You will notice that I measured with an amp that was far, far form clipping and still found only 6V peaks at the extremes. I doubt the amp was running out of steam and compressing the peaks. If we are using a digital source, we have the advantage of knowing its maximum output level. It's maximum is fixed, it can't get any louder.
Yes - Power supplies can make a big difference to the sound of an amp. But an amp that is limited in voltage is not necessarily limited in current. Some are, some aren't. I tend to like those that are not.
Even the most dynamic recordings I own don't have a "loud section" to peak ratio over 22dB. Usually much less.
So if I'm playing loud at an average 85dB the peaks will be 107dB. That's loud for home listening. You want that 22dB of headroom? OK, then figure out what your average voltage level is and multiply by 12.6. That's the peak voltage you need. Most of us won't ever need that much headroom, but there it is. I'll leave you alone with Ohm's law and a calculator to figure out how much power that is - and what amp you need to achieve it.
If you're running 1 ohm speakers, you may not need a lot of voltage, but you will need a lot of current, which means power. Low output impedance will help, too. But this is a thread about high efficiency full range speakers, most of which are nominally 8 ohms. You can easily figure out how much power you are using, the steps to do so have been given in this thread. The next step is to figure out if your amp can actually supply it. That won't change how much you are using, but may change the amp you choose or the power supply you build.
Believe me, if your amp can really supply 22dB over your average loud level, you have plenty of headroom. I can't tell you if your amp can do it, but I can tell you how much you need.
Yes - Power supplies can make a big difference to the sound of an amp. But an amp that is limited in voltage is not necessarily limited in current. Some are, some aren't. I tend to like those that are not.
Even the most dynamic recordings I own don't have a "loud section" to peak ratio over 22dB. Usually much less.
So if I'm playing loud at an average 85dB the peaks will be 107dB. That's loud for home listening. You want that 22dB of headroom? OK, then figure out what your average voltage level is and multiply by 12.6. That's the peak voltage you need. Most of us won't ever need that much headroom, but there it is. I'll leave you alone with Ohm's law and a calculator to figure out how much power that is - and what amp you need to achieve it.
If you're running 1 ohm speakers, you may not need a lot of voltage, but you will need a lot of current, which means power. Low output impedance will help, too. But this is a thread about high efficiency full range speakers, most of which are nominally 8 ohms. You can easily figure out how much power you are using, the steps to do so have been given in this thread. The next step is to figure out if your amp can actually supply it. That won't change how much you are using, but may change the amp you choose or the power supply you build.
Believe me, if your amp can really supply 22dB over your average loud level, you have plenty of headroom. I can't tell you if your amp can do it, but I can tell you how much you need.
Could you please elaborate on how this could be done? I can't see this happening, with all the resistance inherant in the output transformers (to name one of the many limiting factors).
Cheers
Chris
My single ended mosfet certainly can't - it only has 0.5A constant current in the output stage.
By that reckoning, my class A amp should manage the same trick as it is single ended and can sink a lot more current than it can source. I'll have to give it a go. Thanks.
OK, quick check, you'd need about a 500V B+ power supply for that.
But still, on short timescales the typical class A tube amp has much higher power than it can produce on average. I think because on very short time scales we can assume the power supply is only a single 47uF eltro. cap with maybe 1R output impedance. On longer time scales we need to look at the entire power supply.
The question was "How much power do you need? Someone said his 2W amp was enough. My counter argument was that his 2W amp can product more than 2W during a transient. And now I'll add that if you were to add some kind of diode clipping at the speaker terminals that truly did limit power to 2W peak the amp would sound horrible. It's those peaks that go way above 2W that give the sound some life.
How much higher? My guess might be off by a factor of two. But who knows. ti depends on details we don't know.
The real answer when some one asks "how much power?" is to ask back "how sensitive are your speakers?"
I concur with Pano. Why guess? Just get a scope and measure. If you don't have one, borrow one, or buy a cheap used one from the 70's.
My speakers are 98db. I recently built a PP tube amp. At first I had el34s in it. It was dynamic as all hell, but I felt like I was being bashed in the head and in the chest by it. Too much power! I finally made some adjustments and swapped in 6v6's. Much better! When I put it on the scope I found that when I cranked it up to house party loud, I only could find transients up to an 8v peak. Most of the music was in the 1-2v range. That's it, and that's all I really need with these particular speakers. If your amp has too much power, you might be losing some detail in the quiet parts of the music, and you are forced to turn it up a bit. Then when a louder passage comes up, you might have to scramble to turn it down.
For this thread to have any meaning, lets post some measurements. Here's mine:
Speakers: 98db/ 8 ohm Klipsch RF3 II
Absolute max transients at very high volume: 8V
Average voltage range at very high volume: 1-2V
Absolute max transients at moderate volume: 4-5V
Average voltage range at moderate volume: .25 -1V
Yeah, the room matters, and all that stuff, but some measured numbers will give everyone a good ballpark idea.
