Hey man, very interesting. Is this one built to a different philosophy to my fave JLAudio W7s. Specifically the 13w7AE? That one is now way too expensive
Yes, slightly different. Less excursion, more motor force, and heavier mms. Also low and linear inductance, the JL has no inductance management.
Sounds awesome. Any 10s and 12s in the range? I hope it makes its way to over here
So, for comparison, we are going to use the new Dayton UMII-15 with its parameters found here.
I have created 4 drivers from this, with everything being identical except for changes in motor force and mms.
First we model them in Dayton's recommended 3cf sealed at 900 watts of power. You can see that UM15 light loses a slight amount of low end and gains a slight amount of high end vs. the factory UM15. Motor force / mms ratio increased here. When we compare UM15 light power and UM15 heavy power, we see that they both lose output compared to the first two, because the motor force doubling has increased the motor force / mms ratio on both of them even more. Neither of them meet the output of the factory UM15 and UM15 light until 30hz. UM15 heavy power has slightly more low end and loses high end vs. UM15 light power. This is pretty much in line with my statement of a higher motor force / mms ratio making the sub act more like a woofer.
Now we take all 4 of these versions at the same power and shrink the enclosure volume in half to 1.5cf. Uh oh, they all lose low end. Factory UM15 and UM15 light both lose the most low end extension, naturally. UM15 light power and UM15 heavy power have more low end than the first pair, but UM15 heavy power takes the lead from UM light power below 50hz. Naturally, because of its higher mms.
Now, we get to where what I have explained is important. We are going to compare UM15 light (the theory that light moving mass is what you want) vs. UM15 heavy power (high motor force, heavy moving mass) in 1.5 cubic feet with more power budget. We have 1500 watts to work with now. As you can see, the UM15 light has a lot of high end but falls on its face at 40hz and is 3db down by 30hz compared to the original UM15 in its factory recommended airspace at rms wattage. UM15 heavy power almost exactly matches the low end of the factory UM15 in its factory recommended airspace at rms wattage, and it also has more high end than even UM15 light in the factory recommended airspace at rms wattage.
So, to be able to cut the enclosure size of this sub in half and get the same performance, all we needed is: more motor force, more mms, and more power. Pretty much exactly what I have said, yea?
I have created 4 drivers from this, with everything being identical except for changes in motor force and mms.
- Dayton UM15 is the factory spec
- Dayton UM15 light is the factory spec minus 50g on the moving mass
- Dayton UM15 light power is the minus 50g moving mass plus double the BL^2/Re motor force
- Dayton UM15 heavy power is the double motor force with +50g added to the factory moving mass
Now we take all 4 of these versions at the same power and shrink the enclosure volume in half to 1.5cf. Uh oh, they all lose low end. Factory UM15 and UM15 light both lose the most low end extension, naturally. UM15 light power and UM15 heavy power have more low end than the first pair, but UM15 heavy power takes the lead from UM light power below 50hz. Naturally, because of its higher mms.
Now, we get to where what I have explained is important. We are going to compare UM15 light (the theory that light moving mass is what you want) vs. UM15 heavy power (high motor force, heavy moving mass) in 1.5 cubic feet with more power budget. We have 1500 watts to work with now. As you can see, the UM15 light has a lot of high end but falls on its face at 40hz and is 3db down by 30hz compared to the original UM15 in its factory recommended airspace at rms wattage. UM15 heavy power almost exactly matches the low end of the factory UM15 in its factory recommended airspace at rms wattage, and it also has more high end than even UM15 light in the factory recommended airspace at rms wattage.
So, to be able to cut the enclosure size of this sub in half and get the same performance, all we needed is: more motor force, more mms, and more power. Pretty much exactly what I have said, yea?
@hurrication
//Edit:
Well, this alone, explained by your own words, could tell you, that mass is out of equation for improvement. When mass does nothing of significant help, it does nothing of significant help.
You can of course do that. But you then altered the premise of my claim or parameters of it. Bait and switch so to speak. Moving goalpost.
Glad that you have an iron in the fire and you know something I might not. Well, I did electrical developlment engineering, always snatched sound and light products under myself, and I am still somewhat paid for certain developments in audio. Mostly Waveguides now. But, appealing from the position of authority does not solve anything for us. Well, using 15" in such small enclosure is already ridiculous by conventional standards. It exactly proves that to a point, larger cone has advantage, over small cone, that has the implied advantage of pushing the force to a smaller area, allowing the cone to move further, to overpower enclosure air volume. This of course is not linear phenomena, but 15" is already large enough to prove that this is a thing, and there is this element of large cone area with its underlying advantages over a cone that could push further in the restricted air volume.
