How many watts is enough? (1 Viewer)

[Kevin. W]

Don't most recievers in the 70-90w/channel reach reference level sound? If your speakers are 80watts do we really need a reciever or AMP that produces 120w+?
Kevin

 

[Chu Gai]

you need to consider the speaker's sensitivity in conjunction with your room environment. please provide more details concerning what it is that you're getting at.

 

[Bob McElfresh]

Well, here is the issue:
It takes LOTS more power to drive the woofer and produce sounds at 80/70... hz than it does for sounds above ... say 300 hz.
So if you have speakers with only tweeters and mid-range drivers, and a external sub (with it's own power supply), then 70-80 wpc is fine.
Note: for two years I ran my HT system, 2 DefTech 2000TL's with built-in woofers, 2002 center and 2 bookshelf rear speakers with a 80 wpc Yamaha 793. When I upgraded to a Yamaha DSP-A1, not a lot of difference for movies.
But when you dont have a external sub, have full range speakers (with woofers) and then play the lobby shoot-out scene on The Matrix at a interesting volume, well a lot of people were posting "My receiver said 'PROTECT' and turned off!"
Even a 80 wpc receiver will produce good bass for several seconds. But it heats up. After a few minutes, the receiver should shut itself off to prevent damage. (The lobby shoot-out scene is several minutes of intense bass).
This is where a name-brand receiver that is rated at 120 wpc for all frequencies 20-20,000 hz with all 5 speakers being driven is nice to have. Or an external sub and all 5 speakers set to SMALL.
Does this help?

 

[Kevin. W]

My current system setup is:
Denon 1801(pre/pro)
Rotel RMB-1066(6x60, 5x70, or 3x150 bridged)
Paradigm Mini Monitors x2, CC350, PS1200 Sub
Basically I have decided to move to seperates and begun with getting the new Rotel RMB-1066. I picked this AMP because of it ability to drive 6 channels or be bridged into 3x150. I understand that in bridged mode this is more than enough power for my 80watt Paradigms, but I remember reading, its better to have an AMP with more juice than one without to prevent clipping. I watch movies at a volume of -30 to -25 and don't plan on upgrading my speakers anytime soon. I have really noticed a difference in sound quality since getting the AMP. I hear more detail than I did before. But anyway, I guess I trying to figure out if I should bridge the AMP or not to make sure the speakers get enough juice without damaging them in the long run.
Kevin

 

[Todd Hochard]

quote: I watch movies at a volume of -30 to -25 and don't plan on upgrading my speakers anytime soon.[/quote]At that volume level on the Denon, the amp is putting out

 

[AjayM]

quote: It takes LOTS more power to drive the woofer and produce sounds at 80/70... hz than it does for sounds above ... say 300 hz.[/quote]
This is mostly false. Which uses more power to achieve lets say 100db (let's pretend there is no x-over loss, etc and in an anechoic chamber)
12" Woofer spec'd at 95db@1w/1m
1" tweeter spec'd at 85db@1w/1m
A speaker system is rated with a sensitivity number, like 90db@1w/1m that means the whole system will play at that volume, the woofer doesn't need more power or the tweeter less, etc.
Now when I said mostly false, I believe this misconception happens because bass usually has bigger peaks than the mids and highs do, especially with movies (big explosions), etc. So if all the sudden the bass level is 10db louder you're going to need more power to the woofer.

But when you dont have a external sub, have full range speakers (with woofers) and then play the lobby shoot-out scene on The Matrix at a interesting volume, well a lot of people were posting "My receiver said 'PROTECT' and turned off!"
Even a 80 wpc receiver will produce good bass for several seconds. But it heats up. After a few minutes, the receiver should shut itself off to prevent damage. (The lobby shoot-out scene is several minutes of intense bass).

