[jb26]

Hi All,

I'm trying to figure out what amount of watts from my amp I'm regularly using so that I can know if I need more power from my amp.

I have just bought a SPL decibel meter and am trying to work backwards from there to figure it out.

I tried using this website: http://myhometheater.homestead.com/splcalculator.html

And these figures:

Speaker sensitivity 87, speaker impedence 8ohm
SPL average 65, peak 77
Listening position 3m
Speakers against a wall

When I try working these figures through by adjust the 'amplifier power' (effectively X in the equation) I am at 1 or 2 watts.

Can someone tell me if I'm doing the maths wrong, or am I really drawing so little power from the amp at my usual listening levels? (Yes they're low, I have two kids under 4).

Thanks

James

[BobRex]

That's about right for a 10 foot listening distance and your SPLs.  Even if you listened at 87dB peak, you'd still only use 10 watts.

[SoCalWJS]

Which SPL meter are you using and what "weighting" are you using?
Would you describe your listening environment as Dead. Live, or Neutral?


.....but yeah, you're not using much power at all at those levels. The answers to the above questions might change it by a few watts, but doubtful it would even be 10.

[Ericus Rex]

That's about right for a 10 foot listening distance and your SPLs.  Even if you listened at 87dB peak, you'd still only use 10 watts.

Wouldn't 87dB from a speaker with 87dB sensitivity only be 1 watt of power?  What am I missing here?

The general rule of thumb that I've heard is to plan for power transients 10x your average listening level.  In your case it would appear that 20 watts is more than enough.

[SoCalWJS]

Wouldn't 87dB from a speaker with 87dB sensitivity only be 1 watt of power?  What am I missing here?

The general rule of thumb that I've heard is to plan for power transients 10x your average listening level.  In your case it would appear that 20 watts is more than enough.

Sound volume diminishes with distance.

Standard speakers (non line source) lose 3db in volume with each 1M of distance I believe, so he is losing about 9db by being 3M from his speakers.

[WC]

Wouldn't 87dB from a speaker with 87dB sensitivity only be 1 watt of power?  What am I missing here?

The general rule of thumb that I've heard is to plan for power transients 10x your average listening level.  In your case it would appear that 20 watts is more than enough.

Depends on the distance from the speakers. Close to the speakers you will be at 87dB, but as you move further away the spls will decrease.

[opnly bafld]

Sound volume diminishes with distance.

Standard speakers (non line source) lose 3db in volume with each 1M of distance I believe, so he is losing about 9db by being 3M from his speakers.

Depends on the distance from the speakers. Close to the speakers you will be at 87dB, but as you move further away the spls will decrease.

Have you ever measured this in your rooms?

[WC]

Not officially. I used an app on my iPad. The closer to the speakers the louder the SPL readings. It is what I would expect to happen.

[Letitroll98]

Have you ever measured this in your rooms?

Yes, I've found these figures to be rough approximations only, i.e. in a perfect world (room).  You don't drop quite the theoretical anechoic level because the room contributes reflected energy, what the amount and what the figures are I have no clue, but I believe that is the explanation, please feel free to correct any of my inaccuracies.

[mcgsxr]

I was playing around with my daughters ipad mini, and a free app for volume spl etc. 

Measuring movies from the 8-9 foot distance, I was seeing an average in the high 80's and peaks no higher than the mid 90's. 

My Paradigm speakers are reputed to have room sensitivity of 94db.  My Pioneer receiver is rated at 130wpc, but I expect realistic rating of 80 wpc with 5 speakers driven. 

I'd bet on around 10 watts being used most of the time.  Lots of headroom I guess. 

I have not registered a reading over 96 in that room, it just gets too loud for me.  I cannot imagine a THX reference volume in my room!

[sebrof]

And also don't forget that there may be more than 1 speaker

[richidoo]

The speaker impedance is not steady at 8ohms all the time. It will vary with frequency being played at that moment, and hit a low probably at the resonance of the reflex port, or up in the treble.  If you know the minimum impedance of the speaker then use that to calculate your power requirements. There are still variables like phase to consider, but that's splitting hairs and you are looking for a rough estimate anyway.  If you can't find the minimum impedance for your speaker, then assume min. is 4 ohms.

While sound does attenuate with distance, the attenuation varies with the freq. Higher freqs attenuate in shorter distance than low freq sound. Speakers are designed with a certain listening distance in mind, so this attenuation will leave just the right amount of treble left at the sweet spot.

For an 87dB speaker I'd prefer 100W amplifier for regular music playing at moderate levels. 200W would be better for the times when the family is out and you want to jam.  If 87dB is the manufacturer's published sensitivity rating, then I would derate that by a couple dB since most published sensitivity ratings are optimistic.

One last thing, an amplifier's distortion increases with power output. So if you speakers draw 100W at the peaks of music set at moderate volume, then you will be hearing more distortion  during those peaks, they will sound noticeably harsh compared to the smooth relaxing tone of the softer passages.  High power amplifiers remain relaxed and smooth-sounding even at high volume peaks because they are still way below their power limit, cruising at 1/10th power without a sweat and it sounds that way.   A 100W amplifier makes 100W just below the onset of clipping. Clipping causes the distortion to shoot up quickly, but just before clipping the distortion is usually already increasing. When the distortion starts to rise, the type of distortion is unpleasant sounding. So stressed amps sound harsh.  Ideally you want to stay below 0.5% THD distortion at all times with SS, or 1% THD with tubes. But with 87dB speakers playing moderately loud music with 100W will hit the redline often. If you figure you need 100W peak, then go for an amplifier that can make 200-300W so that the sound is always relaxed and smooth. You will have a lot more fun whether listening quietly or loudly. If the stereo sounds good loud, then everyone will want to listen loud because it is more fun. Kids love to dance.

