I checked the Parts Express Web site two days ago and they now have test results for the W4-1052SA posted. The sample they tested appears to have a greater response variance than the two I tested. I also checked Tang Band’s Web site and found no listing for the W4-1052SA. Earlier in this thread, there was what appears to be a copy of a Tang Band W4-1052SA Web site page attached to a post. If that was from the Tang Band site, it is now gone.
Tang Band still has a page for their structured PP diaphragm listed with the model number W4-927SA. The 927 and the 1052 appear identical except for the mounting lip on the basket. I was interested to see that their response graph for the 927 looks much more like mine and Parts Express’ testing of the 1052 than the 1052 did.
If Tang Band has posted a more representative frequency response graph of one of their driver’s performance, then I commend them. If we want manufacturers to publish accurate and detailed performance evaluations, then we must buy drivers so represented. The question is, with the performance shown in the W4-927SA graph attached to this post, would you be as interested in the driver as you were when you saw Tang Band’s original page for the 1052? What manufacturer wants to go first and publish high definition test results that make their drivers look worse than their competitors even if they are acoustically better?
Mark
Tang Band still has a page for their structured PP diaphragm listed with the model number W4-927SA. The 927 and the 1052 appear identical except for the mounting lip on the basket. I was interested to see that their response graph for the 927 looks much more like mine and Parts Express’ testing of the 1052 than the 1052 did.
If Tang Band has posted a more representative frequency response graph of one of their driver’s performance, then I commend them. If we want manufacturers to publish accurate and detailed performance evaluations, then we must buy drivers so represented. The question is, with the performance shown in the W4-927SA graph attached to this post, would you be as interested in the driver as you were when you saw Tang Band’s original page for the 1052? What manufacturer wants to go first and publish high definition test results that make their drivers look worse than their competitors even if they are acoustically better?
Mark
Attachments
http://www.partsexpress.com/pdf/264-828g.pdf
I don't know about you, guys, but those breakups are not my taste...
I don't know about you, guys, but those breakups are not my taste...
Yes amazing what 1-10 smoothing can do. A TON
Look at the 30 off .Talk about dive city
This is one of the issues with the new XBL 4" CSS is doing that I have told Bob and Dan is a issue.
The 4" drivers from TB all are done before 12K none of them really go past that then the off axis on the drops at 10K no higher.
XBL on the photo type that was at CES has more upper end over the rest but not to the 18-20 k that is possible.
Low end well XBL talks.
The upper end and mids better with XBL as these harmonic distortion from XBL is very very low.
So the XBL will work on the 4" just a matter of a few other issues that are being looked at now.
plug changes ,cone changes, cast frame over stamped as on the photo types..
Just to let you all know what is going on with the 4"
Look at the 30 off .Talk about dive city
This is one of the issues with the new XBL 4" CSS is doing that I have told Bob and Dan is a issue.
The 4" drivers from TB all are done before 12K none of them really go past that then the off axis on the drops at 10K no higher.
XBL on the photo type that was at CES has more upper end over the rest but not to the 18-20 k that is possible.
Low end well XBL talks.
The upper end and mids better with XBL as these harmonic distortion from XBL is very very low.
So the XBL will work on the 4" just a matter of a few other issues that are being looked at now.
plug changes ,cone changes, cast frame over stamped as on the photo types..
Just to let you all know what is going on with the 4"
A key point of my last post was whether you really wanted to know what a driver does. If people actually showed you the performance of a driver, would you still buy it?
It is possible to like the sound of a driver that is not accurate. It is possible to dislike the sound of an accurate driver. This is a separate problem. Parts Express does not publish measurement responses for the A11EC80-02F. While different than liking the sound, if you thought the sound was accurate, you are not going to like the measured response of this driver. I am sorry, but it is worse than the Tang Band W4-1052SA. And it is not just a little bit worse.
I like the sound of an accurate driver. The more accurate the driver, the better I like the sound. Because of that I find value in testing. Testing, however, is not perfect. There are ways, from simple to complex, to misrepresent the performance of a driver with available test rigs. It is not enough just to have someone post a bode plot.
In a perfect world, everyone would publish the most detailed and accurate results of their driver's performance. Then we would all know just how far we have to go to get to sound "reproduction." And we have very far to go.
We also have far to go before we can agree on what is needed for accurate sound reproduction. For example, in a full-range driver, just how important is 30 degree off-axis performance?
Mark
MarkMcK said:
I wan't trying to imply that it was the most accurate driver, it's certainly not my first choice in drivers overall. However, it is fairly good for the money..
Given that little bit of information, the only good answer is "that depends'... Given different rooms, enclosure/baffle/horn designs, etc. the answer to that question alone can vary greatly...
Peace
TB North American Distribution
TB also have an official distributor in Burnaby BC Canada. NUERA Acoustics. info@nuera-acoustic.ca
http://www.nuera-acoustic.ca
All prices on the web site are US Dollars and they are happy to ship to the USA.
Check out their clearance specials.
TB also have an official distributor in Burnaby BC Canada. NUERA Acoustics. info@nuera-acoustic.ca
http://www.nuera-acoustic.ca
All prices on the web site are US Dollars and they are happy to ship to the USA.
Check out their clearance specials.
MarkMcK said:
Mark;
That 1st graph on the TB W4-1052SA that I posted earlier in this
thread was not from a TB web page but was rather emailed to
me by request from TB. Seeing that Parts Express had no specs
posted at the time & TB had not even a posting of the driver let alone specs I simply emailed TB asking if specs were available & about a day later I received a courteous reply with that spec
sheet. Seems that 1st graph though may have been more of a prediction or wishful thinking on their part at the time than an actual measured response. Who knows?
I've come across another rather interesting spec sheet discrepancy for a TB driver. For the W69-1042 (6x9 woofer), the graph & the TS parameters posted at TB are in total disagreement with those posted at Nuera. I emailed Neura about 1 & 1/2 months ago questioning the discrepancy & they promptly replied thanking me for pointing it out & that they would question TB & get back to me as soon as they found out but I've not heard from them since. I suppose I'll email TB. I suspect the Nuera posted specs for that driver are an early out dated release & in greater error as for one thing The "EFFECTIVE PISTON AREA" spec of 0.014 M sq., (shown by Nuera's data sheet) is of particular suspect to me since that would be a value more typical of a 6" than a 6"x9" speaker. The value of 0.0226 M sq. (shown by the data sheet on Tang Bands website) makes more sense.
It sucks not being able to trust published specs.
i like TB too....i was looking at the 927 a while back...then at the FE 103 and 108N....
seems to me a lot of us are entranced by Xmax values. I have reason to believe that bigger Xmax numbers do not mean a nessacrily better driver. Also there seems to be a pattern where drivers with higher xmax are less sensitive than drivers with low xmax.
so the questions i ask are:
how much xmax do we need? say to produce 90db/1m/50Hz or 95db/2m/100Hz etc ?
how does one determine the "quality" of the driver before one actually purchases a few units, tets it using FFT, MLSSA, etc... and listening tests?
The aural memory is quite short (assumed since we can listen to a musical score 100 times but watching a movie 3 times is usually boring) and hence how does one judge/compare 10 or even 5 drivers haivng simialr specs (say 3-4" fullrange that operate from 100hz-10k+)?
seems to me a lot of us are entranced by Xmax values. I have reason to believe that bigger Xmax numbers do not mean a nessacrily better driver. Also there seems to be a pattern where drivers with higher xmax are less sensitive than drivers with low xmax.
so the questions i ask are:
how much xmax do we need? say to produce 90db/1m/50Hz or 95db/2m/100Hz etc ?
how does one determine the "quality" of the driver before one actually purchases a few units, tets it using FFT, MLSSA, etc... and listening tests?
