Sensitivity is not a specification based on Re, you simply
don't want to the see the wood because of the trees.
Re simply doesn't exist as a parameter for complete loudspeakers.
Sensistivity for loudspeakers is measured in dB/2.83V.
Z nominal for loudspeakers is an attempt to define equivalent
resistive loading of the louspeaker in order to ascertain the
load driving capability required from the amplifier.
However for a lot of loudspeakers dB/2.83V is not the same
as dB/W, given we are talking how many dB's for one of our amplifiers watts.
In this case the only relevant parameter is the nominal impedance
of the speaker which was used to choose the amplifier initially.
The simple fact is it is completely accepted that :
An 8 ohm nominal loudspeaker dB/2.83V = dB/W
A 4 ohm nominal loudspeaker dB/2.83V = dB/2W
A 2 ohm nominal loudspeaker dB/2.83V = dB/4W
And the same rational is nearly always applied to drivers.
Though vendors of 4 ohm drivers and 4 ohm dual coil wired in
parallel often print the sensistivity as X dB, which misleadingly
is dB/2.83V and the unwary often assume is dB/W.
The honest vendors would use the equivalent db/W figure,
i.e. for 4 ohms subtract 3dB from the dB/2.83V figure and
for 2 ohms subtract 6dB from the dB/2.83V figure.
Your average Joe doesn't care about the definition of efficiency,
they want to know how loud it is per watt without being conned.
Your definitions simply don't help, and are unecessarily confusing.
Whats more your assumptions from the plain
facts of your reference are simply wrong.
The voltage sensistivity of a driver can be easily calculated.
Re is not used to modify this as you suggest for loudspeakers
or drivers to give the "SENSITIVITY" (sic).
sreten.
don't want to the see the wood because of the trees.
Re simply doesn't exist as a parameter for complete loudspeakers.
Sensistivity for loudspeakers is measured in dB/2.83V.
Z nominal for loudspeakers is an attempt to define equivalent
resistive loading of the louspeaker in order to ascertain the
load driving capability required from the amplifier.
However for a lot of loudspeakers dB/2.83V is not the same
as dB/W, given we are talking how many dB's for one of our amplifiers watts.
In this case the only relevant parameter is the nominal impedance
of the speaker which was used to choose the amplifier initially.
The simple fact is it is completely accepted that :
An 8 ohm nominal loudspeaker dB/2.83V = dB/W
A 4 ohm nominal loudspeaker dB/2.83V = dB/2W
A 2 ohm nominal loudspeaker dB/2.83V = dB/4W
And the same rational is nearly always applied to drivers.
Though vendors of 4 ohm drivers and 4 ohm dual coil wired in
parallel often print the sensistivity as X dB, which misleadingly
is dB/2.83V and the unwary often assume is dB/W.
The honest vendors would use the equivalent db/W figure,
i.e. for 4 ohms subtract 3dB from the dB/2.83V figure and
for 2 ohms subtract 6dB from the dB/2.83V figure.
Your average Joe doesn't care about the definition of efficiency,
they want to know how loud it is per watt without being conned.
Your definitions simply don't help, and are unecessarily confusing.
Whats more your assumptions from the plain
facts of your reference are simply wrong.
The voltage sensistivity of a driver can be easily calculated.
Re is not used to modify this as you suggest for loudspeakers
or drivers to give the "SENSITIVITY" (sic).
sreten.
So how does all this crap you just wrote down answer the original question? It doesn't.
There is a fundamental failure to communicate here and it has nothing to do with being wrong. It is a different way of looking at the problem which you do not seem to have properly grasped in the first place.
If I am wrong, Prove my assumptions wrong, not with banal spouting of unconnected facts, but with an actual argument connected to something I have said.
If Re doesn't exist as a parameter for loudspeakers, I suppose Sd, Bl and Mms are fantasies as well?
Z nominal has nothing to do with these ratings. What you call completely accepted has no basis in fact.
I know the voltage sensitivity can be calculated, and I gave you a formula....
What color is the sky in your world?
There is a fundamental failure to communicate here and it has nothing to do with being wrong. It is a different way of looking at the problem which you do not seem to have properly grasped in the first place.
If I am wrong, Prove my assumptions wrong, not with banal spouting of unconnected facts, but with an actual argument connected to something I have said.
If Re doesn't exist as a parameter for loudspeakers, I suppose Sd, Bl and Mms are fantasies as well?
Z nominal has nothing to do with these ratings. What you call completely accepted has no basis in fact.
I know the voltage sensitivity can be calculated, and I gave you a formula....
What color is the sky in your world?
