|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
64.122.72.31
In Reply to: RE: How does speaker sensitivity affect real world SPL? posted by khachai44@gmail.com on May 07, 2017 at 21:26:00
Efficiency is 1 watt/1 meter.
Sensitivity is 2.83 volts at 1 meter.
If the speaker is 8 ohms, the two are the same.
(8 ohms = 2.83volts/0.35 amps; 1 watt= 2.83Volts x 0.35Amps)
If the speaker is 4 ohms, 2.83 volts works out to 2 watts, not one. IOW, there is the expectation that the amp is doubling its power!
So by cutting the impedance in half, the speaker can **appear** to be more efficient, which does not happen. If you want to know how efficient the speaker is, the only way to do that so it makes sense it to limit the input power to 1 watt.
Now to play with this a little, if you have a speaker that is 4 ohms and 90 db sensitivity, its efficiency is actually 87 db. This is because going from 2 watts to 1 watt is a 3 db change.
Conversely, if you have a speaker that is 90 db sensitivity and it is a 16 ohm speaker, its efficiency is 93 db.
This can have a profound effect depending on the type of amplifier you want to use. FWIW. One should also keep in mind that all amps regardless of technology have greater distortion into 4 ohms, and so will sound harsher and less detailed.
Follow Ups:
Ralph,I cant believe after all these years you're still mixing up sensitivity and efficiency, nor can i believe you haven't recieved your 16ohm medal of honor award. BTW , are you now saying distortion makes a difference, like .0024% is better than 1% ..
Yeah , I know.. :)
Edits: 05/08/17
One should also keep in mind that all amps regardless of technology have greater distortion into 4 ohms, and so will sound harsher and less detailed.
That may be true for direct-coupled amps (I can't think of any exceptions). But many transformer coupled amps have the same or higher distortion from their 8 ohm taps into an 8 ohm load than from their 4 ohm taps into a 4 ohm load, at equal power. And it's not just some tube amps, the McIntosh solid state amps exhibit this behavior too.
Just by using the 4 ohm tap instead of the 8 ohm tap (and assuming that in each case the taps are loaded correctly) with some output transformers. I forgot to mention that...
What would be the mechanism for that?
Certainly connecting a 8 ohms load to the 4 ohms tap will cause the LF -3db point to move one octave up because the reflected impedance has doubled but the inductance stayed the same but the reflected impedance with a 4ohm load connected to the 4 ohm tap should be the same as a 8 ohm load connected to the 8 ohms tap and therefore the LF -3db point should stay the same.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
therefore the LF -3db point should stay the same.
Emphasis added; 'should' being the modifier in that sentence, and I agree.
However, in reality the fact is that the larger the turns ratio it seems that there is a loss of efficiency. We've had several custom transformer projects and run into this phenomena. I suspect a YMMV thing depending on the quality of the part.
Boulder
You are right that sensitivity and efficiency are different but your definition of efficiency is wrong. Efficiency is how much sound energy is produced by the electrical energy provided to the speaker. Most of that energy is actually converted to heat (that's why voice coils melt) rather than sound. A speaker with a sensitivity of 90db is only 0.62% efficient - the rest of the energy goes up the chimney. A detailed discussion and the effects of efficiency can be found at on Ron Elliott's webpage.
I married the perfect woman. The downside is everything that goes wrong is my fault.
-
Nope.Sensitivity is, and always has been, expressed in 1W/1m.
"Efficiency" is often used to describe sensitivity while taking into account impedance. (Sensitivity compensates for impedance). This is not really what actual efficiency is... efficiency is really an expression of electrical power in compared to acoustic power out.
A tweeter that is 90db 1W/1m sensitive is 0.6% efficient
A prosound woofer that is 97db 1W/1m sensitive is 3% efficient.The rest goes into heat (I^2*R coil losses and overcoming frictional forces of the mechanical system, which also results in heat).
I always refer to this calculator page for efficiency.
http://www.sengpielaudio.com/calculator-efficiency.htm
Although 32 and 16 ohm speakers were once very common, 8 ohms became a common "nominal" impedance for speakers for many decades. As such, the voltage used for efficiency is 2.83V, as this voltage when applied to an 8 ohm nominal load gives 1 Watt.
P = (V^2) / R
P = 2.83^2 / 8
P = 8 / 8
P = 1WInterestingly, this means an 8 ohm speakers will have the same value for efficiency and sensitivity!
When you apply 2.83 volts to a 16 ohm speaker you get
P = (V^2) / R
P = 2.83^2 / 16
P = 8 / 16
P = 0.5WWhen you apply 2.83 volts to a 4 ohm speaker you get
P = (V^2) / R
P = 2.83^2 / 4
P = 8 / 4
P = 2WEfficiency takes into account the speakers impedance in so far as how it impacts power. Sensitivity compensates for this by adjusting the input voltage to keep input power constant.
