i've read quite a lot regarding the accuracy of the quoted specs for audio nirvana drivers.
is the general consensus still that they are generally poor accuracy if not total fabrications?
thanks.
is the general consensus still that they are generally poor accuracy if not total fabrications?
thanks.
I haven't seen any user measures posted for a while; some of those I've seen in the past deviated significantly from the published, but IIRC Freddi had some 8in units that were reasonably close to the data sheets, and I believe Nelson likes his well enough so YMMV.
Last edited:
i ask because i noticed this.
super 10 on commonsense audio
QES - 0.3
QTS - 0.278
QMS - 3.758
VAS - 238.783
BL - 9.065
super 10 listed elsewhere, and stated as the specs from 'audio nirvana'
QES - 0.28
QTS - 0.25
QMS - 3.4
VAS - 173
BL - 11.2
super 8 on commonsense
QES - 0.169
QTS - 0.162
QMS - 3.836
VAS - 66.495
BL - 11.84
super 8 elsewhere
QES - 0.184
QTS - 0.175
QMS - 3.38
VAS - 73.3
BL - 11.02
my suspicions were fully aroused when i asked about shipping prices and got a full sales pith about how i couldnt buy a 'better sounding speakers, anywhere, for any price'
i'm not sure what to think with this. they're reasonably priced but if those values change so much that would make me think that the quality control is terrible and if the quality control is terrible and they're as cheap as they are then the speaker itself probably isn't going to be as good as they're being made out?
how do you reasonably design a cabinet (which is hard enough for me to understand and do as it is) if the 2 speakers you get have wildly varying specs?
We need to compare how well other full range driver specs compare with the published specs. Now, I'm sure the MA audio guys will scream at the top of their voices about how their specs match up. But look at drivers that are similar to ANs. The fostexes, Lowthers, hi-eff Tang Bands. They are likely to show similar variation.
Cabinet design also provides a large margin of error for driver parameters, contrary to what cabinet designers will have you believe. Important parameters are the Vas and Qts. And look at the absolute variation, not the percentage. For example, the two super8 specs you've provided above, the Qts difference is 0.013 and the Vas difference is about 7. IMO, this will not even show up on the frequency response, let alone be audible.
It is obvious the ANs are not as well made as the more expensive drivers. But take them for what they are. A cheap alternative to the Lowthers. Plus they have NP's endorsement. What more can you ask for?
I'd say ignore the naysayers. Most of them have a financial interest anyway, and they wish the ANs would go away.
To get a good balance from the ANs, you'd need a good horn. Don't go for BR, you'll have no bass. Some EQ for flattenning the top end also helps. Nelson has provided the details if you can find them.
Cabinet design also provides a large margin of error for driver parameters, contrary to what cabinet designers will have you believe. Important parameters are the Vas and Qts. And look at the absolute variation, not the percentage. For example, the two super8 specs you've provided above, the Qts difference is 0.013 and the Vas difference is about 7. IMO, this will not even show up on the frequency response, let alone be audible.
It is obvious the ANs are not as well made as the more expensive drivers. But take them for what they are. A cheap alternative to the Lowthers. Plus they have NP's endorsement. What more can you ask for?
I'd say ignore the naysayers. Most of them have a financial interest anyway, and they wish the ANs would go away.
To get a good balance from the ANs, you'd need a good horn. Don't go for BR, you'll have no bass. Some EQ for flattenning the top end also helps. Nelson has provided the details if you can find them.
What Mark will say, and I believe is that if you do not exactly reproduce the conditions under which the drivers were measured. In my case, I complained for a long time that the Fostex measurements were way off what I measured. But I found that if I used the factory measurements for design, it worked perfectly.
The problem with user measurements is that they are done at different times under different condition with different measuring equipment. Well of course the answers are going to be different. If you are going to make claims about QC, you need to measure at least a half dozen drivers under the same conditions. P10 has done that with several drivers. Once upon a time he posted the results here. Perhaps Dave will comment?
Bob
The problem with user measurements is that they are done at different times under different condition with different measuring equipment. Well of course the answers are going to be different. If you are going to make claims about QC, you need to measure at least a half dozen drivers under the same conditions. P10 has done that with several drivers. Once upon a time he posted the results here. Perhaps Dave will comment?
