Low powered amps and speaker damage

[J Larsen]

Hi,

I have recently purchased a 40W tube amp (PrimaLuna ProLogue 2) to power my B&W 704s, which have a sensitivity rating of 90dB. The PrimaLuna is replacing a 60W SS amp which has malfunctioned.

I know that 40W is lower than what most people would use to supply the B&Ws. Can you tell me if I am thinking about this the right way?

A) 90dB sensitivity means the speakers require 1W to produce 90dB as measured 1 foot away.

B) I do not want to attempt to drive my speakers at a volume that would exceed the power capabilities of the amp, because that would lead to clipping and possible speaker damage.

C) 32W is sufficient power for the B&Ws to produce 105dB as measured at 1 foot, because 32 = 2^5, and 5*3db = 15dB. 90dB+15dB=105dB.

D) as long as I don't try to play music louder than 105dB, there is no risk of speaker damage.

E) 105dB would get me quickly evicted from my NYC apartment, so I have nothing to worry about.

Is this logic sound?

Thanks in advance.

[Æ]

90 dB at one meter, not one foot.

One of the nice characteristics of just about every tube amp is that they clip gracefully. No real worries about loudspeaker damage. Furthermore, tube amps often have more real power to begin with.
Say for instance you have a nondescript solid state amp that can deliver 100 watts into an 8 ohm load at 1KHz (per industry and government standards). Then drive that same solid state amp with a full range signal or pink noise, all of a sudden that 100 watt solid state amp can only deliver something like 25-30 watts per channel. Of course it kind of depends on the amp, but ultimately solid state amps clip far more easily. Some people don't fully appreciate what I just said, but Henry Wolcott easily demonstrated this to me with my own KENWOOD L-07MII 150 watt solid state monoblocks. My amps would deliver well over 150 watts into 8 ohms at 1KHz and something like 270 watts into 4 ohms, but when driven with pink noise into an 8 ohm load, they crapped out at maybe 35 watts.

Originally Posted by J Larsen

Hi,

I have recently purchased a 40W tube amp (PrimaLuna ProLogue 2) to power my B&W 704s, which have a sensitivity rating of 90dB. The PrimaLuna is replacing a 60W SS amp which has malfunctioned.

I know that 40W is lower than what most people would use to supply the B&Ws. Can you tell me if I am thinking about this the right way?

A) 90dB sensitivity means the speakers require 1W to produce 90dB as measured 1 foot away.

B) I do not want to attempt to drive my speakers at a volume that would exceed the power capabilities of the amp, because that would lead to clipping and possible speaker damage.

C) 32W is sufficient power for the B&Ws to produce 105dB as measured at 1 foot, because 32 = 2^5, and 5*3db = 15dB. 90dB+15dB=105dB.

D) as long as I don't try to play music louder than 105dB, there is no risk of speaker damage.

E) 105dB would get me quickly evicted from my NYC apartment, so I have nothing to worry about.

Is this logic sound?

Thanks in advance.

[J Larsen]

Thanks AE, I appreciate the reply.

Just out of curiosity, if we were talking about a 40W SS amp, would my logic be sound? I have a technical background (MS theoretical physics from ages ago) but not a strong background in electronics or the technical side of audio. Trying to learn, though!

[Æ]

Originally Posted by J Larsen

Thanks AE, I appreciate the reply.

Just out of curiosity, if we were talking about a 40W SS amp, would my logic be sound? I have a technical background (MS theoretical physics from ages ago) but not a strong background in electronics or the technical side of audio. Trying to learn, though!

I'm not exactly sure what it is you are asking me, but low powered solid state will clip easily unless the amp is known to have exceptionally good headroom.
Like I said earlier, if you have a good tube amp, don't worry about it clipping. I have a little 8 watt tube amp that gives out, runs out of power, but it never goes into a harsh clipping mode.

[daver]

As far as the 90 dB question . . . yes, it does depend on power input, and at what distance it is measured. Unfortunately, manufacturer's are not held to a "standard".