The aural memory is quite short (assumed since we can listen to a musical score 100 times but watching a movie 3 times is usually boring) and hence how does one judge/compare 10 or even 5 drivers haivng simialr specs (say 3-4" fullrange that operate from 100hz-10k+)?
I believe the 927 and 1052 are identical except for the mounting rim on the basket. If NUERA will carry the 927 you might consider this driver instead of the 1052. The 1052 may be a model number exclusively for Parts Express.
Just one word of caution about these baskets, a thick baffle will close the basket vent voids and alter the performance of the driver. A thin baffle insert where the driver is mounted or a relieved rear baffle to keep the basket cutouts as open as possible is a good thing.
Navin, I could be wrong about this but I do not believe that frequency has much to do with Xmax to maximum spl comparisons. It will take the same excursion to produce a given spl at 100 Hz as at 200 Hz as at 1 kHz. It is just that music has more energy in the lower frequencies than in the higher. It is just that when playing music the higher spls will be at the lower frequencies.
There are exceptions to this of course. Those exceptions are the more extreme T/S alignments, those with eq, and those trying to produce sound below Fs. Since you specified a frequency of 50 Hz when the driver Fs is 65 Hz, you may be trying to do this.
As for how we decide on a driver, I can only tell you that I test and evaluate everything myself. I have been doing impulse testing for so long I can qualitatively evaluate a driver by its click sound. I use mathematical manipulations of the clicks only to help me design corrections and adaptations for the driver performance. I also have well documented standards of comparison for listening tests. It does help to have a reference of high accuracy. For most people going it alone, I know it is very difficult to tell that you are making progress toward greater accuracy.
Drivers are also hard to compare without a set standard or convention of testing. While we now know better how the 1052 performs acoustically, we don't know how it compares to the Fostex or any other extended range driver. Even the more expensive Jordan’s and Bandor’s test differently than you might expect or assume. That does not mean that they are not quality drivers or do not sound nice, it just means that it is hard to decide without doing your own evaluation (testing and listening).
Fortunately highly sophisticated analysis equipment is now easily and affordably available to make quasi-anechoic testing available to all. You do still have to learn how to use the system and to understand what the results mean, but that can be learned without having to be taught.
Mark
Just one word of caution about these baskets, a thick baffle will close the basket vent voids and alter the performance of the driver. A thin baffle insert where the driver is mounted or a relieved rear baffle to keep the basket cutouts as open as possible is a good thing.
Navin, I could be wrong about this but I do not believe that frequency has much to do with Xmax to maximum spl comparisons. It will take the same excursion to produce a given spl at 100 Hz as at 200 Hz as at 1 kHz. It is just that music has more energy in the lower frequencies than in the higher. It is just that when playing music the higher spls will be at the lower frequencies.
There are exceptions to this of course. Those exceptions are the more extreme T/S alignments, those with eq, and those trying to produce sound below Fs. Since you specified a frequency of 50 Hz when the driver Fs is 65 Hz, you may be trying to do this.
As for how we decide on a driver, I can only tell you that I test and evaluate everything myself. I have been doing impulse testing for so long I can qualitatively evaluate a driver by its click sound. I use mathematical manipulations of the clicks only to help me design corrections and adaptations for the driver performance. I also have well documented standards of comparison for listening tests. It does help to have a reference of high accuracy. For most people going it alone, I know it is very difficult to tell that you are making progress toward greater accuracy.
Drivers are also hard to compare without a set standard or convention of testing. While we now know better how the 1052 performs acoustically, we don't know how it compares to the Fostex or any other extended range driver. Even the more expensive Jordan’s and Bandor’s test differently than you might expect or assume. That does not mean that they are not quality drivers or do not sound nice, it just means that it is hard to decide without doing your own evaluation (testing and listening).