Bob
Edit -Bob beat me to it, and a lot more eloquently. I'll leave my post up though.
Purely based on the published spec., depends on the model as to whether it necessarily 'needs' a back horn or not. It would make life easier in some cases (on the understanding that back-horns entail their own compromises, like anything else does), not necessarily so in others. Since a back-horn is just an extreme variation on a vented box, like any other such cabinet they're only useful over a limited BW -up to ~300Hz, give or take, and if the driver's mass corner is higher than that once the amplifier's output impedance or wire resistance etc. is accounted for, then you'll either need Eq, or a short front horn to fill-in the gap between the bass horn's upper corner frequency & the driver's inherent Fh[m].
Agreed re comparisons, and as I often point out, comparisons are only really useful if you're comparing like with like (or as-near-as). For the AN units, of current manufacturers, that basically means Fostex. FWIW, my experience has been that the Fostex data is pretty solid, and variation within about 10%, which is workable. Some exceptions of course, but as a general rule that seems to hold good, suggesting that is the production tolerance they work to. GM, I and many others have used their published data for years with good results, suggesting that it's pretty representative.
Speaking generally, a point which is not always appreciated is that T/S parameters to an extent change according to measurement conditions. Most of the equipment that DIYers have access to measures with a low voltage, with the result that Q, Fo for example. often appear higher than the figures supplied by manufacturers, who generally use more voltage when measuring. Anecdote: a couple of years back, DaveD measured a pair of FE127Es on his equipment, with the aforementioned results. That same pair of drivers were then immediately packed up & shipped to Mark Fenlon (Mark Audio) who measured them on the gear he uses to test his own units. Result: data which was much closer to the factory figures. Very close in fact. Question: Which was 'correct?' Answer: Both. See above regarding successful use of their published data.
Another issue at work -if you back-calculate the figures provided on some manufacturer data sheets, you will sometimes find it appears to contradict itself. That may be because they've used average values across a batch. Or they may be being economical with the actualite. Or a combination of these, and other factors.
Purely based on the published spec., depends on the model as to whether it necessarily 'needs' a back horn or not. It would make life easier in some cases (on the understanding that back-horns entail their own compromises, like anything else does), not necessarily so in others. Since a back-horn is just an extreme variation on a vented box, like any other such cabinet they're only useful over a limited BW -up to ~300Hz, give or take, and if the driver's mass corner is higher than that once the amplifier's output impedance or wire resistance etc. is accounted for, then you'll either need Eq, or a short front horn to fill-in the gap between the bass horn's upper corner frequency & the driver's inherent Fh[m].
Agreed re comparisons, and as I often point out, comparisons are only really useful if you're comparing like with like (or as-near-as). For the AN units, of current manufacturers, that basically means Fostex. FWIW, my experience has been that the Fostex data is pretty solid, and variation within about 10%, which is workable. Some exceptions of course, but as a general rule that seems to hold good, suggesting that is the production tolerance they work to. GM, I and many others have used their published data for years with good results, suggesting that it's pretty representative.
Speaking generally, a point which is not always appreciated is that T/S parameters to an extent change according to measurement conditions. Most of the equipment that DIYers have access to measures with a low voltage, with the result that Q, Fo for example. often appear higher than the figures supplied by manufacturers, who generally use more voltage when measuring. Anecdote: a couple of years back, DaveD measured a pair of FE127Es on his equipment, with the aforementioned results. That same pair of drivers were then immediately packed up & shipped to Mark Fenlon (Mark Audio) who measured them on the gear he uses to test his own units. Result: data which was much closer to the factory figures. Very close in fact. Question: Which was 'correct?' Answer: Both. See above regarding successful use of their published data.
Another issue at work -if you back-calculate the figures provided on some manufacturer data sheets, you will sometimes find it appears to contradict itself. That may be because they've used average values across a batch. Or they may be being economical with the actualite. Or a combination of these, and other factors.
Last edited:
Well said, Scott!
Yes, the ANs (and the Lowthers) probably could use a front horn. See Thorsten's design.