According to these b&w spec's it says 90 dB with 2.83V, at 1m.
http://www.bowers-wilkins.com/displa...rid=1112&sc=hf

The 2.83V would translate to 1W, if the impedance is 8 ohms at whatever frequency they measured the 90 dB at 1 meter. Note that the minimum impedance is 4.1 ohms, but no information about where that frequency is, or if the 90 dB was measured there or elsewhere.

[critofur]

Clipping is what's responsible for a lot of speaker damage. But, AFAIK, speakers will usually SOUND a little bad before they blow.

I imagine that most blown speakers occur when people either don't notice or don't care that they're sounding bad and just continue to play them loud, because they want it loud more than they want quality?

+1 on the above comment about tube amps clipping "gracefully".

Depending what music you listen to, the recordings, and how loud, I think that hardly any amps have the dynamic range + "reserve power" to fully reach the dynamic peaks in music, so maybe most music playback involves very brief moments of clipping / un-intentional compression by the playback system without reaching the point where the little "Clipping Indicator" LED on some amps would actually light up long enough to see it. The way that this sort of clipping sounds may one of the main reasons the sound of tube amps is appealing compared to solid state, though the "RMS" levels the music being played at may be significantly lower than the amps rated power...

[billfitzmaurice]

Originally Posted by J Larsen

B) I do not want to attempt to drive my speakers at a volume that would exceed the power capabilities of the amp, because that would lead to clipping and possible speaker damage.
Is this logic sound?

Yes and no. If you clip the amp, or anything else within the signal chain, the result is a higher than normal power density in the harmonics, which can result in damage to tweeters and less than robust midrange drivers. But unless you have particularly fragile tweeters chances are you'll hear distortion well below the point where damage will occur.

[J Larsen]

Originally Posted by billfitzmaurice

Yes and no. If you clip the amp, or anything else within the signal chain, the result is a higher than normal power density in the harmonics, which can result in damage to tweeters and less than robust midrange drivers. But unless you have particularly fragile tweeters chances are you'll hear distortion well below the point where damage will occur.

Thanks for that. Am I correct that clipping would be expected at volumes exceeding 105dB based on steps A and C in my logic? I'm just trying to understand how speaker sensitivity and amplifier power ratings interact.

[Æ]

Originally Posted by J Larsen

Thanks for that. Am I correct that clipping would be expected at volumes exceeding 105dB based on steps A and C in my logic? I'm just trying to understand how speaker sensitivity and amplifier power ratings interact.

Well, you can always calculate this or that, but the only way for sure to know when an amp is clipping or just beginning to clip would be to measure the amp while it's playing. Anyway your ears are more than capable of determining when things are beginning to fall apart.
The maximum clean SPL you can generate also depends on how far away from the speakers you are and the size of your room etc. Many variables. Get yourself a SPL meter, Radio Shack sells them.

[billfitzmaurice]

Originally Posted by J Larsen

Thanks for that. Am I correct that clipping would be expected at volumes exceeding 105dB based on steps A and C in my logic? I'm just trying to understand how speaker sensitivity and amplifier power ratings interact.

Add 10dB to the base sensitivity for 10 watts, 3dB for each doubling of power above that. 40 watts gets you to 106dB. 106 dB is mighty loud, so is 95dB for that matter.

[monte7998]

I was told years ago by a dude who owned a stereo shop:

90db...1w
93db...2w
96db...4w
99db...6w
102db...12w
105db...24w
108db...48w
111db...96w

I takes 3db for us to hear a diffrence in volume and a 3db increase takes double the wattage. How correct he was on this im not sure.

[billfitzmaurice]

Originally Posted by monte7998

I was told years ago by a dude who owned a stereo shop:

90db...1w
93db...2w
96db...4w
99db...6w
102db...12w
105db...24w
108db...48w
111db...96w

I takes 3db for us to hear a diffrence in volume and a 3db increase takes double the wattage. How correct he was on this im not sure.

4x2 is 8, not 6. Except in California, of course.

[Andy_G]

Originally Posted by billfitzmaurice

Yes and no. If you clip the amp, or anything else within the signal chain, .

