• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

How much amplifier power is required?

Willem

Major Contributor
Joined
Jan 8, 2019
Messages
3,658
Likes
5,276
The claim that you really do not need that much amplifier power and that low power amplifiers somehow let you listen deeper into the music is one of those audiophool myths. I just found back the link to the Harbeth video of a demonstration some years ago of their M40.1 flagship speaker in the Netherlands: https://www.harbeth.co.uk/library/movies_hug/HAL-HV0515-01a_Power.mov
The monoblocks that were used had peak power meters, and on highly dynamic music these registered peak power output of more than 700 watts per channel. Conversely, at the same loudness but with some Corelli chamber music the maximum was only 28 watts.
 

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague
As always, some level of exactness in expressions is needed. All depends on speaker sensitivity, 82dB/2.83V/m sensitive speakers will need 100x more power than speakers with 102dB/2.83V/m sensitivity to make same SPL.
 
OP
W

Willem

Major Contributor
Joined
Jan 8, 2019
Messages
3,658
Likes
5,276
Sure, but the demo is a convincing enough anecdote (under ceteris paribus conditions, of course). Moreover, the Harbeth M40.1 have quite average sensitivity. And of course there is also room size etc.
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,580
Likes
38,281
Location
Gold Coast, Queensland, Australia
If I had to put a number on it, I'd say 200w/ch@8R, 350W/ch@4R. That will cover a broad range of speaker efficiencies and give you plenty of power for whatever mood you are in.

Back in the 70s, the minimum power suggested with efficient speakers was 30W/ch and typical upper range amplifiers were 100W-300W/ch. Not much has changed except speakers are now 5-10dB less efficient, so the upper numbers are where you should start, not the lower end. Especially if it's a hobby and speaker upgrades/changes are on the cards.
 

Daverz

Major Contributor
Joined
Mar 17, 2019
Messages
1,294
Likes
1,451
I found that Pan Sonic track on Youtube (could not find it on Qobuz)


So not sure of the fidelity there on YT, but turning it up as loud as I can tolerate (honestly not that loud as some of you, I'm sure, but I note that the guys in the video are able to converse without screaming), I never see the clip indicators come on my Bryston 3B-SST (250W into 4 Ohm). Speakers are Buchardt S400s (88 dB sensitivity, 4 Ohm nominal).

Pan Sonic does have several albums on Qobuz. Normally if I heard these sounds coming out of my speakers I'd be afraid of blowing a tweeter. :oops:
 

RichB

Major Contributor
Forum Donor
Joined
May 24, 2019
Messages
1,946
Likes
2,611
Location
Massachusetts
I found that Pan Sonic track on Youtube (could not find it on Qobuz)


So not sure of the fidelity there on YT, but turning it up as loud as I can tolerate (honestly not that loud as some of you, I'm sure, but I note that the guys in the video are able to converse without screaming), I never see the clip indicators come on my Bryston 3B-SST (250W into 4 Ohm). Speakers are Buchardt S400s (88 dB sensitivity, 4 Ohm nominal).

Pan Sonic does have several albums on Qobuz. Normally if I heard these sounds coming out of my speakers I'd be afraid of blowing a tweeter. :oops:

Yes, this track scares me. I played it years ago and suggest starting at low volume.
Here are the Stereophile measurements:

https://www.stereophile.com/content/harbeth-m401-loudspeaker-measurements

My estimate of the big Harbeth's voltage sensitivity on its tweeter axis was 84.5dB(B)/2.83V/m, which is within experimental error of the specified 85dB/2.83V/m.

I measured the power at 2.83 volts and measured the volume of sine-waves at my 12 foot listening position, then used a spreadsheet with my typical listening levels. The Benchmarks are 100 WPC into 8 Ohms and 180 WPC into 4 ohms when not bridged (as I am using them). I can clip music but only after turning it up to levels that I do not enjoy. With movies, I have played some bass heavy movies like Aquaman full range and the Salon2 woofers are moving quite a bit without illuminating the clip indicators. I have the option of bridging the AHB2's but, thus far, I don't find the need.

Measured SPL at Listening Position.jpg


So at this point, I am getting by with my meager wattage :p

- Rich
 

Wes

Major Contributor
Forum Donor
Joined
Dec 5, 2019
Messages
3,843
Likes
3,788
Sure, but the demo is a convincing enough anecdote (under ceteris paribus conditions, of course). Moreover, the Harbeth M40.1 have quite average sensitivity. And of course there is also room size etc.

There is also the impedance of the speaker as a f(freq.) and what music you try to play - cannon fire takes a bigger amp...
 

bunkbail

Addicted to Fun and Learning
Forum Donor
Joined
Feb 27, 2018
Messages
520
Likes
662
If I had to put a number on it, I'd say 200w/ch@8R, 350W/ch@4R. That will cover a broad range of speaker efficiencies and give you plenty of power for whatever mood you are in.
A little off topic but I want some explanation regarding power. I have an 200W@8R poweramp and on the official page of my speakers states that the recommended amplifier power is "25-120W". In this case, what happens if I do hit the 200W mark? Can I damage the drivers that way?
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,580
Likes
38,281
Location
Gold Coast, Queensland, Australia
A little off topic but I want some explanation regarding power. I have an 200W@8R poweramp and on the official page of my speakers states that the recommended amplifier power is "25-120W". In this case, what happens if I do hit the 200W mark? Can I damage the drivers that way?