Regarding the interaction with David of CommonSense, I too have found him to be an eccentric guy. Not all that he says makes sense, which is funny considering the name of his company. But that doesn't mean his products are bad. And I'll take his sales pitch anytime over the unethical hard selling that happens in the full-range forum. At least David is open and blunt about it.
Yes, the ANs (and the Lowthers) probably could use a front horn. See Thorsten's design.
Regarding the interaction with David of CommonSense, I too have found him to be an eccentric guy. Not all that he says makes sense, which is funny considering the name of his company. But that doesn't mean his products are bad. And I'll take his sales pitch anytime over the unethical hard selling that happens in the full-range forum. At least David is open and blunt about it.
I'm sure he'd like to if it didn't result in more of ....
a slight bit hyperbolic, perhaps ?
not that I've particularly researched the matter, but does Mr Dicks post on any forums, and if so do those conversations get any more animated than herein?
Exactly right. T/S measurements vary as measured by different people not just because of inaccuracies of the measurements due to different measurement procedures, but because the T/S parameters themselves DO in fact change with drive levels.
Specifically, apparent (more on that in a moment) suspension compliance Cms can vary greatly with excursion which means Fs, Qms, and Vas all change with drive level.
I talked about this quite a bit in another thread starting in post 13:
http://www.diyaudio.com/forums/multi-way/182688-help-understanding-physics-vas.html#post2461764
In post 18 I presented the measurements of two drivers measured at different drive levels, an 8" full range driver with cloth surround, and a 12" woofer with rubber surrounds:
http://www.diyaudio.com/forums/multi-way/182688-help-understanding-physics-vas.html#post2461984
The full range driver varied the most - Fs varied from 48Hz at very low drive levels (millivolts) all the way down to 34.5Hz at a drive level of 2.3 volts. (About half Xmax at resonance) There was an almost 2 to 1 change in Vas and Qms as well. There was a smaller but still quite significant change with the 12" woofer.
In post 28 I graphed the change in Fs with drive level:
http://www.diyaudio.com/forums/multi-way/182688-help-understanding-physics-vas.html#post2462768
Although it wasn't really resolved in that original thread I now believe that the reason for the change in Fs with excursion is simply due to the effects of a non-linear compliance on a harmonic oscillator. (eg the cone mass and suspension compliance)
If the compliance is perfectly linear the resonant frequency of a harmonic oscillator is independent of amplitude, however if the compliance is non-linear with excursion the resonant frequency will actually change with the amplitude of the oscillation.
Depending on the shape of the non-linearity it can either decrease or increase resonant frequency with increasing excursion - it seems that at least with these two drivers (perhaps most drivers) the compliance non-linearity is such that Fs falls with increasing excursion.
Since the moving mass must remain constant this non-linearity manifests as an apparent change in the compliance Cms. In reality the Cms is not a single unchanging value but varies with excursion.
The important thing to realise is that although Cms, Fs, Qms, and Vas all change, the only independant variable that is changing is Cms, all the others are dependant variables which are maintaining their relationships with each other.
As soon as the driver is in a box the compliance is greatly linearised, so any apparent change in Cms with amplitude is swamped, so I think in practice the apparent differences in T/S parameters as measured at different drive levels matters very little - any set of measurements taken at different drive levels will give a very similar bass alignment in a box simulator.
From what I can see while most DIY'ers measure T/S at very low drive levels, many manufacturers seem to measure and specify the T/S parameters at a high drive level - perhaps at close to Xmax excursion.
For example the Visaton W300S manufacturers figure for Fs is 23Hz - if I measure it at a very low drive level I get 29Hz, but if I measure it at 5.3v RMS - half Xmax excursion at resonance, I get 24Hz, much closer to the manufacturers claim and I'm sure if I increased the drive a bit more it would drop to 23Hz.
The bottom line is that T/S parameters do vary with drive level, it's not just "inaccuracies" due to different measuring equipment or techniques. In the tests I performed in the other thread the measurement technique was identical for all measurements - the only thing varied was the drive level.
Last edited:
I have had a direct experience of this. I have measured literally 1000+ drivers. Rarely do the numbers i obtain look like the factory specs.
I measured a pair of FE127eN and then shipped them Mark Fenlon who then measured them using one of the 2 packages that are standard in the industry (LMS or LEAP). His results were well within tolerence of factory specs, mine for the same driver were way different.