A reciever will go into Protect mode when it's over-driven or is over-heating. Over-heating is easy to cure, especially since almost every picture I see of people's home theater's the reciever has only a 2-4" of space above it and is usually covered on 3-4 sides (bottom, sides and sometimes the back) that leaves very little room for ventilation. Most reciever manufactures recommend like 12" of space above the reciever for the heat to dissipate.
Over-driving, well a 70wpc amp isn't going to be able to cut it if you want to listen loud, say you are listening to the Matrix on your 70wpc reciever, hooked up to your 85db@1w1m efficient speakers...but you're listening loud, let's say it's averaging around 85db (plus it makes this easier to explain)....
85db 1watt
88db 2watts
91db 4watts
94db 8watts
97db 16watts
100db 32watts
103db 64watts
106db 128watts
So you can see if you have a large transient you can run into problems. Of course this isn't taking into account that most people sit more than 1 meter away from their speakers, and for a cone driven speaker you lose I belive 6db for every meter you are away from it, so if you sit 3 meters away (around 9ft) your 85db speakers are really only 73db with 1 watt (at 3 meters), so to get back to your original 85db range you're using 16watts of power, at 64watts of power you're only at 91db...to exceed the 70w reciever and jump to 128watts you're only at 94db.
It's very easy to over-drive a small amp like this.
Here's a good small read on transients and amp power in regards to music that came over an email list yesterday.
Art Ludwig looked at three pieces of music that he likes to listen to, and found a peak power to RMS ratio of about 30:1 to 40:1, see;
http://www.silcom.com/~aludwig/EARS....nd_loud_sounds
on his site. Even with a 100dB/1W efficiency, giving you a, say, 90dB eff in the listening position, you'll need 40W to cover the peaks when listening at an RMS level of 90dB (atleast for the three pieces of music he tested with).
A more normal example, I loose about 10dB eff to the distance between the listening position and my speakers, my speakers have a 92dB eff, and I like to sometimes listen at an RMS level of ~95dB. It seems I would need about 20W for the RMS level and up to 800W to cover the peaks. My current
amp only gives me 85W, so in theory I have to keep the RMS SPL in the listening position at a lowly ~85dB or below, due to lack of power.
Bottom line - big enough is .. very big! A whole lot bigger than I thought before reading that page on Art's site.
Hope this helps some more.
Andrew
------------------
[Edited last by AjayM on September 07, 2001 at 12:56 PM]

 

[Yumbo]

here's the J6P spiel:
depends on the size of your room, after matching speaker capacity. efficiency also comes into effect.
think of it like your car:
small room, small engine
big room, big engine
big engine, big capacity (speakers)
both will give you better efficiency on a relative scale.
when you rev a small engine, it makes more noise but will sound 'quick'.
when you rev a big engine, it will purr, and may seem 'slow' but will be a more pleasant ride.
quick = punchy, bright (for boombox movies)
slow = silky, warm (for subtle movies as well)
essentially, sounds will be more natural, detailed and hardly distort. your light bill also goes up, lol.
common sense and the size of your budget should prevail.
eg. I used a 70W amp in a small room, and a 100W amp in a big room - they sound different, but both are good in their own environment.
I'm just a bit perplexed at people listening at -25 and -10 and 0! whats's going on?
reference on a good amp should be no more than -28 (10 o'clock), average listening at -40 (9 o'clock).
otherwise, your amp is poor or speakers aren't matched.
------------------
Yumbo - IMDVD

 

[AjayM]

I'm just a bit perplexed at people listening at -25 and -10 and 0! whats's going on?
reference on a good amp should be no more than -28 (10 o'clock), average listening at -40 (9 o'clock).

Where did you get that idea? I guess first what kind of reciever are you using? My denon goes past 0 and up into a fair amount of positive numbers and the lowest it will go is around 40-42 (hardly average listening level).
And the reference to the clock for the volume adjustment only works on some recievers. But set reference for where-ever you want to.
But achieving reference volume has nothing to do with the little number on your display or where your volume knob is at, to make so many db takes so much power. If you crank the knob down, you end up going into the speaker calibration area and cranking up all those adjustments, or vice versa...the end result is still the same.
Andrew
------------------

 

[RicP]

reference on a good amp should be no more than -28 (10 o'clock), average listening at -40 (9 o'clock).

This is completely dependant on your PRE-AMP, not your amp, and the display has nothing whatsoever to do with how much SPL a system is putting out.
Just pay no attention to what your pre-amp/reciever volume knob says and use a Sound Pressure Level meter to determine what "volume reading" your system produces reference level at.
Alos speaker wattage ratings are essentially meaningless. The Ohm rating and the sensitivity are the important numbers.
It will take more Wattage (on average) to power low Ohm (low resistance), low sensitivity speakers to the same SPL's as a high Ohm, high sensitivity speakers being powered by a smaller amp.
------------------
http://www.ricperrott.com
Ric Perrott - My DVD's

 

[Todd Hochard]

and the display has nothing whatsoever to do with how much SPL a system is putting out.