On transients, unlikely. At a drunken party, certainly.
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,580
Likes
38,281
Location
Gold Coast, Queensland, Australia
I measured the power at 2.83 volts

Your spreadsheet may need some adjustment.

2.83V is 1 watt into 8 ohms. your speakers are closer to 4 ohms. (3-5 at bass where the power is). 2.83V at 4 ohms is 2 watts, twice as much. You need to double all your power numbers.

1578524306651.png
 

digitalfrost

Major Contributor
Joined
Jul 22, 2018
Messages
1,521
Likes
3,086
Location
Palatinate, Germany
- Buy an SPL meter
- Measure how loud you like to hear, typically
- Find the efficiency of your speakers
- Calculate how many watts are needed

I think power requirements are generally overestimated. There is a reason most consumer amps have around 100W and most people are fine with that. Also, this whole idea of needing more power for peaks is generally irrelevant for modern music since a lot of it is compressed to death anyway.

86dBC is plenty loud for most people and it's also a common efficency rating for speakers. Add 3dB for stereo and some room gain, you're gonna cruise well below 1W/channel for most of the time. Of course, large rooms, large listening distances can change this significantly, but I think this select group of people is rare, and they probably know who they are.
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,580
Likes
38,281
Location
Gold Coast, Queensland, Australia
I think power requirements are generally overestimated.

In my opinion and based on years of measurements of amplifiers, I'd say the complete opposite. Whatever power you think you need, double it.

That's where my 200W@8R number came from. And with most speakers now being nominally 4-6R, that's where my 350W@4R came from.

100W@8R amplifiers are easily and regularly clipped into 85-87dB speakers with dynamic, well recorded music during typical audiophile listening sessions. If you're happy with clipping and distortion on transients and really loud passages, 100W may be OK.
 

digitalfrost

Major Contributor
Joined
Jul 22, 2018
Messages
1,521
Likes
3,086
Location
Palatinate, Germany
Well you certainly know more about amplifiers than me. But, if that's the case and we all got these golden ears, how come nobody notices/cares? I just measured an audiophile compilation, dynamic range seems to be 14dB. I think we agree in this day and age, that's pretty good dynamic range.

Correct me if I'm wrong where, but if I am listening at 86dBC with 1W and I get a 14dB peak, that would require 25W right? I mean, I see how that escalates quickly if you need more than the 1W baseline, and hey at 10W we arrive at 250W peak, which is pretty much in the range you stated.

Well, I listen in the near field, I literary sit 1m away from my speakers. If I sat 2m away, it would be 6dB quieter, right? So I'd need 4W for the same power, and for peaks I'd need 100W?

Also, since +3dB needs a doubling of amplifier power, wouldn't it make more sense to buy speakers that are loud enough by themselves, or go with multiple subwoofers to have more power amplifiers at hand in the bass region? It seems to me, if affordable amplifiers don't have enough power, people got the wrong speakers in the first place.
 

RichB

Major Contributor
Forum Donor
Joined
May 24, 2019
Messages
1,946
Likes
2,611
Location
Massachusetts
Your spreadsheet may need some adjustment.

2.83V is 1 watt into 8 ohms. your speakers are closer to 4 ohms. (3-5 at bass where the power is). 2.83V at 4 ohms is 2 watts, twice as much. You need to double all your power numbers.

View attachment 45069

Correct. I did not completely describe the process here.
The XMC-1 volume setting of -31 produced 2.83 volts so that row is the anchor and was set to 2 watts to account for the 4 ohm load.

The 2.83 volts SPL played in stereo with room gain happens to match the speaker efficiency and I suspect it might be a good rule of thumb for many.

Measured SPL at Listening Position.jpg

These results were surprising given the 5400 cubic foot room volume.
Since every 3 dB doubles the power a couple missing the effect of stereo and room gain can greatly increase (by a factor of 4) the estimated power requirements. I was also able to illuminate the clipping levels with familiar (compressed) recordings at power levels that were in-line with the spreadsheet estimates.

- Rich
 

HammerSandwich

Major Contributor
Forum Donor
Joined
Nov 22, 2018
Messages
1,137
Likes
1,497
- Buy an SPL meter
- Measure how loud you like to hear, typically
- Find the efficiency of your speakers
- Calculate how many watts are needed
Alternatively, try the DIYAudio Voltage Test.

...regularly clipped into 85-87dB speakers with dynamic, well recorded music during typical audiophile listening sessions.
Yes, high-DR tracks are the killers, as hinted in the OP. Because only a little music qualifies, I don't hit my system's top 6dB very often, and a lot of my friends would never play those tracks.

@digitalfrost, you're on the right track. But "how much is required" scales to the worst-case scenario, not "in this day and age, that's pretty good dynamic range." Off the top of my head, I checked Sweet Sue from Swing Live. It peaks at -0.4dBFS with RMS at -26dB.
 