This really rammed home something Dan Wiggins has been saying for some time... T/S are not scalars, they are curves, largely a function of drive, but also weather (temp, humidity, and air pressure).
If i see measures that come close to the factory, at this point a supposition on my part, that the T/S curves are close to horizontal.
I do my 1st iteration of any design with the factory numbers. Then i work from there. Rarely have i had to modify a box for reasons that would reflect on the T/S. I do tend to design boxes that are purposely aimed at working on dynamic changes in T/S so that helps.
I have only had a chance to measure 3 pair of Audio Nirvana, listen to 2 pair, and have designed a box for an AN i have never seen (the one build of that design has the builder quite happy with them from all reports)
I try out as many drivers as i can. I am going to try more ANs (Cast 8 & Cast 15 high on the list), but the queue is deep -- i still have the SEAS 8" here collecting dust, and only recently has the 3" Scan-Speak had any attention. Most energies (and cash flow) are directed at trying to keep up with demand for drivers that we do know work well.
Besides not being impressed with either of the AN i have heard (mind you, both the stamped basket cheapies) there are 2 things that bother me. David Dicks does not use T/S to design his boxes (a dramatic change in the vent parameters for one of the pair i did hear resulted in significant gains) so is not likely invested in their reality, but more so, when i get a pair of drivers that measure like this:
dave
Attachments
Let me get home tonight and find my measurements of the bog standard AN8 I have. Both drivers measure pretty close to each other.
Dave, I don't doubt that these are honest measurements because one of the units I bought had the coil rubbing against the gap (David replaced the driver). So, there may be problems with the sample you received. But as seller of fostex and MA drivers, it is not right to post your personal measurements of a competitor's driver. I remember Mark Fenlon having the same complaint about Zaph's measurements of his drivers. You don't see NP posting measurements of ML amps, do you?
And someone has deleted my previous post. Do I dare ask for a reason or who did it?
Dave, I don't doubt that these are honest measurements because one of the units I bought had the coil rubbing against the gap (David replaced the driver). So, there may be problems with the sample you received. But as seller of fostex and MA drivers, it is not right to post your personal measurements of a competitor's driver. I remember Mark Fenlon having the same complaint about Zaph's measurements of his drivers. You don't see NP posting measurements of ML amps, do you?
And someone has deleted my previous post. Do I dare ask for a reason or who did it?
This is an interesting discussion for me. Thanks to all.
If a drivers T/S parameters change with drive level, is it possible there is non linearity to be concerned about. I've been hypothesizing lately that Klippel testing is related to this issue. Thoughts?
If a drivers T/S parameters change with drive level, is it possible there is non linearity to be concerned about. I've been hypothesizing lately that Klippel testing is related to this issue. Thoughts?
ok,
the threads i read on the subject made these discrepancies out to be a lot worse than this thread has now pretty much concluded at.
i was only asking as i'm considering audio-nirvana drivers as a next step up the ladder financially and hopefully performance wise for a beginner to probably abuse in their pursuit of their dream system (which they do not yet know what that dream is). i was slightly alarmed at quite a lot of negativity aimed at them and claims of their creator being, in short, a liar.
the threads i read on the subject made these discrepancies out to be a lot worse than this thread has now pretty much concluded at.
i was only asking as i'm considering audio-nirvana drivers as a next step up the ladder financially and hopefully performance wise for a beginner to probably abuse in their pursuit of their dream system (which they do not yet know what that dream is). i was slightly alarmed at quite a lot of negativity aimed at them and claims of their creator being, in short, a liar.
Mark had an objection in his forum where he keeps anything not MA out. People were using Zaph's amateur measures to question the professional level measures that Mark posted. I'd object as well. I see no reason not to post data i have accumulated. It is what it is and i have stated i use factory data for design. I use impedance curves for QC & driver matching where relative numbers are fine..
If people want to buy them and D Dicks will sell them to me i'll do ANs up, just like any other special order driver.
dave
I have had a direct experience of this. I have measured literally 1000+ drivers. Rarely do the numbers i obtain look like the factory specs.