Ron E said:
I think the original poster realised the answer to his question in post #4. It seems we haven't got it yet. (I just realised that he asked for a conversion from dB@2.83 *watts* to dB@1 watt.) I think he understood back then what we haven't stopped arguing about yet.
I think it is all about "what is commonly done" and of course we can have different opinions on this. I think we all understand that 2.83 V into 8 ohms is 1 watt. That's all there is to it.
End of story (for me
)I second Svante.
Speaking as a member who has learned much from both Ron and Sreten, I think we are putting too much into a battle of Beranek vs modern usage.
I too thought that efficiency was always presented as a percentage. Here is an excerpt from a thread where Dan Wiggins of the highly respected Adire Audio gives his thoughts about sensitivity versus efficiency, and how it is written. He continues on for a couple of posts.
http://www.diyaudio.com/forums/showthread.php?postid=334361#post334361
Ron E said:
My sky is pragmatic, yours is your own, in a minority of one.
If you can't accept the way things are commonly percieved,
its your problem not mine. I'm just telling you the way it is.
I'm not going to be polite about this because....................
god I can't be bothered.
(other than I refer to a loudspeaker as a complex system,
and a driver as a single entity.)
sreten.I'm going to have one last go at this and then give up :
For a single driver :
Voltage sensitivity is easily calculated - all seem to agree.
Ron E equation gives the efficiency when the drivers impedance
is resistive and equal to Re, for a single driver this is typically its
point of lowest efficiency above its high pass cutoff and below
its low pass cutoff.
This can easily be defined as a point of reference for calculations.
Drivers are described as having a nominal impedance which is
typically 20% to 30% higher than Re. This is because the average
loading of a driver must be higher than its minimum loading.
Ron E contention is that a driver with an Re of 8R has dB/2.83V
than is equal to dB/W, this is certainly true at the drivers point
of lowest efficiency given the constraints defined above.
My contention is an 8 ohm nominal driver has dB/2.83V = dB/W.
Ron E' s equation shows that a typical 8 ohm driver with Re = 6R
is actually 1.26dB less efficient than this at its minimum efficiency,
(given constraints defined above) which is certainly true.
Ron E thinks the minimum efficiency defines the reference
efficiency, I think its just a reference point for calculations.
For example : a loudspeaker review for an 8 ohm nominal speaker
will happily convert dB/2.83V to dB/W, whilst advising that for
4 ohm speakers real dB/W is actually 3dB less than the dB/2.83V.
For complete loudspeaker systems Re is not an available
parameter, and due to the fecklessness of some crossover
designers the point of minimum efficiency may be entirely
unrelated to the Re of the bass (or any other) driver(s).
Again the point of Z nominal for a complete loudspeaker system
is that it is derived from the measured impedance curve, not
the parameters of the drivers.
This is all commonsense. I don't see what the problem is other
than Ron E refuses to accept the concept of average impedance,
average efficiency and the I accept somewhat inexact way the
nominal impedance is used to estimate these.
sreten.
For a single driver :
Voltage sensitivity is easily calculated - all seem to agree.
Ron E equation gives the efficiency when the drivers impedance
is resistive and equal to Re, for a single driver this is typically its
point of lowest efficiency above its high pass cutoff and below
its low pass cutoff.
This can easily be defined as a point of reference for calculations.
Drivers are described as having a nominal impedance which is
typically 20% to 30% higher than Re. This is because the average
loading of a driver must be higher than its minimum loading.
Ron E contention is that a driver with an Re of 8R has dB/2.83V
than is equal to dB/W, this is certainly true at the drivers point
of lowest efficiency given the constraints defined above.
My contention is an 8 ohm nominal driver has dB/2.83V = dB/W.
Ron E' s equation shows that a typical 8 ohm driver with Re = 6R
is actually 1.26dB less efficient than this at its minimum efficiency,
(given constraints defined above) which is certainly true.
Ron E thinks the minimum efficiency defines the reference
efficiency, I think its just a reference point for calculations.
For example : a loudspeaker review for an 8 ohm nominal speaker
will happily convert dB/2.83V to dB/W, whilst advising that for
4 ohm speakers real dB/W is actually 3dB less than the dB/2.83V.
For complete loudspeaker systems Re is not an available
parameter, and due to the fecklessness of some crossover
designers the point of minimum efficiency may be entirely
unrelated to the Re of the bass (or any other) driver(s).
Again the point of Z nominal for a complete loudspeaker system
is that it is derived from the measured impedance curve, not
the parameters of the drivers.