I could be wrong, but I think this is a common usage of the word "efficiency" when it's technically incorrect. In reality, it's really sensitivity with and without using the 8 ohm reference for power. Sensitivity 1W/1m and sensitivity 2.83V/1m.
1W/1m = let's adjust voltage to have a common 1W input.
2.83V/1m = let's apply 2.83V and allow power to vary with impedance.Since amplifiers are voltage sources, the 2.83V/1m has real-world applications, where sensitivity really speaks more to the speakers actual *efficiency*.
As such, a 2 ohm speaker can be less efficient than an 8 ohm speaker but still produce higher SPL at 2.83V. This is because you get 4W at 2 ohms and 1W at 8 ohms with 2.83V applied.
Cheers,
Presto
Edits: 05/09/17 05/09/17 05/09/17
Your opening sentence ("Sensitivity is, and always has been, expressed in 1W/1m.") is incorrect.
Sensitivity is in fact 2.83 Volts at one meter.
In every other way your post is correct.
Efficiency as a measurement standard went out some years ago (1980s), to be replaced by sensitivity. But you still see it used all over the industry. There is continuing confusion on the topic.
The history of it goes back to the 1950s, when EV and MacIntosh began to push for a voltage source amplifier standard. Prior to that time, all speakers had to have level controls for the midrange and tweeter as the actual voltage response of the amplifier to be used was unknown (IOW they were not there to adjust to the room). By adopting a voltage source as a reference, the controls could be eliminated. But the older efficiency standard held on well into the 1980s and is still used today.
"Sensitivity is, and always has been, expressed in 1W/1m."Okay, that might well be incorrect. Here, let's try this:
"Sensitivity, in my travels through domestic speaker specifications, has almost always been expressed in 1W/1m"
They will also say "1m SPL@2.83V". It's rare to see them cite 4V for 16 ohm drivers or 2V for 4 ohm drivers even though if they give 1W/1m as those are exactly the applied voltages. I've seen a lot of specs in pro-sound circles for four ohm driver which are at 2.83V / 1m. 2.83V vs. 1W can't be the same if you say "The 8 ohm version is 3db less sensitive than the 4 ohm version". If you use voltage as the reference, there is a 3db difference. If you use power (1W) as a reference they will have the same sensitivity, if all other design details besides coil impedance are observed.
This is why the *word* sensitivity (at 1m) can't mean 2.83V *and* 1W at the same time unless it's an 8ohm nominal driver or system.
This is why I thought they reserved "sensitivity" for 1W/1m and perhaps threw people off calling 2.83V/1m/8 ohms sensitivity as well.
I wonder if life would be easier if we said "1m power sensitivity in 1W/1m" and "1m voltage sensitivity in 2.83V/1m"
EV has SPL at 10 feet with, what was it, four volts applied?
From the EVM12S 12" mid-woofer specification sheet:(EIA Standard SE-103 Section SE3).
The thing is, it's easy to get 1W/1m from the 10 foot measurement - just increase the SPL per the inverse square law for the SPL at 1m instead of 10 feet. (It's already at 1W).
If the person seeing the spec knows what it is and can compare apples to apples with other drivers/systems he's okay. For everyone else, it's not surprising there is confusing surrounding the terms and their application.
Cheers,
Presto
Edits: 05/10/17 05/10/17
We're pretty much on the same page except that first bit.
Okay, that might well be incorrect. Here, let's try this:
"Sensitivity, in my travels through domestic speaker specifications, has almost always been expressed in 1W/1m"
That's a bit closer, which is to say that individual manufacturers are making the same mistakes that many consumers do, which is that they freely mix the two terms.
Whenever you see the '1W/1m' thing expressed as 'sensitivity' that is the manufacturer screwing up. These days, many of them are so young they don't even know there was a prior measurement standard.
So... what are the *correct* terms?
We know what SPL 2.83V/1m and SPL 1W/1m means technically...
What are the terms for EACH.
They are not the same by definition although they will yield the same result for an 8 ohm nominal device.
The best descriptions I have come across is:
2.83V/1m -> Voltage Sensitivity
1W/1m -> just "Sensitivity" but perhaps should be "power sensitivity"
Cheers,
Presto
-
Ralph ;
" These days, many of them are so young they don't even know there was a prior measurement standard."
Careful Sir...your wrinkles are showing !!!
LOL .
Couldn't resist !
-
...for the link to Rod Elliott's site. My link to his bucking transformer article hasn't worked for a long time, and I thought the site was gone forever.
Thanks again.
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: