No offense but it's clear you're not aware of how driver parameters are inter-related.
Hoffman's iron law- you can have 2 out of three: Deep bass, Efficiency, Small Box.
Car audio drivers are easy to design to optimize the first and third.
no offence taken, i know my way round transducer design quite well...
in general you have to make space/eff/bandwidth compromises. but there are ways to improve over the current performance level available in the market,
and you are indeed making my point for me, many woofers (especially aftermarket Automotive) have low Fs and Low Vas but are woefully inefficient,
in general you have to make space/eff/bandwidth compromises. but there are ways to improve over the current performance level available in the market,
and you are indeed making my point for me, many woofers (especially aftermarket Automotive) have low Fs and Low Vas but are woefully inefficient,
This driver also falls into that category as well. Look at the sensitivity rating. The N0 is <0.19%.
I'm not saying it isn't a good driver. It's on my short list of drivers I'd like to test. The areas where it looks like it should excel are linear displacement, inductance, deep bass distortion and power handling. Efficiency and output headroom outside of the deep bass range are not its strong suit though.
You may be surprised what a large increase in the motor force can do to a drivers real world performance.
you're forgetting that N0 is calculated based on mid band efficiency, and is a product of Fs, Vas and Qe, it has no relation to actual in use efficiency of a subwoofer.
Ideally there needs to be a new standard to define a single figure efficiency of subwoofers,
most people struggle to grasp multiple parameters to describe subwoofer performance.
this driver is a high end subwoofer, the performance above ~100hz is irrelevant, as is the mid band efficiency (N0 or dB/1w etc.)
I understand totally what a large increase in motor force can do for real world performance,
if a stronger motor would benefit the driver, it would have one..
as it is, the only way a stronger motor would be of benefit would be if the cone area was increased (18" driver), or the Moving mass increase to tune better in an even smaller box,
but a 15" driver that works in 60L is small enough, any smaller and you'd struggle to physically fit the driver in the cabinet!
Ideally there needs to be a new standard to define a single figure efficiency of subwoofers,
most people struggle to grasp multiple parameters to describe subwoofer performance.
this driver is a high end subwoofer, the performance above ~100hz is irrelevant, as is the mid band efficiency (N0 or dB/1w etc.)
I understand totally what a large increase in motor force can do for real world performance,
if a stronger motor would benefit the driver, it would have one..
as it is, the only way a stronger motor would be of benefit would be if the cone area was increased (18" driver), or the Moving mass increase to tune better in an even smaller box,
but a 15" driver that works in 60L is small enough, any smaller and you'd struggle to physically fit the driver in the cabinet!
I am not forgetting. That may be but reference efficiency still has some correlation with the overall efficiency in the bass range as well.
What IS relevant is the performance from 40-120Hz which is the lions share of most of the bass content in media. This is the range being discussed on my end. This driver is neither efficient or sensitive over this range. That is my point. It's overall a lower efficiency driver.
We will have to disagree here on a number of points. My opinions on transducer & speaker design and optimization are quite different from the long held traditional views based on F3, constant voltage response shape and small signals.
Out of curiosity are you affiliated with Tymphany? Do you work in the industry?
I do indeed work in the transducer design industry,
it seems you're confident on the performance of this driver having never tested it or simulated in cabinet,
during real life testing we found that this driver beat competing drivers by 12dB at 30Hz, for the same power,
SPL in cabinet, at subwoofer frequencies are significantly higher than similar sized subwoofers (or larger), that is higher than usual efficiency in my book.
it seems you're confident on the performance of this driver having never tested it or simulated in cabinet,
during real life testing we found that this driver beat competing drivers by 12dB at 30Hz, for the same power,
SPL in cabinet, at subwoofer frequencies are significantly higher than similar sized subwoofers (or larger), that is higher than usual efficiency in my book.
I thought you might.
Why do you make this assumption? Why would I not have run some sims with this driver? On top of the factory information there is 3rd party data on systems using this driver and other data that is not public.
Did they have comparable SD and Xmax? A 12dB efficiency difference is big. The most obvious explanation for such a difference is that one system was near it's impedance maximum at 30Hz and the other had a much different impedance curve with the peak well away from 30Hz placing the efficiency maximum in a different range. This is why I never try to look at efficiency at a single frequency. It always favors one system over another.
The STW-350 has a maximum efficiency of around 1.3% in a very small sealed enclosure. In a larger sealed cab for maximum 30Hz efficiency it should be <1%. 12dB less than 1% is about 0.06%.
Let me reiterate I'm in no way saying this is not a good driver. I'm not bashing on it. All I said is I would be more interested in it with a stronger NdFeb based motor.
it's Neither,
remember that Xmax is not the limit of linear excursion, it is unrelated to the ability or performance of a driver at that excursion,
it's merely the length that the VC winding width overhangs the to plate of the driver, i wish we could replace this parameter!
be very aware many manufacturers 'interpret' the measurement of xmax in different ways,
for some reason the pro manufacturers use VC overhang + 1/2 or 1/3 of the gap height, there is no justifiable reason for that!
some other companies use Peak to peak values for the bigger numbers,
some use arbitrary calculation methods, i've even seen maximum peak to peak mechanical excursion limit published as Xmax!
you can only get an idea of the performance by considering the Xmax (VC overhang) and gap height,
a) imagine a VC 20mm long and a 1mm gap height, the maximum excursion must be limited to 9.5mm, and published Xmax is 9.5mm
b) for the same 20mm VC and a 18mm gap height the maximum excursion is 18mm, yes the published Xmax is only 1mm
the subject is much more complex than this basic calculation, but typically for a subwoofer a long gap is very advantageous even though it reduces the published xmax.
the driver being discussed has a very long gap, hence the low published Xmax,
it will however have very low distortion and very large excursion capability,
a good rule of thumb is to look at the VC overhang, and also look at the displacement where 1mm of VC ramains int he gap,
using the example above this gives,
a) 9.5 and 9.5
b) 1.0 and 18.0
just fyi, i'd choose driver (a) as a midrange (less than 5mm excursion)
and i'd choose driver (b) as a woofer or subwoofer (as it would have higher excursion)
Are any vendors selling these drivers yet?
Gbullimore's post in reply covers it well.
This driver might "only" have a 10.5mm coil overhang but the magnetic gap is very tall. What this means in general is that while the BL may start to drop as the coil moves beyond the coil overhang, it should drop very gradually. The BL variation for subwoofers is typically considered acceptable at 70% per Klippel analysis. With such a tall magnetic gap a 30% loss of BL should not occur until far beyond 10.5mm.
Technically the coil overhang is 10.5mm. As a measure of functional xmax as most other drivers are rated though, this is way conservative. Most other bass drivers are not listing the mathematical coil overhang as the xmax these days.
Gbullimore's post in reply covers it well.
This driver might "only" have a 10.5mm coil overhang but the magnetic gap is very tall. What this means in general is that while the BL may start to drop as the coil moves beyond the coil overhang, it should drop very gradually. The BL variation for subwoofers is typically considered acceptable at 70% per Klippel analysis. With such a tall magnetic gap a 30% loss of BL should not occur until far beyond 10.5mm.
Technically the coil overhang is 10.5mm. As a measure of functional xmax as most other drivers are rated though, this is way conservative. Most other bass drivers are not listing the mathematical coil overhang as the xmax these days.
That's the old (traditional) way to define xmax.
There are new ways, like klippel analysis, this will give a much more accurate picture of usable excursion determined by actual distortion measurements.
Peerless chose to use the old method for the 10.5 mm xmax rating but with drivers like this it's common sense to use klippel testing instead. It will give more accurate results AND a higher number that looks better on a spec sheet.
You don't need to replace the parameter, just use a definition that more accurately defines the performance vs distortion.
While it is true that many companies use different formulas or even newer distortion analysis type of definitions for xmax, those in the know look at spec sheets and note how the xmax was determined. Without that data nothing is even remotely comparable except maybe different drivers produced by the same manufacturer.
Maybe one day there will be a new standard that everyone agrees on, until then drivers like this should be klippel tested and the spec sheet should clearly state how the xmax was determined.
the issue with specifying with the Klippel suggested method is that it's not possible to 'prove' this value for a design, only for a specific unit.
VC overhang is a physical parameter that does not vary,
it's possible to get variable results with the klippel method, unit to unit variation, and also pre and post run-in, will affect the measured value.
remember that most audio companies require specs before run in, so they can test the driver at IQC, whereas most hobyists prefer post run-in parameters, (audio co. engineers run in then test TS by themselves)
for industry engineers, most will understand the Gh and Xmax listed in geometrical terms, although there are some that have very little grasp of driver technology!
personally i would reference 'traditional' Xmax and the 1mm VC in gap as the 2 important displacement characteristics,
VC overhang is a physical parameter that does not vary,
it's possible to get variable results with the klippel method, unit to unit variation, and also pre and post run-in, will affect the measured value.
remember that most audio companies require specs before run in, so they can test the driver at IQC, whereas most hobyists prefer post run-in parameters, (audio co. engineers run in then test TS by themselves)
for industry engineers, most will understand the Gh and Xmax listed in geometrical terms, although there are some that have very little grasp of driver technology!
personally i would reference 'traditional' Xmax and the 1mm VC in gap as the 2 important displacement characteristics,
These are not really problems.
To get rid of unit to unit variation test 100 drivers and average the results. This can even be updated by testing 100 drivers from every batch made and updating the spec sheet. Too much work? Boo hoo, I want accurate specs.
Run a pre break in klippel test for quality control purposes and a post break in test for the spec sheet. Too much work? If a company thinks so they are lazy and not willing to do what is required to get an accurate spec.
Audio company engineers and diy'ers do indeed test the t/s themselves sometimes but this has nothing to do with anything. Nobody can measure xmax by themselves unless they rip apart the driver and measure the physical characteristics or run tests (like klippel) to determine linear excursion vs distortion.
Understanding the traditional xmax definition is all fine and good, but like you yourself say - it's not an accurate indication of linear xmax, especially for drivers like this one. So what good is it?
The only real way to get an accurate linear excursion spec is to test BY MEASURING the excursion vs distortion.
To get rid of unit to unit variation test 100 drivers and average the results. This can even be updated by testing 100 drivers from every batch made and updating the spec sheet. Too much work? Boo hoo, I want accurate specs.
Run a pre break in klippel test for quality control purposes and a post break in test for the spec sheet. Too much work? If a company thinks so they are lazy and not willing to do what is required to get an accurate spec.
Audio company engineers and diy'ers do indeed test the t/s themselves sometimes but this has nothing to do with anything. Nobody can measure xmax by themselves unless they rip apart the driver and measure the physical characteristics or run tests (like klippel) to determine linear excursion vs distortion.
Understanding the traditional xmax definition is all fine and good, but like you yourself say - it's not an accurate indication of linear xmax, especially for drivers like this one. So what good is it?
The only real way to get an accurate linear excursion spec is to test BY MEASURING the excursion vs distortion.
I would also like to mention that the Klippel parameters test is the one that engineers should be using to DESIGN drivers. The Klippel performance test is the one that should be used to define xmax.
Unfortunately I've never even seen a Klippel performance test, all you ever see is the parameters test. The performance test is actually a lot easier to do (there's nothing to interpret) and it gives a more accurate representation of actual linear xmax because it directly measures total distortion, not single parameter distortion.
Unfortunately I've never even seen a Klippel performance test, all you ever see is the parameters test. The performance test is actually a lot easier to do (there's nothing to interpret) and it gives a more accurate representation of actual linear xmax because it directly measures total distortion, not single parameter distortion.
it's not about laziness, we test every driver 100% on line and keep the records for ever,
as this thread shows many people are confused by such things as Xmax, and sensitivity spec. so having a pre and post run-in spec sheet would cause a lot of confusion.
as for updating the spec sheet for every 100 drivers produced... would you want to buy driver where the spec sheet constantly changes, even by a small amount?
the issue measuring Xmax using distortion as the limit, is that this will vary depending on test enclosure,
e.g, with Klippel Xmax testing, you can get a very different value depending if you test the driver in free air or a small sealed enclosure, large sealed enclosure or vented enclosure,
at the moment the most widely used standard in the industry is the geometric Xmax (coil overhang)
there is no issue with using this, as as long as you provide gap height, all the information is there to be able to accurately compare drivers performance,
as this thread shows many people are confused by such things as Xmax, and sensitivity spec. so having a pre and post run-in spec sheet would cause a lot of confusion.
as for updating the spec sheet for every 100 drivers produced... would you want to buy driver where the spec sheet constantly changes, even by a small amount?
the issue measuring Xmax using distortion as the limit, is that this will vary depending on test enclosure,
e.g, with Klippel Xmax testing, you can get a very different value depending if you test the driver in free air or a small sealed enclosure, large sealed enclosure or vented enclosure,
at the moment the most widely used standard in the industry is the geometric Xmax (coil overhang)
there is no issue with using this, as as long as you provide gap height, all the information is there to be able to accurately compare drivers performance,
which Klippel test do you refer to as 'performance test'
we use the full suite of Klippel testing to design our drivers, and also advanced FEA tools like Comsol,
Klippel LSI testing is very useful but as you say results are rarely published, this is because they are easy to mis-interpret,
you can also choose the input parameters to either show the driver to be better or worse than it really is,
Nice. You said you are a design engineer, are you actually working for Peerless?
No. The pre break in test would never be published anywhere. It would just be used for quality control purposes. The post break in spec would be the one published on the spec sheet.
That's not at all what I said. Test 100 samples from each BATCH and then update the spec sheet for each BATCH.
I would LOVE to buy a driver where the spec sheet continually changes to reflect updated parameters. Just so long as the batch number is clearly labeled on the driver itself, you could look up the spec sheet for the batch number of the driver you have and you would have very accurate specs.
I realize this is a lot further than any company would be willing to go, but it would certainly be ideal for people that can't measure their own t/s.
Yes, obviously an enclosure will change things dramatically. But you can just use the klippel standard test baffle since it's already a known reference. If you want to test in multiple different types of enclosures that would be ideal, but again, not something that any manufacturer is going to do.
Yes, the traditional method is still most widely used. But it DOES NOT give an accurate picture of linear excursion. If it did you would not be arguing that this driver has more than 10.5 mm of usable excursion.
And like you said, most people just look at the xmax spec itself, a single number. Most people don't even know that companies are playing around with the traditional formulas to add an extra 1/3 or 1/2 to the xmax rating.
There is ALL KINDS of cheating going on with xmax specs, even from respected companies like B&C. Ideally there should be a standard but in lieu of that, companies can at least define linear excursion as accurately as possible. Coil and gap measurements WILL NOT get you there, especially for this type of driver. You get 10.5 mm, not a number a lot higher that would accurately reflect linear excursion.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.