This is all commonsense. I don't see what the problem is other
than Ron E refuses to accept the concept of average impedance,
average efficiency and the I accept somewhat inexact way the
nominal impedance is used to estimate these.
sreten.
My issue with using nominal impedance is that it is an arbitrary number that is decided by whoever writes the specifications of the driver. While one may choose to use nominal impedance for rating efficiency or sensitivity, the use of this number creates problems when one desires to compare the sensitivities of drivers with different manufacturers - or even different marketing people who wrote the spec. 8 ohm drivers typically vary between a DC resistance of ~4.8 - 7ohms.
For complete loudspeaker systems, I would say that the rating of 1W/1m is even more arbitrary, and the 2.83V rating is the only one which makes sense. I wouldn't even try to decide using manufacturer specifications if a complete loudspeaker were more sensitive than another or not. Only an actual measurement would suffice. There is no legal standard for such ratings that I am aware of, and the manufacturer can choose to add several dB to account for the effects of placing the system in a reverberant environment.
My take on the question was that one wishes to convert the specs from a driver manufacturer for comparison purposes. I think that due to the arbitrary nature of nominal impedance, it is not hard to see that using calculated efficiency or sensitivity is the most fair and unambiguous method to compare individual drivers, especially those from different manufacturers. After realizing that calculated sensitivity was a much better metric for comparison purposes, I read that acousticians have been debating this topic for 50 years - hence the concepts of "Power Available Efficiency" from Beranek (1954), "Basic efficiency" from Thiele (1961) and "Reference efficiency" from Benson (1971). All come to the same conclusion that I have.
This definition of efficiency or sensitivity does not represent the actual electroacoustic efficiency at any point (not even the minimum above resonance). In my experience, it does represent the rated sensitivity or efficiency of the driver as quoted in specification sheets of many manufacturers.
So in end, we agree on most things. If one wishes to use nominal impedance, one must accept that their result, or the results of their comparison, is nominal.
For complete loudspeaker systems, I would say that the rating of 1W/1m is even more arbitrary, and the 2.83V rating is the only one which makes sense. I wouldn't even try to decide using manufacturer specifications if a complete loudspeaker were more sensitive than another or not. Only an actual measurement would suffice. There is no legal standard for such ratings that I am aware of, and the manufacturer can choose to add several dB to account for the effects of placing the system in a reverberant environment.
My take on the question was that one wishes to convert the specs from a driver manufacturer for comparison purposes. I think that due to the arbitrary nature of nominal impedance, it is not hard to see that using calculated efficiency or sensitivity is the most fair and unambiguous method to compare individual drivers, especially those from different manufacturers. After realizing that calculated sensitivity was a much better metric for comparison purposes, I read that acousticians have been debating this topic for 50 years - hence the concepts of "Power Available Efficiency" from Beranek (1954), "Basic efficiency" from Thiele (1961) and "Reference efficiency" from Benson (1971). All come to the same conclusion that I have.
This definition of efficiency or sensitivity does not represent the actual electroacoustic efficiency at any point (not even the minimum above resonance). In my experience, it does represent the rated sensitivity or efficiency of the driver as quoted in specification sheets of many manufacturers.
So in end, we agree on most things. If one wishes to use nominal impedance, one must accept that their result, or the results of their comparison, is nominal.
sreten, kelticwizard, thanks for yor input.
As opposed to mfg with "corrupted" information, I've found an example of mfg offering very good information*: http://www.d-s-t.com (home of Peerless, Vifa, Scan-Speak). It seems their graphs show SPL at 2.83V, and an impedance curve is showed as well. Since I'll be using an active XO solution I don't care that much about impedance, as long as it doesn't go so low as to make the amp run out of steam. What I want is ruler-flat SPL from X to Y Hz, measured at a specific voltage [the amp outputs volts, the amperes is just a "consequence" of the speakers impedance], and now I've understood how to read mfg graphs, thanks everyone for enlightening me!
* = I'm not affiliated with DST in any way, I just happen to like their graphs.
As opposed to mfg with "corrupted" information, I've found an example of mfg offering very good information*: http://www.d-s-t.com (home of Peerless, Vifa, Scan-Speak). It seems their graphs show SPL at 2.83V, and an impedance curve is showed as well. Since I'll be using an active XO solution I don't care that much about impedance, as long as it doesn't go so low as to make the amp run out of steam. What I want is ruler-flat SPL from X to Y Hz, measured at a specific voltage [the amp outputs volts, the amperes is just a "consequence" of the speakers impedance], and now I've understood how to read mfg graphs, thanks everyone for enlightening me!
* = I'm not affiliated with DST in any way, I just happen to like their graphs.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.