Last edited:
Couldn't agree more. By far the simplest and most graphic method. If the peaks aren't clipped and distortion is acceptable you have enough power.
It would be of interest to know what the spl was for the various levels (i.e., very high volume and moderate volume). Measurements at the listening position and at 1m from the speakers would be most helpful.
Regards,
Bob
I don't believe the figures you state.
It looks like you don't know how to measure average from a waveform.
Just use a voltmeter for averages and use the scope for peaks.
Using just the peak voltage information, since the averages are being disputed.
8Vpk into 98dB/W @ 1m 8ohm is 107dBpk. Yes loud.
4Vpk is 101dBpk. I would still class this as loud.
Last edited:
I shouldn't have used the word "average" because I know am inadvertently touching on the terms "average power" and "peak power". I know I need to divide by sq rt 2 and all that. But all I mean is that when I look at the scope at very high volume levels, most of the music is happening around in the 1-2V range. Then when a bass drum, or other loud transient kicks in, in very extreme cases, it may hit 8V. Yes, this is at very very high volume levels, mostly suitable for house parties where you are 10 ft away, and not for regular listening.
In the moderate case, and I yes I suppose still reasonably loud, most of the music is happening in the .25-1V. When a very loud transient such as bass drum, or low frequency sound of some sort kicks in, it may kick up to 4V just for that brief moment. But this is the upper extreme.
I know these observations aren't perfect, but I do believe it gives a pretty good idea of the power needed. Basically, in the moderate case (still plenty loud), I would be OK with 1W or so. In the house party loud case we're talking 4 watts RMS or so. [(8V/1.414)^2/ 8ohms = 4W rms]
Please correct my calculation if it doesnt seem correct. But hopefully it gives a ballpark idea. Other's observations may be different, so it would be interesting to see a sampling of other reports with different sensitivity speakers.
Unfortunately, I don't have an SPL meter, but maybe I can get one.
Ancan, thanks very much for doing the measurements and posting them. It's nice that you were able to look at them with the scope.
Yes, 8 volts into 8 ohms is 8 watts. 8-8-8. An amp that can do a clean 4W RMS should be able to provide 8 watts peak. That's a lot of power for 98dB/1W/1M speakers, as Andrew points out. It must have been loud!
Anyone else care to measure? You don't have to have an oscilloscope, just a volt meter with a 20V AC scale. It would be nice to see measurements from someone with inefficient speakers and a big amp.
Yes, 8 volts into 8 ohms is 8 watts. 8-8-8. An amp that can do a clean 4W RMS should be able to provide 8 watts peak. That's a lot of power for 98dB/1W/1M speakers, as Andrew points out. It must have been loud!
Anyone else care to measure? You don't have to have an oscilloscope, just a volt meter with a 20V AC scale. It would be nice to see measurements from someone with inefficient speakers and a big amp.
Just to repeat and seeing Pano agreeing.
8Vpk into .. etc, when driving two speakers results in ~102dBpk at a 2.5m listening distance. That is still loud.
With a range of music types expect a few peaks to be slightly higher, maybe another +3dB or +6dB .
That is even louder, but fortunately of very short duration.
An F5 on +-25Vdc supplies can easily output transients of 20Vpk.
8Vpk into .. etc, when driving two speakers results in ~102dBpk at a 2.5m listening distance. That is still loud.
With a range of music types expect a few peaks to be slightly higher, maybe another +3dB or +6dB .
That is even louder, but fortunately of very short duration.
An F5 on +-25Vdc supplies can easily output transients of 20Vpk.
Last edited:
It may be even more Andrew, depending on the room. Even at that he may have been running at an average of 85-89dB. Quite loud in a small room. It's when you get into huge rooms that the power requirements really go way up. Thus the need for massive amps and efficient speakers in P.A.
😱 Relax ,
I'm working on it , pulling my LMS/Leap unit out, putting the speakers back together, It can measure volts, avg and peak along with db, can't show clipping like a scope , i do know on large scale music, 6-8 amps avg, plus 18 db ..😀
I have inefficient subwoofers and fairly efficient Fostexes (see my sig for details). I'll do some measurements some time soon, and let you guys know what I get.
Chris
Thanks Chris, I was hoping to see some measurements for mid to low efficiency stuff. Like 85dB/2.83V or lower. Maybe some of the 3-4" fullrange speakers. Tho I doubt they will be played very loud, it would still be nice to see how much power they use when playing as loud as they do in normal conditions.
That might give us a rough figure for "Low Efficiency Speakers, How much power do they really need?" I was thinking fullrange, because this is the Fullrange forum, but others are welcome.
That might give us a rough figure for "Low Efficiency Speakers, How much power do they really need?" I was thinking fullrange, because this is the Fullrange forum, but others are welcome.
- Status
- Not open for further replies.