Well designed solid state amps' distortion stays lower for longer before rising into clipping. You can get more watts out of them before the nasty sounding distortion is heard. Tube amps' distortion rises sooner, and more proprotionally with power, but the type of distortion is less offensive, so even at max power it probably won't sound as bad as a SS at full tilt. But tubes cost about double SS for same power. 

The good news is that there are a lot of very good high power amplifiers available now, especially used.

[JLM]

Several observations:

Rightly so, the discussion is centering on the peaks.  It's the normally the peaks that cause 'clipping' (overload of the system capacity to accurately follow the smooth waveforms that represent sound).  When system capacity is exceeded the top and bottom of the smooth sine wave are flattened out (looking like they've been clipped off).  Notice I said 'system' (any component in the chain of reproduction can overload, but like in this discussion we normally think of the power amp).  Clipped signals are especially hard on speaker drivers as they are physically trying to copy the signal by moving in and out but with clipped signals are forced to instantly stop and start (which causes various forms of damage to the drivers).

Most state that live jazz/classical peaks are around 105 dB and rock about 110 dB, but some insist that they're all closer to 130 dB.  Again most would say that typical dynamic swings run from 10 dB for rock to 30 dB for classical.  All this relates to just how 'fast' is your meter?  Are the instantaneous peaks being missed?  Many meters have a slow/fast option (always use 'fast').  In other words the actual peaks could be much higher than indicated by your meter, hence the above recommendations for bigger amps.  Note that how well an amp handles peaks is based on several design factors.  Tubes handle peaks more gracefully (rounding off the edges of the clips), most solid state not so much (a 'hard' clip).  Large capacity power supplies also help soften the clip.  OTOH realize that distortion from tube amps are lower at lower loads (where we listen) and are of the more agreeable even order harmonics where typical solid state designs display more distortion at lower level signals.  BTW Rich makes a good point about impedance varying with frequency and that the published nominal impedance specifications are thrown around too easily.

One factor in the above comments that hasn't been directly addressed in room gain.  My experience is that in a 'normal' room, say 12 ft x 20 ft with typical furnishings, the overall room gain plus consideration of the contribution from the 2nd channel (as per sebrof above) roughly equals the loss from listening more than 1 meter away.  So 77 dB peaks in a similar room from 87 dB/w/m efficiency speakers should only take 0.1 watts.  But frankly 65 dB average/77 dB peaks is very tame listening in terms of average and peaks, so I wonder about your numbers/meter.  (Most audiofests I've attended run roughly 80 - 85 dB on average.)  I'm an old mellow audiofart, but even I listen louder than 65/77 dB, and for anything beyond background/casual listening want my system to handle at least 105 dB for realistic classical/jazz peaks (in your case 160 watts per channel).

Here's a quick chart for converting dB of gain and watts, just fudge in something for room gain/distance/2nd channel and add to your speaker efficiency to compare watts to dB:

-30 dB   0.001 watts
-20 dB   0.01 watts
-10 dB   0.1 watts
   0 dB   1 watt
 10 dB   10 watts
 15 dB   32 watts
 20 dB   100 watts
 25 dB   320 watts
 30 dB   1,000 watts

[Ericus Rex]

Sound volume diminishes with distance.

Standard speakers (non line source) lose 3db in volume with each 1M of distance I believe, so he is losing about 9db by being 3M from his speakers.

But the first meter is within the spec, right?  87dB 1 watt/1 meter, so we're only talking about 2 meters diminution, right?  So that's 6dB and so since power must double for every 3dB of gain and since he's using less than 1 watt average we're talking under 4 watts at 3 meters.  Is this about right?

[sebrof]

But the first meter is within the spec, right?  87dB 1 watt/1 meter, so we're only talking about 2 meters diminution, right?  So that's 6dB and so since power must double for every 3dB of gain and since he's using less than 1 watt average we're talking under 4 watts at 3 meters.  Is this about right?

Yes, this is about right. The calculators change a little, people's listening levels change, room size and speaker sensitivity changes. But one thing that is constant is that when people plug in the numbers they are surprised at how little power they actually use.
Then people start talking about rock concerts and such and then everyone needs a 1000 watt amp.   

[jb26]

Hi All,

Firstly, thank you all for the knowledgeable contributions, I'm glad to hear that in terms of ballpark my maths is about right. 

A few comments:
- I am indeed listening at very moderate levels, it's due to kids in the room and little ability to shut off their bedrooms - our next house won't have the same issues!
- While the meter is certainly not 'professional grade', from the discussions here and the circumstances of testing I'm confident it's reasonably accurate.  For example, I had a crying kid scream at one point into the 80s, at a quick peak, and she was clearly louder than the music that was playing at the regular 65/75 levels.
- I'm amazed that I am using so little wattage

In terms of the advice through the various posts, here's what I'm taking away:
- that for most of my listening I could get away with a very low wattage amp
- that an amp is better running at a low percentage of capacity, so a larger amp will deal with my peaks better anyway
- that for future proofing, when my volume levels can be higer because the kids aren't as close to the stereo, a larger amp is better
- that about 150wpc would be the sweet spot to cover these issues off
- if I was to stay solid state an A/AB amp that ran A for about 5 watts, and AB up to 150 sounds would be great
- that until I need more volume consistently I can probably continue to get away with my 25wpc amp

Again, thanks all, and I'll be happy to hear more from you guys, this has been a great discussion so far, I've learnt a lot.

JB