Fortunately highly sophisticated analysis equipment is now easily and affordably available to make quasi-anechoic testing available to all. You do still have to learn how to use the system and to understand what the results mean, but that can be learned without having to be taught.
Mark
MarkMcK said:
Mark, Thanks for you quick and detailed answer. I dont have the theotritcal background you do but my reasoning concerning SPL goes like this....
Producing SPLs is a function of the volume of air moved. (Xmax x Sd). However the ear is not a linear machine (Fletcher Munson curves and it's variants document this). Hence one needs to move move air at lower freq to have the same percived SPL.
Now how do we determine what SPL we need to produce 95db at 2 m at 100Hz (since 50Hz is below the resonant freq of the driver in question). I prefer figures at 2 m only because i dont know a situation where one will be 1 m from each speaker. 3m would be more realistic but might be too strenous a parameter.
BTW let me state this now...a speaker that say is 88db sens with a 6mm Xmax might not produce as much SPL as a speaker that is 100db sens with a 1mm Xmax. So SPL is not only a function of the volume of air moved but also a function of the speaker sensitivity.
A mix up.
Hi Navin,
You are slightly off target.
The max SPL possible is determined by the max linear excursion that is available. The sensitivity does not play a part in that.
Sensitivity will however tell you how many electrical watts you need to put into the speaker to achieve that max SPL. A less sensitive speaker will need more electrical watts to give you the same max SPL - with the same Xmax.
Max SPL is excursion limited as frequency goes down. You need more Xmax to get the same SPL as frequency goes down. Since Xmax is fixed the max available SPL will drop with decreasing frequency.
Sensitivity will not determine that. However sensitivity is tied to the Q which is also involved in determining how the bass roll off is and thus how much power is required to reach Xmax. Like a Q of 0.5 needs more power to reach Xmax ( at a low frequency ) than one with 0.7 .
Hi Navin,
You are slightly off target.
The max SPL possible is determined by the max linear excursion that is available. The sensitivity does not play a part in that.
Sensitivity will however tell you how many electrical watts you need to put into the speaker to achieve that max SPL. A less sensitive speaker will need more electrical watts to give you the same max SPL - with the same Xmax.
Max SPL is excursion limited as frequency goes down. You need more Xmax to get the same SPL as frequency goes down. Since Xmax is fixed the max available SPL will drop with decreasing frequency.
Sensitivity will not determine that. However sensitivity is tied to the Q which is also involved in determining how the bass roll off is and thus how much power is required to reach Xmax. Like a Q of 0.5 needs more power to reach Xmax ( at a low frequency ) than one with 0.7 .
Re: A mix up.
thanks i meant the same as what you mentioned maybe i was not clear. afterall we input watts into a speaker. so max SPL wuld be related to teh number of watts the speaker can heandle (at a particualr freq) before it's excursion is limited (or it has reached it's thermal limits) + the sens per watt. so if a speaker is 90db sens. and at 50hz will max out at 20W the max SPL is 103db.
question is if we know the sens and Xmax of a speaker how do we calculate the Max SPL at say 100hz.
ashok said:Sensitivity will however tell you how many electrical watts you need to put into the speaker to achieve that max SPL.
Max SPL is excursion limited as frequency goes down. You need more Xmax to get the same SPL as frequency goes down.
thanks i meant the same as what you mentioned maybe i was not clear. afterall we input watts into a speaker. so max SPL wuld be related to teh number of watts the speaker can heandle (at a particualr freq) before it's excursion is limited (or it has reached it's thermal limits) + the sens per watt. so if a speaker is 90db sens. and at 50hz will max out at 20W the max SPL is 103db.
question is if we know the sens and Xmax of a speaker how do we calculate the Max SPL at say 100hz.
Greetings all,
The last few posts have taken this thread into a new direction. To bring it back and make clear the question I believe is being asked let me try to restate it.
Just how loudly will the W4-927SA or W4-1052SA play at different frequencies as measured (or listened to) two-meters away from the loudspeaker?