I have quite few developments behind me to assure you that even if I am not right, this is not ordinary "nuh uh from the street". There are merits to my claim and it is an educated opinion. But we will see going forwward, who is failing where, I am happy to learn and more than that. Now onto your sims...
//Edit:
Yeah, apparently, seemingly, but no. Let me check and build my case. You are looking just at sensitivity, ignoring the rest, and that´s the issue. I would agree without the more mms. That all is DOH obvious. Not so with the Mms. It is quite branching off, but hope I will manage.
Last edited:
Hey man, great to see the shootout sims, thank you for taking the time to argue at the point with data. It also confirms one of my observations that I have found extremely difficult to successfully draw attention to examine the 'why' in detail by knowing types of folks like yourself. This is exactly what I see in person. A heavy class sound quality driver like the W7 gets seriously loud with the notes from an acoustic kick and 4-string electric bass guitar. Unfortunately, the responses always seem to be offtopic and about the sub 50Hz
So, here are overlayed SPL, cone excursion and power input graphs of the Dayton speaker in 1,5cu.ft volume (usable band 20-120Hz, due to the Hornresp limitations). One is original, one is 50g lighter on Mms.
For the response, the EQd solution works the same. For cone excursion, the sim shows that the lighter version would have little lesser excursion, but I disregard that, as for the same SPL, same excursions should be produce.
For the power graph, the Upper range above 50Hz speaks for the lighter version, 20-50Hz range speaks for the heavier version.
Funnyly enough, for different case, I argued that my solution eats less power, heats up less, and program content is not RMS power, and so this claim of power and heating was disregarded, with some members arguing that it doesn´t matter and I basically lost that argument to the higher power draw being ignored. Now here we have it again, only now I am on the higher power draw side, It´s the only thing You can pick on, and it cannot be lost both ways. Which is it? 🙂
Now, I up the lighter cone model Bl, and EQ to the same response.
Hornresp is limited by amount of eualizers, so now lets focus on 10-60Hz range, where I equalized it out.
See, lighter cone and stronger motor after EQ can do same SPL, but with muchless power. Analogically, If you add the power back in the low end of the speaker, the SPL will go up there too, and so light cone strong motor solution can give you more SPL down there too.
But unequalized high Bl driver will show less SPL in the bass region. It does not make it less subwoofer. It makes it a device with altered impedance, that needs different voltage to be driven properly. It DOES play bass, and power by power, it does play bass more and better.
The thought process that strong motor will make it less of a subwoofer, and light cone will make it less of a subwoofer is not fully true. The strong motor has more damping and needs more voltage for the same power input to keep the voltage response, and the lighter cone only loses around the impedance peak of the heavier cone speaker Fs moved down a bit, not in the whole band. So I cannot agree with your assesment.
I would heavily recommend untangling this and solving thee Mms issue and Bl issue separately. The math should not change, and things should be more obvious.
For the response, the EQd solution works the same. For cone excursion, the sim shows that the lighter version would have little lesser excursion, but I disregard that, as for the same SPL, same excursions should be produce.
For the power graph, the Upper range above 50Hz speaks for the lighter version, 20-50Hz range speaks for the heavier version.
Funnyly enough, for different case, I argued that my solution eats less power, heats up less, and program content is not RMS power, and so this claim of power and heating was disregarded, with some members arguing that it doesn´t matter and I basically lost that argument to the higher power draw being ignored. Now here we have it again, only now I am on the higher power draw side, It´s the only thing You can pick on, and it cannot be lost both ways. Which is it? 🙂
Now, I up the lighter cone model Bl, and EQ to the same response.
Hornresp is limited by amount of eualizers, so now lets focus on 10-60Hz range, where I equalized it out.
See, lighter cone and stronger motor after EQ can do same SPL, but with muchless power. Analogically, If you add the power back in the low end of the speaker, the SPL will go up there too, and so light cone strong motor solution can give you more SPL down there too.
But unequalized high Bl driver will show less SPL in the bass region. It does not make it less subwoofer. It makes it a device with altered impedance, that needs different voltage to be driven properly. It DOES play bass, and power by power, it does play bass more and better.
The thought process that strong motor will make it less of a subwoofer, and light cone will make it less of a subwoofer is not fully true. The strong motor has more damping and needs more voltage for the same power input to keep the voltage response, and the lighter cone only loses around the impedance peak of the heavier cone speaker Fs moved down a bit, not in the whole band. So I cannot agree with your assesment.
I would heavily recommend untangling this and solving thee Mms issue and Bl issue separately. The math should not change, and things should be more obvious.