...unless, of course, you calibrate reference level (in my case, 75db from Video Essentials) to correspond to the 0db indication on the front panel. With my setup, it was a breeze to do.
I can't imagine that I would ever listen to a movie or music at anything above reference level (the 75db "average" I stated above) in my room. After all, I like having my hearing.
Todd

 

[RicP]

...unless, of course, you calibrate reference level (in my case, 75db from Video Essentials) to correspond to the 0db indication on the front panel. With my setup, it was a breeze to do.

Sure, IF you're capable of doing that. I couldn't get my system to register 75db at 0db on the dial if I tried, it's much too loud. I can only kick each channel down 10db in the calibration, and I register over 88db at 0db with all channels zeroed.
But anyway, the point is it's virtually impossible to say to someone that reference levl is at "x" setting on your volume knob...there are just WAY too many variables.
------------------
http://www.ricperrott.com
Ric Perrott - My DVD's

 

[Bob McElfresh]

(Ahhhh... It begins )

AJ: I'm afraid I must stand by my assertation that it takes more power to drive a woofer than a tweeter.

Just looking at efficiency numbers does not allow you to conclude that

the woofer doesn't need more power or the tweeter less, etc

You seem to be saying that it takes nearly the same ammount of power to produce a sound at any frequency.

I'd like to challenge this (if you are up to a friendly debate ) by asking you a few questions:

• Why do external subs come with their own power supplies that range from 200-2,500 watts, but the mid-range and tweeters are not self-powered?
• Those personal "shreeker" alarms produce 120 db of sound from a 9-volt transistor battery. If you took out the little tweeter and wired in a woofer, how much sound would you get? Why or why not?
• When you look at receiver power ratings, why are the more reputable brands careful to report watts-per-channel from "20 - 20,000 hz" if it takes the same power to produce a sound at any frequency?
• Why is a 4 ohm speaker more likely to over-heat a receiver than a 8 ohm speaker? Why?[/list=a]

 

[AjayM]