I measured a pair of FE127eN and then shipped them Mark Fenlon who then measured them using one of the 2 packages that are standard in the industry (LMS or LEAP). His results were well within tolerence of factory specs, mine for the same driver were way different.
This really rammed home something Dan Wiggins has been saying for some time... T/S are not scalars, they are curves, largely a function of drive, but also weather (temp, humidity, and air pressure).
If i see measures that come close to the factory, at this point a supposition on my part, that the T/S curves are close to horizontal.
I do my 1st iteration of any design with the factory numbers. Then i work from there. Rarely have i had to modify a box for reasons that would reflect on the T/S. I do tend to design boxes that are purposely aimed at working on dynamic changes in T/S so that helps.
I have only had a chance to measure 3 pair of Audio Nirvana, listen to 2 pair, and have designed a box for an AN i have never seen (the one build of that design has the builder quite happy with them from all reports)
I try out as many drivers as i can. I am going to try more ANs (Cast 8 & Cast 15 high on the list), but the queue is deep -- i still have the SEAS 8" here collecting dust, and only recently has the 3" Scan-Speak had any attention. Most energies (and cash flow) are directed at trying to keep up with demand for drivers that we do know work well.
Besides not being impressed with either of the AN i have heard (mind you, both the stamped basket cheapies) there are 2 things that bother me. David Dicks does not use T/S to design his boxes (a dramatic change in the vent parameters for one of the pair i did hear resulted in significant gains) so is not likely invested in their reality, but more so, when i get a pair of drivers that measure like this:
dave
i totally misunderstood this at first glance.
those specs you've shown me there are like the degree of variation that people have reported. given its you who has measured both units then the scope for variables is reduced when compared to someone in a colder climate 1000's of miles away measuring them and then comparing them to your measurements. at which point i understand how 2 values could look different to the factory specs.
given this, and based upon your listening experience. for the money are they worth perusing as a stepping stone towards something else or even stumbling upon the system i desire? other more significant quality drivers that are still below the uber driver (1000 euro per pair) cost are still considerably more per pair.
mark audio alpair 12 - 300euro/pair
fostex 208 sigma - 500/pair
fostex fe206 - 230/pair
tang band w8 - 500/pair
sonido - 300/pair
AN super 10 casts - 220 euros/pair
AN super 8 casts - 200 euros/pair
Just to be clear, those measures are of 2 Super8 sent to me by a client for treatment. The measures were taken under identical circumstances within 5 min of each other. I would not consider them absolute numbers, but can be used to evaluate one driver relative to the other. The relative difference was maintained thru-out treatment.
On my personal listening experience i cannot recommend any of the stamped basket ANs (Super 8, 10 or 12). The cast baskets you have listed come highly recommended(once treated) by Bud. I'd really like to hear them.
The Sonindo i have not heard, they are pretty. I would not personally choose to live with any of the others you have listed (A12 is unobtainium, we'll see if A12.2p can improve enuff to catch up), each brand has a smaller, less expensive driver that i prefer. Any trade-off against ultimate loudness and dynamic impact can be countered with helper woofers (at the same time improving the already superior mid/top). The statement system i am currently working on is Alpair 7.3eN XOed to 4 x SDX7 per side (active)
dave
Dave and gaf, here's my measurement of the AN8 pair I have.
Based on my listening, I think the ANs are a fine set of drivers. They need some work on the treble, as do most full rangers.
I want to reiterate to you gaf that all 'recommendations' by Dave have to be taken with the understanding that he sells the driver he recommends. Please know that.
Based on my listening, I think the ANs are a fine set of drivers. They need some work on the treble, as do most full rangers.
I want to reiterate to you gaf that all 'recommendations' by Dave have to be taken with the understanding that he sells the driver he recommends. Please know that.
Attachments
This is untrue. Mark's concern was with Zaph making measurements of his drivers, not with Zaph posting them in his forum. Let's be clear about that. And Zaph is by no means an amateur. If Zaph's measurements are amateur level, what should we call your measurements?
And i have to state that i'm a diyer 1st when posting here (i fell into selling stuff). If i really like a driver i go about finding a way to add it to my stable of standard drivers. I have significant knowledge & experience to share, and i make no bones about my specific personal preferences.
dave
- Status
- Not open for further replies.