Last edited:
Congrats, you have just done exactly what the makers of cheap subwoofers do!
hmm.. lemme just add this band-aid EQ right here to try and get some low end back
Now, keep that same sim with those two and include a model with the same high bl plus 50g more mms than original, at the higher power level. Also add a factory original sub in a 3cf sealed box. I assume you left those out because it would invalidate your claim that more mms is bad. Show us the Fb/qtc of your pair as well as how much group delay that EQ added. Both of those are important and can not be overlooked. What you have presented is just a narrow scope of half the variables so you can try and get a "see, i told you so" in. All you have shown is how to apply a band-aid to get a higher bl driver to have the same response as a factory original. Not relevant to total objective performance, zoom out.
hmm.. lemme just add this band-aid EQ right here to try and get some low end back
Now, keep that same sim with those two and include a model with the same high bl plus 50g more mms than original, at the higher power level. Also add a factory original sub in a 3cf sealed box. I assume you left those out because it would invalidate your claim that more mms is bad. Show us the Fb/qtc of your pair as well as how much group delay that EQ added. Both of those are important and can not be overlooked. What you have presented is just a narrow scope of half the variables so you can try and get a "see, i told you so" in. All you have shown is how to apply a band-aid to get a higher bl driver to have the same response as a factory original. Not relevant to total objective performance, zoom out.
I have already untangled mms and bl in my sim showcase, there are no mental gymnastics that can unpack it any other way than what I have already presented.
So "all we need" is twice the magnet weight, a new voice coil design that can handle around double the power as the previous design, and a heavier cone that can withstand the increased forces put on it.
Generally available at around twice the cost, more if linear response in the smaller cabinet is also desired 😉
Not exactly twice the cost, you would be shocked at the markup that some speaker companies put on their subwoofers. But yes, generally you design the voice coil size from the start to support the power handling needed to meet the design goals. The cost difference between a 2.5" coil and a 3" coil from PE in bulk is no more than $10 per unit.
I have some 3" coil motor designs with over 200 n^2/W motor force using this $50 retail magnet, its cost in bulk from JDA magnetics is around $22 each. Also have some 4" coil motor designs with around 150 n^2/W motor force with this $50 retail magnet, but its bulk cost is around $35 each.
I have some 3" coil motor designs with over 200 n^2/W motor force using this $50 retail magnet, its cost in bulk from JDA magnetics is around $22 each. Also have some 4" coil motor designs with around 150 n^2/W motor force with this $50 retail magnet, but its bulk cost is around $35 each.
Last edited:
@hurrication
Expensive is not a measure of performance or quality in audio fields too often. This is not a scientific argument.
Again, the most powerful systems like IPAL certainly use very high Bl drivers that do not add mass on their cones intentionally.
If we cannot discuss why Higher Bl driver doesn´t produce less bass, or why more mass is a net loss, but we need combo of X circumstances to get to the seemingly better performance, then we don´t have an argument. This is a mockery of science, math and logic, and I show you how. All out of the window.
You claimed something along the lines of "More Bl makes the driver to play less bass, making it effectively a woofer from the subwoofer", then you proceed to make a case with a driver with double motor force to make more bass. Can't make this satire up. This is ridiculous and you are not seeing any of that. It would even be funny if you weren't serious.
Congratulations, you managed to twist it enough to show what you wanted to see. What´s not wanted to see is just dismissed. No wonder snakeoil lives well and good in the hifi world still. Neat!
Don´t care what they do, I do my developments. If the solution is the same, it´s not a problem. Many people are doing things not knowing why.
Expensive is not a measure of performance or quality in audio fields too often. This is not a scientific argument.
I see your ways the same way, I just don´t need to be nasty about it. Solving the issue on the signal part is not just cheap, but a logical step epensive manufacturers use.
It would invalidate nothing, it makes it more difficult to unpack the underlying problematics. I did it to not branch off and make it confusing and unseparable from the true issue. Your approach is hostile and unscientific. If you can´t untangle and argue a behavior of an isolated parameter, you´re not doing a good scientific job.
If you knew anything about audio, you would have known that Group Delay would be same or better.
Again, the most powerful systems like IPAL certainly use very high Bl drivers that do not add mass on their cones intentionally.
If we cannot discuss why Higher Bl driver doesn´t produce less bass, or why more mass is a net loss, but we need combo of X circumstances to get to the seemingly better performance, then we don´t have an argument. This is a mockery of science, math and logic, and I show you how. All out of the window.