quote: Just looking at efficiency numbers does not allow you to conclude that
the woofer doesn't need more power or the tweeter less, etc[/quote]
It absolutly does. Let's drop the whole transient's side of the equation and use a simple test tone. Individual speakers are rated for effeciency just the same as whole speaker systems. So in order to make your woofer and tweeter produce sound at the same volume you need to match them up (or add an L-pad into your x-over). Let's look at two speaker systems.
Woofer rated 85db 1w/1m
Tweeter rated 95db 1w/1m
If you don't add something to the tweeter to match up the level to the woofer your tweeter will always play 10db louder (and won't sound good).
Woofer rated 90db 1w/1m
Tweeter rated 90db 1w/1m
Both the woofer and tweeter play at the same volume no matter how many watts go through it.
The amplifier doesn't know to send 10db more power to the woofer than the tweeter, it's not that smart.
quote: You seem to be saying that it takes nearly the same ammount of power to produce a sound at any frequency.[/quote]
Sure does if your speakers are rated the same.
quote: I'd like to challenge this (if you are up to a friendly debate ) by asking you a few questions:[/quote]
I'm always up for friendly debate.
quote: Why do external subs come with their own power supplies that range from 200-2,500 watts, but the mid-range and tweeters are not self-powered?[/quote]
One has an amp screwed to the back and the other has an amp on the shelf. The speaker doesn't power itself. As to why most subs use their own amps, well first most people have a single sub, so you'll have a fun time trying to recreate a mono signal without something to sum the channels together if you tried to hook them to your main speakers.
Also we come back to the effeciency side, an average 12" woofer is much less sensitive than a mid or tweeter. Take my newest DIY speaker, it's using a pair of Focal 7k4411 speakers and a Focal TLR tweeter, this system is rated at 95db effeciency, find me a sub woofer that will match that. I could use a pair of Shiva's (about 88db effecient), but that drops me down to a 2-Ohm load. The new Adire Maelstrom is rated at 98db effecient which wouldn't be a bad match and could probably dropped in with a minimal amount of level matching (but a lotta x-over work, and a huge cabinet).
If you want to talk power levels, the amps attached to my mains is rated around 375-400watts per channel, the amp on my Tempest sub is rated at 250watts per channel. My system is properly calibrated using test tones, VE and Avia and is pretty flat from about 20Hz up past 20kHz and it will get very, very loud...and the sub with almost half the power keeps up just fine well into the 100's db wise.
So why do you see these subs rated with silly amount of power? I would bet dollars to donuts that it has mostly to do with EQ work (another can of worms that is past the scope of this thread topic)....or it has to do with Hoffman's law (a sub is a compromise between space, extension and efficency...pick 2).
quote: Those personal "shreeker" alarms produce 120 db of sound from a 9-volt transistor battery. If you took out the little tweeter and wired in a woofer, how much sound would you get? Why or why not?[/quote]
A transistor batter? Those tiny piezo speakers are hugely effecient, I wouldn't be suprised if there was less than 1watt of power going to it. So let's say you wired in an 8ohm woofer, and lets say the thing makes 1w @ 8ohms, you will get sound, if the woofer is rated at 85db 1w/1m you will get 85db of sound at 1 meter.
quote: When you look at receiver power ratings, why are the more reputable brands careful to report watts-per-channel from "20 - 20,000 hz" if it takes the same power to produce a sound at any frequency?[/quote]
This has nothing to do with a speaker, and opens another whole can of worms. If the amp's power fluctuates over the frequency spectrum than you will have a very uneven sound. Let's do an outrageous example.
Cheapy amp A makes 50watts of power at 30Hz, 100watts of power at 1500Hz and back to 50watts at 15,000Hz, this means your speaker system will be 3db lower in volume at 30 and 15,000 hz (assuming a perfectly flat speaker response).
quote: Why is a 4 ohm speaker more likely to over-heat a receiver than a 8 ohm speaker? Why?[/quote]
This is a function of the amp again and not the speaker. The amp needs more current to create the same power at a lower impedence (sort of a generalization, but applies to the "average" amp), again nothing to do with the speaker.
Alot of this is based in theory anyways, none of us listen to test tones, etc. Also the enclosure and x-over have things to do with the effeciency of a speaker as well, so does the room the speaker system sits in. And there's probably a lot of things I didn't explain very clearly...the benefits of staying up to late.
Andrew
[Edited last by AjayM on September 07, 2001 at 11:08 PM]

 

[Bob McElfresh]

Darn, you have some good reasons.
Lets try this....
You said:

The amp needs more current to create the same power at a lower impedence (sort of a generalization, but applies to the "average" amp), again nothing to do with the speaker.

This is not true.
The speaker IS the impedence in the circuit.
Your speaker is a coil of wire. And a coil of wire has an impedence.
But if you hooked a ohm meter to a speaker with a 8 ohm impedence, will it read 8 ohms? No. It will read nearly 0.0 ohms. Thats because to DC, a coil of wire is a direct short.
Fact: Impedence is not a DC phenonom, but an AC (like a audio signal).
When you start to make the DC signal an AC signal, the impedence increases. As the frequency goes up, so does the impedence. (At some high frequency, the coil of wire will act like a broken wire and not let any current flow.)
Fact: impedence is FREQUENCY dependent.
Fact: low frequency into a coil is nearly a dead short and draws lots of current. High frequency into a coil has lots of impedence.
Fact: Power = Voltage X current.
Conclusion: low-frequency signals draw more power than higher frequency signals from your amp.
Second argument: Put your hand in front of a tweeter while it is producing sound. Can you feel the air moving? No.
Put your hand in front of the port on a sub while it is producing sound. Can you feel the air moving? Yes.
Since power is the ability to do work and moving air takes work, the driver that moves more air is doing more work.
And more work takes ... MORE POWER! (I feel like Tim Allen )

 

[Jeff Loughridge]