Right now I do not know the answer to that. I do know that an upper limit can be determined by two factors. One is the thermal limit of the driver. The second is the Xmax (how much air the driver can displace). Neither of these are absolute. For example, if you can tolerate increasing distortion, a driver can excurse beyond Xmax and will still produce a higher SPL.
I also know that the answer is going to be conditional and complicated. The various resonant structures of the driver and the enclosure will impact both magnitude of output and cone displacement. (The idea of piston motion is just a myth. I have never seen a driver acting at a piston for any range of frequencies, no matter how narrow).
Think about what happens with a ported enclosure, at box resonant frequency, the output magnitude is large, but the driver's displacement is small. Go below that frequency and suddenly much more excursion is needed to produce the same volume of sound. That is why some of the more recent simulators also graph Xmax along with frequency output.
I also have an idea or theory as to how I can determine the Xmax/thermal/max SPL relationship. No promises, but I will see if I can find the time to conduct the experiment.
Mark
The last few posts have taken this thread into a new direction. To bring it back and make clear the question I believe is being asked let me try to restate it.
Just how loudly will the W4-927SA or W4-1052SA play at different frequencies as measured (or listened to) two-meters away from the loudspeaker?
Right now I do not know the answer to that. I do know that an upper limit can be determined by two factors. One is the thermal limit of the driver. The second is the Xmax (how much air the driver can displace). Neither of these are absolute. For example, if you can tolerate increasing distortion, a driver can excurse beyond Xmax and will still produce a higher SPL.
I also know that the answer is going to be conditional and complicated. The various resonant structures of the driver and the enclosure will impact both magnitude of output and cone displacement. (The idea of piston motion is just a myth. I have never seen a driver acting at a piston for any range of frequencies, no matter how narrow).
Think about what happens with a ported enclosure, at box resonant frequency, the output magnitude is large, but the driver's displacement is small. Go below that frequency and suddenly much more excursion is needed to produce the same volume of sound. That is why some of the more recent simulators also graph Xmax along with frequency output.
I also have an idea or theory as to how I can determine the Xmax/thermal/max SPL relationship. No promises, but I will see if I can find the time to conduct the experiment.
Mark
Roscoe,
Thanks for the advice. Rest assured, however, that since I am going to be conducting empirical experiments instead of just modeling or simulating, that if frequency is a tightly coupled variable I will be able to determine it. I will try to answer my restatement of the question (see my previous post, paragraph number 2).
In the meantime, if you would be willing to explain why and how frequency determines max SPL, I would be glad to read your thinking on this. Assuming that the driver is operating within the range of Fs to mass/inductive loading limit, what is the mechanism that requires more excursion to produce the same volume displacement at lower frequencies? I need someone to explain this because it just doesn't make any sense to me.
I hope people are not confusing perceived loudness and SPL. The two are not the same.
Thanks,
Mark
Thanks for the advice. Rest assured, however, that since I am going to be conducting empirical experiments instead of just modeling or simulating, that if frequency is a tightly coupled variable I will be able to determine it. I will try to answer my restatement of the question (see my previous post, paragraph number 2).
In the meantime, if you would be willing to explain why and how frequency determines max SPL, I would be glad to read your thinking on this. Assuming that the driver is operating within the range of Fs to mass/inductive loading limit, what is the mechanism that requires more excursion to produce the same volume displacement at lower frequencies? I need someone to explain this because it just doesn't make any sense to me.
I hope people are not confusing perceived loudness and SPL. The two are not the same.
Thanks,
Mark
MarkMcK said:
It's just basic physics. I won't take the time to explain it in detail here, it should be available in any acoustics text. But look at the empirical evidence: If it DIDN'T take more excursion (displacement actually, which is excursion X area) then if we can get 100dB at a meter out of a 1" dome tweeter at 5kHz, why do we need a larger woofer to reproduce the same SPL at 50Hz?
Peace
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.