You claimed something along the lines of "More Bl makes the driver to play less bass, making it effectively a woofer from the subwoofer", then you proceed to make a case with a driver with double motor force to make more bass. Can't make this satire up. This is ridiculous and you are not seeing any of that. It would even be funny if you weren't serious.
Congratulations, you managed to twist it enough to show what you wanted to see. What´s not wanted to see is just dismissed. No wonder snakeoil lives well and good in the hifi world still. Neat!
Last edited:
Yay, going the “twice” on one chassis makes less barbaric waste of isobaric as it can allow for excursion to be built higher. Awesome way to evolve or negate the need for doing the shellfish 🔑
If it would be so easy to put just the coil of double length into the system and have double the power. That pesky thermodynamics and mechanics too, is somehow ruining that idea.
I read “double motor force” and “increasing weight”
Any reason for omitting the weight condition he mentioned? From what I can read from here is you are omitting his second requirement to make his statement fit with a false match
Is an Fb of 39Hz truly in "subwoofer" territory? It seems quite high. The expected natural response is shown below, and the −3dB point is a very high 45.6Hz.
Of course, if we add an 80Hz 4th-order Linkwitz–Riley low-pass filter, we get the following response. The action of the low-pass filter now extends the −3dB point to 31.3Hz, albeit with a loss of 3.5dB in sensitivity. Here the input power requirements have gone up by 220%.
If we instead apply a 50Hz 4th-order Linkwitz–Riley low-pass filter, we get the following response. The −3dB point is now 24.8 Hz, albeit with a loss of 6.8dB in sensitivity. Here the input power requirements have gone up by about 470%.
And we still don't have a −3dB point of 20Hz for this subwoofer. I guess that's the penalty that needs to be paid when cramming a large woofer into a small enclosure.
Attachments
Last edited:
@Randy Bassinga
The specific reason is to show what Bl does, what mass does, and to untangle the problem and get to the bottom of things.
The approach of multiple variables allowed for a logical fallacy here. Big one.
If more motor force causes less bass, and there is no math to support any kind of nonlinearity of such statement, then ad absurdum with no motor you'd have most bass.
If weight adds on bass response, then put a truck of bricks on the cone, it will play louder.
And yes, we need to go there obviously, because going just a bit doesn't light any bulbs yet.
If we cannot find a math that gives us reasons for different optimum of my assumption, then this assumption holds! Ridiculous? Well obviously not. Or is it? Which is it?
The specific reason is to show what Bl does, what mass does, and to untangle the problem and get to the bottom of things.
The approach of multiple variables allowed for a logical fallacy here. Big one.
If more motor force causes less bass, and there is no math to support any kind of nonlinearity of such statement, then ad absurdum with no motor you'd have most bass.
If weight adds on bass response, then put a truck of bricks on the cone, it will play louder.
And yes, we need to go there obviously, because going just a bit doesn't light any bulbs yet.
If we cannot find a math that gives us reasons for different optimum of my assumption, then this assumption holds! Ridiculous? Well obviously not. Or is it? Which is it?
No, I said verbatim :
" For subsonic subwoofers that don't need to play more than an octave, this ratio can be between .2 to .3. For audiophile subwoofers that can play 2-3 octaves, this ratio needs to be .3 to .4. And for a very wide bandwidth, .45+. Once you get around .45+ the subwoofer wants to start being more of a woofer and you begin to compromise low end extension. "
I have explained this as motor force / mms ratio multiple times in this topic, but the problem is you just ignore some of the variables and re-write the discussion in your head to support your outdated thoughts about subwoofer design.
My friend, you are the one who is twisting things to try and validate your thought process. Unfortunately, your method of design is irrelevant and outdated and companies are hiring people like me to re-engineer the cheap band-aid subwoofer solutions as something that is actually workable. When you have a small sealed subwoofer design that is going into production, I will begin to listen to what you have to say as being valid.
It is in the same playing field as many other hifi sealed subwoofer designs of similar usage, with the others needing more airspace for a similar alignment. Given the constraints of 1.25cf sealed volume, it is very good. There is nothing out there that I am aware of with this low end curve in that small of an airspace.
Correct, This subwoofer, and any small sealed subwoofer at that, is not designed for this type of low end response. Results like that can easily be achieved with a low tuned vented design, but with that you fall way outside the size and airspace restraints.
Same thing for the case. This claim means more motor force = less bass. Still ridiculous claim.
There is a misunderstanding here. I can see you are not listening. The response is not for you much.
Again being successful in production and marketing has very little to do with factuality, truth, math and science and such. Look at how much profit McDonalds does, and they don't exactly make best food, do they. Especially if you are interested financially in the product, you have a good reasons to be biased about it. The money listens and talks.