Bob,
A complete, succinct, and 100% accurate explanation. It takes a lot of power to move air, and at 20Hz to 120 Hz, the characteristic impedance of the driver is very low.
This can also be demonstrated by measuring the current drawn by an amplifier, while producing the same SPL. Using a full range speaker, the current needed to produce 90 dB at 20 Hz vs. 10 kHz is substantial. Using a slow sweep, you can watch the current draw quite easily.
I must admit I don't know if driver efficiency is measured with a steady state tone or pink noise, but if it is a full range speaker it seems to me it should use pink noise.
------------------
Jeff Loughridge
Director of Engineering
WPGC/WHFS
Infinity Broadcasting Corp
[email protected]
[Edited last by Jeff Loughridge on September 08, 2001 at 11:14 PM]

 

[Steven Lin]

While the above is true for a pure inductor, the impedance of speakers usually doesn't monotonically increase with frequency. It usually has dips and peaks throughout the audio range.

 

[Rick P.]

Depending on our point of view, you're both right.
quote:
--------------------------------------------------------------------------------
You seem to be saying that it takes nearly the same ammount of power to produce a sound at any frequency.
--------------------------------------------------------------------------------
Sure does if your speakers are rated the same.
There's the catch. In audio amp terms, 1 watt is defined as an input of 2.83 V (which equals 1 watt at 8 ohms). Voltage is used as a reference because the impedance of a loudspeaker changes with frequency, and therefore so will the power dissipated in the load. As Bob pointed out, the impedance of a tweeter (or any coil) increases with frequency, therefore the current through the coil is going to drop and electrical power decreases. However, the voltage output across the speaker terminals of an amplifier should be constant across the entire audio range (hence the 20 HZ - 20 kHz bandwidth power rating of good amps) at any given volume setting. When a tweeter has an output rating of 93 dB @ 1 watt & 1 meter, it really means is that the output is 93 dB given an input voltage of 2.83 V. What is the actual power dissipated in the tweeter? Who knows - it depends on the frequency, characteristics of the speaker and crossover network, resonance point, etc ...
So, while the electrical power required to drive your tweeter may drop, the voltage required by your tweeter (and crossover network) to maintain a given output level will be the same as that required by your low-impedance woofer. And since you can't get a low-power consumer amp with a high voltage output (like you can with tube amps), you're stuck with having to buy a high-power amplifier just for it's higher output voltage.
You could of course re-design the crossover network and bi-amp / tri-amp your speakers to allow you to use a much smaller amp for your high-frequency drivers, but this is isn't something your average joe six-pack is likely to do.
Rick
[Edited last by Rick P. on September 09, 2001 at 10:56 AM]

 

[AjayM]

So what happens if I have a ribbon speaker or something like a Maggie, that has a flat 4Ohm load across it's frequency band?
And to throw a little more into the mix, let's take my sealed Tempest subwoofer. I haven't measured it, so I'm using data from the LspCAD sim, but it remains a pretty constant 4.7-5Ohm load across the frequency spectrum, except at 31Hz there's this peak at 33Ohms.
Now look at my tweeter, Link Removed
There's a small bump at the fs point (the x-over is an octave above that), and it's pretty flat up into the high teens at around 8Ohms.
I don't see these varying ohm readings at specific frequency's (other than the woofer's peak at 31Hz).
Rick gave a good explanation about the whole speaker system, if you have a 93db 2.83v/1m speaker system, if you apply 2.83v to it all the drivers will play the same volume, if you quadruple the voltage going into it all the drivers play the same volume. Now the x-over may have some resistors and such mixed in with the tweeter to play the same level as the woofer, but it may not (the x-over in my latest DIY adventure has no resistors in it, yet anyways).
Andrew
[Edited last by AjayM on September 09, 2001 at 11:31 AM]

 

[Rick P.]