This is one reason why you have to be careful with using digital sends and processors. In digital, a clip is purely a flat top curve.. not good. Analog (and tubes).. you have a little bit of leeway.. ie.. they clip 'gracefully"

[Æ]

Experienced audiophiles, those with "Golden Ears" can sometimes hear differences as little as 1dB.

Originally Posted by monte7998

I was told years ago by a dude who owned a stereo shop:

90db...1w
93db...2w
96db...4w
99db...6w
102db...12w
105db...24w
108db...48w
111db...96w

I takes 3db for us to hear a diffrence in volume and a 3db increase takes double the wattage. How correct he was on this im not sure.

[billfitzmaurice]

Originally Posted by Æ

Experienced audiophiles, those with "Golden Ears" can sometimes hear differences as little as 1dB.

Point of fact, when the decibel was invented Bell Labs (or whatever they called themselves back then) used the smallest deviation that could be heard by an average person as 1dB, but did so with a test tone, not broadband content. With a level source of broadband, like pink noise, it's nearly impossible to discern a 1dB change in level. If that content is dynamic, like music, it is impossible to discern a 1dB change. As to what oddiophiles can hear, that's anything their can convince themselves they do.

[curt_c]

Originally Posted by billfitzmaurice

With a level source of broadband, like pink noise, it's nearly impossible to discern a 1dB change in level. If that content is dynamic, like music, it is impossible to discern a 1dB change. As to what oddiophiles can hear, that's anything their can convince themselves they do.

As Einstein said: “It’s all relative”, and has much to do with psychoacoustics. It has been my experience in voicing speakers that the ear ‘locks’ onto a specific band of frequencies, perhaps a peak in response in the midrange, and uses that as the ‘reference’ level for the overall system sensitivity. This suggests that when comparing two speaker systems, how ‘sensitive’ they are compared to each other can vary depending on the music content and which frequency the brain determined as the reference. This is well documented, and used at great advantage in commercial speaker designs aimed at the masses.

When I voice speakers I like to set the levels for equivalent power responses (By ear, of course.) when I switch between the SUT and the reference speaker. (That would be the Speaker Under Test.) This was quite evident when I was voicing the Maverick’s, as the relative level between them and the reference speaker seemed to vary between 3 dB and 4 dB depending on the spectral content of the source material chosen. This difference was quite obvious, and repeatable in subsequent trials.

Because of this apparent ability of the brain to select a frequency band for a reference, very small changes in amplitude at frequencies where the ear is most sensitive are clearly audible, but much less so at the frequency extremes. I find using female vocalists particularly telling, and very small changes in response in that passband, even 0.5 dB, can be quite obvious.

As far as changing a single source levels by 1 dB, or even the 0.5 dB steps I have available on my system, I find it audible in the short term only.

C

[Æ]

That is why I said "sometimes."
Back in 2002 we had a DIY gathering up here in the Sacramento area at Dan "Lefty" Carroll's house. A couple of guys drove up all the way from Southern California (Los Angeles) to attend. I auditioned/critiqued this one guys pair of really nice 2 way bookshelf loudspeakers. I noticed and mentioned that the treble was a bit low, he pulled out a computer generated graph and explained to me that he had purposely designed it that way because of his unique listening environment. A lively room.
I asked him what was the difference on his graph between the tweeter and midbass, he told me about 1dB!
That was the best part of a decade ago, my hearing sensitivity isn't quite so good any more, my top end is rolling off quite a bit more now, but I still have highly polished "silver ears."
It's not so much your ears, but your brain that does the work, interpretation.

Originally Posted by billfitzmaurice

Point of fact, when the decibel was invented Bell Labs (or whatever they called themselves back then) used the smallest deviation that could be heard by an average person as 1dB, but did so with a test tone, not broadband content. With a level source of broadband, like pink noise, it's nearly impossible to discern a 1dB change in level. If that content is dynamic, like music, it is impossible to discern a 1dB change. As to what oddiophiles can hear, that's anything their can convince themselves they do.