Let's not.
Let's instead consider your average listening level, the efficiency of your speakers and the dynamic range of your source material.
If you wish to be able to reproduce 20db peaks above the RMS SPL level of the source, you will need 100 times the wattage that you need to reproduce the RMS SPL level.
Now, take your speaker efficiency. Suppose it is 80 dB/1 watt@1meter. If your average level is 80 dB, that 20 dB peak will require 1x100 = 100 watts to reproduce.
Suppose your speakers are 90 dB/1watt@1meter. Again, if your average level is 80 dB, that 20 dB peak will require 0.1 x 100 = 10 watts to reproduce.
If your speakers are 100 dB efficient, with an 80 dB average level, you only need a 1 watt amp to reproduce a 20 dB peak.
--------
The original poster is keen on the threshold of pain -- 120 dB. Start with speakers that are 100 dB efficient or more. Then 100 watts will do.
90 dB needs 1000 watts. BUT, it is unlikely that these speakers will play that loud. Forget it with 80 dB efficiency.
Also, what is absent from the analysis is the listening distance. In my room, I listen at 2m. I have two speakers, which makes up for the loss of SPL due to distance. I also have a small room, so it is easily loaded -- SPL is higher than it would be in a barn.
As was stated earlier, absorption of walls/floors/ceiling also matters.
However, the main factor that you need to address for success remains speaker efficiency, not amplifier power.
No - it's 5 (pairs , that is). Or one pair for yer' cheap speakers. 😀
Most PE (diy) drivers are 88-91db ... <90db is average.
Build a bookshelf - 1 pair big BJT - 80-100W.
Tower w/multiple woofers (paradigm style) .. a 200W will be fine.
(good)Subs like 250+
All for 89db drivers to hit 110-115db in a room.
Why I give 100/200/300W (2/3/5 pairs) as options on my new amp.
The clipping is even more important , I went through more tweeters
on my 95W amp than on my 200+ w ones.
I also took some junk tweeters to test different input stages. a Doug Self
amp will smoke a tweeter on hard clip almost immediately.
New fancy self-clamping amps , you have to clip hard and long - to do the same.
An over-powered amp (that does not clip) , the voice coil will just heat up
and burn gradually. Best is the new "fancy amp" that is also overpowered.
Then if you DO clip occasionally , it will be with less voice coil frying harmonics.
OS
Most PE (diy) drivers are 88-91db ... <90db is average.
Build a bookshelf - 1 pair big BJT - 80-100W.
Tower w/multiple woofers (paradigm style) .. a 200W will be fine.
(good)Subs like 250+
All for 89db drivers to hit 110-115db in a room.
Why I give 100/200/300W (2/3/5 pairs) as options on my new amp.
The clipping is even more important , I went through more tweeters
on my 95W amp than on my 200+ w ones.
I also took some junk tweeters to test different input stages. a Doug Self
amp will smoke a tweeter on hard clip almost immediately.
New fancy self-clamping amps , you have to clip hard and long - to do the same.
An over-powered amp (that does not clip) , the voice coil will just heat up
and burn gradually. Best is the new "fancy amp" that is also overpowered.
Then if you DO clip occasionally , it will be with less voice coil frying harmonics.
OS
I did something interesting (in my view at least 😛). I got the sample data from some songs and put them in a spread sheet. I then took an average of the samples and compared them to the maximum sample value to find the dynamic range of these songs and I was actually surprised at how big the range actually is. For the noisiest part I could find of a cannibal corpse track the peak value was 3.4 times greater than the average value. For 'kalimba' (windows sample music) the peak was actually 5.3 times greater than the average. I did some other tracks as well as tested different parts of the same tracks but the max/average ratio came out pretty much the same.
