please forgive my lack of knowledge. Regards Deaf.

Deaf i don't agree with ur figures..the figures pertainig to the efficiency of horns and direct radiators....the from what i have read and know the figures are greater than waht u have quoted..almost by 10 fold..

try this site for this
http://www.the12volt...s.asp?TID=73962&PN=1

Apparently for a speaker Efficiency = 10E(Sensitivity in dB - 112)/10
this translates to

Eff(%)=Sens(db)
6.7=108
5.0=105
3.7=102
2.7=99
2.0=96
1.5=93
1.1=90
0.8=87
0.6=84
0.5=81

ie on an average only 1% of the power generated by an amp is getting congverted ti Audio (Generally speaking)

Now since a classic Class AB speaker is only around 60-85 % efficient (Compared to the 10-20 % of a class A)

the power consumed by the system is pretty high compared to what you get ( can assume a constant power of around 15-30 W by each non amplification component like Preamps/CDps and DACs)

Hence if the input power into a system is 500 W
only (500-50)*.65*.01 = 3 W (Approximately) is getting to be ouctual energy converted to power.

And this is at 1 m from the speaker

Dear Ben
I have said,from and upto,which means anything in between too.The electrostics of old were very inefficient hence my figures.Ofcourse PA horns can have upto 30-40% efficiency,but I think we are talking hi-fi here.As I said I could be wrong and a lot more study is required.
Regards Deaf.

As we can see Arj has really started his research,and his figures are spot on,also what I guessed was correct too.For 100% efficiency a driver has too have a sensitivity of 112db/1watt/1mtr.
Regards Deaf.