RichB

Major Contributor
Forum Donor
Joined
May 24, 2019
Messages
1,946
Likes
2,611
Location
Massachusetts
- Buy an SPL meter
- Measure how loud you like to hear, typically
- Find the efficiency of your speakers
- Calculate how many watts are needed

I think power requirements are generally overestimated. There is a reason most consumer amps have around 100W and most people are fine with that. Also, this whole idea of needing more power for peaks is generally irrelevant for modern music since a lot of it is compressed to death anyway.

86dBC is plenty loud for most people and it's also a common efficency rating for speakers. Add 3dB for stereo and some room gain, you're gonna cruise well below 1W/channel for most of the time. Of course, large rooms, large listening distances can change this significantly, but I think this select group of people is rare, and they probably know who they are.

Just to clarify the point, popular music is not just compressed but often slammed and clipped at maximum volume. These tracks simply cannot get any louder.

Here is Adele's Skyfall recorded at 96kHz/24Bits bought from HDTracks (in case you thought this was a problem for 16 bit audio only):

AdeleSkyFallDetails.jpg
AdeleSkyFallAudacity.jpg


HD Audio, my tuccas. :p

- Rich
 

RichB

Major Contributor
Forum Donor
Joined
May 24, 2019
Messages
1,946
Likes
2,611
Location
Massachusetts
Well you certainly know more about amplifiers than me. But, if that's the case and we all got these golden ears, how come nobody notices/cares? I just measured an audiophile compilation, dynamic range seems to be 14dB. I think we agree in this day and age, that's pretty good dynamic range.
Correct me if I'm wrong where, but if I am listening at 86dBC with 1W and I get a 14dB peak, that would require 25W right? I mean, I see how that escalates quickly if you need more than the 1W baseline, and hey at 10W we arrive at 250W peak, which is pretty much in the range you stated.

DR14 records are good these days. My method measured SPL at my listening position using full scale sign wave files (at maximum volume). I used 250Hz, 1kHz, and 4kHz tones and they were essentially the same so I went with 1kHz. If you are listening a 86dB that is already loud.

Well, I listen in the near field, I literary sit 1m away from my speakers. If I sat 2m away, it would be 6dB quieter, right? So I'd need 4W for the same power, and for peaks I'd need 100W?

Not exactly because of room gain and stereo (twice the speakers).
Here is a decent SPL calculator:
https://myhometheater.homestead.com/splcalculator.html

Here is what happens when I plug in the values for my room and the (4 Ohm) Salon2s (there not really in the corners but it gives the most gain):

Salon2SPLCalculator.jpg


Here are the power computations based on my 2.83 volt level SPL measurements:

Measured SPL at Listening PositionClipping.jpg


Very close. Since the AHB2 have clip indicators also back up these power estimates.

- Rich
 
OP
W

Willem

Major Contributor
Joined
Jan 8, 2019
Messages
3,658
Likes
5,276
What all this shows is the importance of power to reproduce peak levels. I listen to a lot of classical music and much of that has and is recorded with big dynamic range. I also use my stereo system to provide audio for home theatre, and there dynamic range can be even bigger.
Fortunately these days powerful amplifiers can be cheap: I recently bought a 2x250 watt Yamaha P2500s for just 300 euros. It sounds absolutely fine, and that was to be expected from the excellent AP measurements of its bigger brother, the 2x350 watt P3500s that sold for only 350 euros.
 

JRG

Member
Joined
Oct 2, 2019
Messages
48
Likes
91
Location
La Rioja, Spain
Not everyone needs high levels of power for their domestic hi-fi systems.
Fundamental factors are the size of the listening room and if we have neighbors to disturb,...

In Spain most people live in flats of about 80 to 120 m2 with neighbors on the other side of the wall, the floor and the ceiling. And the music equipment is usually located in the living room, which is around 20 to 35 m2 without a real soundproofing.

In that common situation you can´t listen at loud levels without disturbing the neighbours and cannot take advantage of big power amps.
Obviously there are people who live in single-family homes with big dedicated listening rooms that do not have to worry about neighbors ... but, at least in Spain, that is the exception, not the norm.

For example, I have a second hifi system in a 15m2 room (5 x 3 m) with a pair of Klipsch RP 160M speakers and I'm really left over with 35W and even less.... because it is impossible to maintain pressure levels of more than 90-95dB at the listening point without disturbing the neighbors... and that pressure is achieved with very little power in 15m2 with a standard 90 dB/W/m speakers that most people use.

And my main system wich is in the living room has the same problem: 35m2 and neighbors across the wall. I just can't listen to music loud enough to the point of needing a lot of power. Very rarely pass half the pot in my Yamaha AS501,... and most the time is at just 25-25% with my Tannoy Revolution XT 8F.

In short,... i think that people who live in apartments with neighbors and without soundproof rooms does not need a lot of power to listen at reasonable levels.
The case of the Harbeth 40 video is exceptional in the sense that their are using a really big dedicated listening room in comparison of most home hifi systems (at least in Europe, where people live in flats as i've commented)
 
Top Bottom