If the peaks of your average music is at least four times greater than the average level it would mean that your amplifier should be able to output 16 times the wattage during peaks, probably way more if you're only listening to classical music. This difference in peak vs. average output power seems really big, I'm tempted to try and make an amplifier with a tiny transformer, but that has a rather high secondary voltage to see what an 5W average, but 80W peak amplifier sounds like 🙂
If the peaks of your average music is at least four times greater than the average level it would mean that your amplifier should be able to output 16 times the wattage during peaks, probably way more if you're only listening to classical music. This difference in peak vs. average output power seems really big, I'm tempted to try and make an amplifier with a tiny transformer, but that has a rather high secondary voltage to see what an 5W average, but 80W peak amplifier sounds like 🙂
Enough to play 105dB at the listening position (for classical music), add 10dB for louder music types.
How efficient are the speakers?
85dB=100W
95dB=10W
105dB=1W (Klipschorn)
How efficient are the speakers?
85dB=100W
95dB=10W
105dB=1W (Klipschorn)
According to the Musical Fidelity calculator (a card nomograph style give away) one would need 10kW +10kW into a pair of 87dB/W @ 1m speakers to get 120dB at a 3m listening position assuming no power compression of the speakers.
Add on an allowance for 3dB of power compression and increase the speaker sensitivity to 100dB/W @ 1m, one would still need 1kW+1kW to reach 120dB @ 3m.
Reduce the requirement to 100dB @ 3m and one needs 5W+5W (after removing the 3dB power compression allowance).
Add on an allowance for 3dB of power compression and increase the speaker sensitivity to 100dB/W @ 1m, one would still need 1kW+1kW to reach 120dB @ 3m.
Reduce the requirement to 100dB @ 3m and one needs 5W+5W (after removing the 3dB power compression allowance).
Pano already did, though it's expressed in voltage rather than power. With most amps, speakers, and listening combinations that doesn't matter as current limiting won't occur.
For power dB is 10 log10(), not 20 log 10(). You may also find Emmanuel Deruty's dynamic range investigations of interest.
You might take a closer look at how loop gain varies across different amplifier topologies. Lot of variability with parts selection and how local and global feedback is used but, generally and loosely speaking, there is something of a power versus loop gain tradeoff. It's particularly keen for high loop gain implementations where the tradeoff is over integrated output device selection in composite class AB and high order class D loops.
Yep. OP also seems keen on overbuilding, a common trait among DIYers. There is some variance from specification to specification but most agree prolonged exposure over 80dB or so results in permanent hearing loss. 120dB is typically considered the onset of instantaneous permanent damage. Perhaps OP wishes only to listen once? 😛
I've seen a pair of Altec 604s (time aligned) support a 5000 seat auditorium well with a pair of 30 watt dukane tube amps and it was compfortably loud. Also most movie theaters in the height of the Cinerama/stereo sound era with RCA/Altec plane wave horn (ubangi and other) used amps seldom above 25 watts and most of those theaters seated at least 1500 people. But most of our living-listening spaces are the far cry from those theaters/auditoriums. As a business I think most sound companies try to sell as much equipment as they possible can irrespective of what sounds good or will do the job. It is interesting to study the old (1948) Western Electric designed speakers like the 755 full range. This 8 inch speaker is highly regarded as a fine performer even now. Few speakers nowdays are designed and made as well.
I have 3 100 w DIY Leach amps in a three way electronically cross-over system. I have a seperate stereo sub channel below 125hz with a pair of Leach DIY super amp 300w (double barralled). The Leach amps all drive woofers. The mids and tweeters are driven my Lux MB3045 tube amps designed by Tim deParavacini. Just my way of doing things. Ray
I have 3 100 w DIY Leach amps in a three way electronically cross-over system. I have a seperate stereo sub channel below 125hz with a pair of Leach DIY super amp 300w (double barralled). The Leach amps all drive woofers. The mids and tweeters are driven my Lux MB3045 tube amps designed by Tim deParavacini. Just my way of doing things. Ray
"100 w DIY Leach amps"
On ±52V, the home built Leach is 50W into the 16Ω impedance of the 110dB/W RCA, plenty of power:
(now in a small community theater in Topeka, this set of RCA were originally in the 2,000 seat Paramount Theater in Cedar Rapids Iowa)
On ±52V, the home built Leach is 50W into the 16Ω impedance of the 110dB/W RCA, plenty of power:
(now in a small community theater in Topeka, this set of RCA were originally in the 2,000 seat Paramount Theater in Cedar Rapids Iowa)
Last edited:
What if we are only talking sub woofers? There's no risk of running into any pain thresholds below 80hz, right?