So what happens if I have a ribbon speaker or something like a Maggie, that has a flat 4Ohm load across it's frequency band?
If you have a speaker with a uniform impedance plot, then you can calculate power as P = V^2 / z, where z is your 'constant' impedance. You bring up an interesting point about Maggies. Although the various planer drivers in a Maggie are not exactly 'flat' at 4 ohms, each panel is essentially a resistive load of about 4 ohms. As I recall, the high-frequency panels in a Maggie actually have a lower resistance than the bass panel so they will dissipate more power.
And to throw a little more into the mix, let's take my sealed Tempest subwoofer. I haven't measured it, so I'm using data from the LspCAD sim, but it remains a pretty constant 4.7-5Ohm load across the frequency spectrum, except at 31Hz there's this peak at 33Ohms.
Now look at my tweeter, Link Removed
There's a small bump at the fs point (the x-over is an octave above that), and it's pretty flat up into the high teens at around 8Ohms.
OK. Your Tempest has a sensitivity of about 92 dB @ 2.83 volts and a power rating of 750 watts RMS, while your tweeter has an efficiency of 95 dB @ 2.83 volts with a power rating of 15 watts RMS. At 8 ohms, your tweeter can be driven with constant V = sqrt (15*8) = 11 volts to produce an output level of 95 + 10 log (11/1) = 105.4 dB. To produce the same output level, your tempest would need ALog ((105.4-92)/10) = 22 watts. If you were listening to white noise, you would quickly enjoy the sound of silence, a waft of blue smoke and a funky smell from your tweeter if you tried to drive it beyond this level. Your Tempest, on the other hand, could go to 120 dB or so in the right enclosure.
This might seem like quite a discrepancy, but it really isn't a problem if you consider the power spectrum of 'normal' sounds. Looking at music or movies through a frequency analyzer, you'll notice that there isn't a lot of content above 3500 Hz or so. Given this, you can safely use your tweeter's maximum power rating of 150 watts instead of its RMS rating in the above calculation and see that it can now produce output levels up to 117 dB. Instantaneous power dissipation will be quite high, but the RMS power dissipated will still be less than its 15 watt RMS rating.
It's fairly common knowledge that you are more likely to blow a tweeter with an underpowered amp than with one that's too powerful. The reason for this is that an amp in clipping introduces a lot of high-frequency information (required to produce a square wave), which easily exceeds the driver's RMS power rating.
So ...
How many watts is enough? If you're using a single amp, then 'enough' watts could be defined as the power required to exceed the maximum (not RMS) power rating of all your drivers. You'll probably never use it, but you will have enough headroom for really dynamic scenes. Too much power isn't a problem if you're sensible with the volume control. It will also leave you 'room' for the future when you get a big raise and can afford really inefficient speakers.
Rick
------------------

 

[AjayM]

quote: OK. Your Tempest has a sensitivity of about 92 dB @ 2.83 volts and a power rating of 750 watts RMS, while your tweeter has an efficiency of 95 dB @ 2.83 volts with a power rating of 15 watts RMS. At 8 ohms, your tweeter can be driven with constant V = sqrt (15*8) = 11 volts to produce an output level of 95 + 10 log (11/1) = 105.4 dB. To produce the same output level, your tempest would need ALog ((105.4-92)/10) = 22 watts.[/quote]
Right, exactly double the wattage to compensate for the 3db more effecient tweeter. If the Tempest were 95db effecient and the TLR tweeter was 92db effecient then the woofer would need less power to drive the same level as the tweeter, which means that bass (in this scenario) doesn't need more power than the highs.
quote: This might seem like quite a discrepancy, but it really isn't a problem if you consider the power spectrum of 'normal' sounds. Looking at music or movies through a frequency analyzer, you'll notice that there isn't a lot of content above 3500 Hz or so. Given this, you can safely use your tweeter's maximum power rating of 150 watts instead of its RMS rating in the above calculation and see that it can now produce output levels up to 117 dB. Instantaneous power dissipation will be quite high, but the RMS power dissipated will still be less than its 15 watt RMS rating.[/quote]
Exactly my point all along, for music/movies (Where there is more lower frequency energy than high) you use more watts for bass than highs, but that's because the volume is higher.
Edit: To go with the above (hate posting in a rush). Bass may have a higher volume, but that doesn't always mean that you will use more watts, the effeciency of a speaker system has to come into play also, an Adire Maelstrom has an effeciency of 98db, if your main speakers were rated at 80db you have a ridiculous 18db difference in there (or you just used 64watts to get to the same level). There's more to a system than just wattage, if you had a set of 110db sensitive speakers than a 50watts reciever is all you will ever need, if you have 80db speakers that 50watts isn't going to cut it.
So how many watts is enough? More is always better (highly scientific isn't it?).
Andrew
[Edited last by AjayM on September 09, 2001 at 10:20 PM]