When you hit around 130+ dB it can get uncomfortable, and over 140dB permanently damage your hearing in an instant.
"As regards OSHA and absolute levels, music is not the same as noise. A noise level that would have you cringing in a factory, and tensing your ears, does not have the same (hearing damage) effect on music. There are studies in the literature that show this, but the dB police don't want to hear this, and so they ignore it.
Dance music can take a lot of flack. The SPL levels are mostly bass though. OSHA 'A' scale readings roll-off the bass at a 6dB/oct rate below 1Khz. So a 105dBA exposure for one hour could hit 130dB at 40hz and be OK. Too damn loud, you bet. Damaging levels, not according to OSHA. It's usually the excessive treble that gets crowd complaints (the bass gets the neighbors complaining).
Car stereo with its 150+ dB peak levels is another issue. I will not get in a car that can play louder than about 130dB, even if I am wearing full coverage ear protection, bone conduction will get you. "
"As regards OSHA and absolute levels, music is not the same as noise. A noise level that would have you cringing in a factory, and tensing your ears, does not have the same (hearing damage) effect on music. There are studies in the literature that show this, but the dB police don't want to hear this, and so they ignore it.
Dance music can take a lot of flack. The SPL levels are mostly bass though. OSHA 'A' scale readings roll-off the bass at a 6dB/oct rate below 1Khz. So a 105dBA exposure for one hour could hit 130dB at 40hz and be OK. Too damn loud, you bet. Damaging levels, not according to OSHA. It's usually the excessive treble that gets crowd complaints (the bass gets the neighbors complaining).
Car stereo with its 150+ dB peak levels is another issue. I will not get in a car that can play louder than about 130dB, even if I am wearing full coverage ear protection, bone conduction will get you. "
Your right its not the right place for this thread or any thread like it. It asks too many questions and provides too little information. No information on the efficiency of the speakers in question and no idea of the square footage of the room in question. Just another mindless thread in my opinion.
I don't purchase any amplifier that is less than 200 watts a channel if its a SS amplifier. If its a tube amp I am comfortable with my Allen model 75's or my modified Quicksilvers.
"100 w DIY Leach amps"
On ±52V, the home built Leach is 50W into the 16Ω impedance of the 110dB/W RCA, plenty of power:
(now in a small community theater in Topeka, this set of RCA were originally in the 2,000 seat Paramount Theater in Cedar Rapids Iowa)
Thats efficiency high SPL with minimum power.
djk: Are those RCA electro dynamic woofers in that Ubangi cabinet or permanent magnet? Thanks Ray
The permanent magnet types.
The networks were re-worked for nearfield listening in a 30' x 50' garage, they seem to be happy in their new home with the same setting.
Supertweeter crossed at 4Khz/18dB, flat to 20Khz.
The networks were re-worked for nearfield listening in a 30' x 50' garage, they seem to be happy in their new home with the same setting.
Supertweeter crossed at 4Khz/18dB, flat to 20Khz.
I'm pretty much willing to bet that you can't get a speaker to mimic even a slow chemical explosion. Twice.
A gunshot too close can perforate an eardrum. You might be able to do that with a close coupled IEM, particularly if you were happy to vaporise the coil.
A gunshot too close can perforate an eardrum. You might be able to do that with a close coupled IEM, particularly if you were happy to vaporise the coil.
- Status
- Not open for further replies.