Hmm, seems I'm returning to Cms questions again,
I was reading: Active Compensation of Transducer Nonlinearities And came across Slide 32 on page 12.
I got the link from:Laser Instead of an Accelerometer? Is it possible?
XBL² technology seems to give a very flat BL curve, and excellent inductance curves. (See Tumult BL & Inductance Curves ). LMS also, at least via words, seems to indicate that the techonology could also give excellent BL curves (See TC Sounds LMS Technology )
I'm just wondering, is there a box configuration, or some tweak, that could 'linearize' Cms value for a subwoofer? If the first and third issue, according to Kippel, can be overcome with motor design, what can be done about Cms?
Oborous said:
I'm just wondering, is there a box configuration, or some tweak, that could 'linearize' Cms value for a subwoofer? If the first and third issue, according to Kippel, can be overcome with motor design, what can be done about Cms?
Nice paper link - thanks!
http://72.14.209.104/search?q=cache...+magnetic+suspension&hl=en&gl=us&ct=clnk&cd=1
A box configuration, no not possible. Although a woofer in a small sealed box can utilize the pressure inside to override the effects of an imperfect suspension. By this I mean the distortion generated by the air pressure pushing on the cone area has way more effect than cms non-linearities. Of course this is just masking the problem with another.
"Some tweak" yeah, if you do alot of testing you can create algorithms to counteract nonlinearities in both BL and cms, this would be done by ramping up or down the signal going to the woofer's amp in a manner that would cancel any nonlinearites. Extremely difficult and I would bet it would be fairly expensive, this is why those who want accurate bass use low excursion woofers and lots of them to get spl and extension.
High excursion introduces alot of problems in the design stage overcoming them usually leads to an expensive woofer. So it can be cheaper to buy many lesser woofers, think one 18" tumult ($700) vs. 8 dayton 15"(~$700). Which would you rather have?
nunayafb said:
Actually the air spring is much more linear than the suspension and the nonlinearity of air can be neglected if the woofer Vd = Sd*Xmax is less than 10% of the box volume. The linearizing capability of the air volume stiffness, if greater than the suspension stiffness, is the origin of the term "acoustic suspension". So the text in the first paragraph above starting with "By this..." is incorrect.
"Algorithms to counteract nonlinearities" is kinda vague.... All you need to do is apply negative feedback 😉
IF you have a really linear suspension, any BL assymetry or nonlinearity will cause dynamic offset problems - which are much more severe than Cms -caused distortion. Dynamic offset is also known as oil-canning.
---Actually the air spring is much more linear than the suspension and the nonlinearity of air can be neglected if the woofer Vd = Sd*Xmax is less than 10% of the box volume.---
Linkwitz gives a formula for calculating the H2 distorsion generated by the air non-linearity in a closed box.
A woofer has second and third order non-linearities. I am not aware of a third order non-linearity for the air. If the internal volume of the box, Vab, is smaller than the Vas of the speaker mounted in it, the compliance of the air will dominate in the response of the system. So there should be less third harmonic when the unit is enclosed than when used on a pannel.
Linkwitz gives a formula for calculating the H2 distorsion generated by the air non-linearity in a closed box.
A woofer has second and third order non-linearities. I am not aware of a third order non-linearity for the air. If the internal volume of the box, Vab, is smaller than the Vas of the speaker mounted in it, the compliance of the air will dominate in the response of the system. So there should be less third harmonic when the unit is enclosed than when used on a pannel.
http://www.linkwitzlab.com/thor-design.htm
http://www.musicanddesign.com/driver_box.html
Note that you can alter driver/system compliance with the "box" (or driver depending on how you look at it) - even one with a relativly small volume.
Consider a box within a box where the interior box is a very small chamber (cavity-coupled to the driver) and the walls are lossy (..perhaps made of fiberglass paneling with some small perforations). I wouldn't think that it would make driver compliance anymore linear though (..actually I think it might do the opposite).
http://www.musicanddesign.com/driver_box.html
Note that you can alter driver/system compliance with the "box" (or driver depending on how you look at it) - even one with a relativly small volume.
Consider a box within a box where the interior box is a very small chamber (cavity-coupled to the driver) and the walls are lossy (..perhaps made of fiberglass paneling with some small perforations). I wouldn't think that it would make driver compliance anymore linear though (..actually I think it might do the opposite).
Actually I was referring to the much higher restoring force of the acoustic suspension dominating the behavior of the total suspension. If you have two springs in parallel one being the suspension and the other being the air the one with the higher spring constant will dictate the behavior of the speaker the most, apparently worded poorly the first time.
Leave it to an M.E. describe negative feedback in the stupid way that I did.
nunayafb said:
Hiya,
Excellent posts everyone, I'm learning alot...
I've come to the same conclusion, that for subwoofers, you're probably better off with even numbers (and multiples) of subwoofers, and probably smaller subwoofers to generate a given low frequency SPL. But, if I can't afford the real estate of that many subs, what are my options.... so I'm looking into what I can do to get as good of sound quality as possible at a reasonable SPL, in a small package without comprimizing anything except for SPL.
I am wondering about a box for a high-excursion subwoofer, operating it only within the BL-linear region i.e. utilize a Adire Tumult ~24mm X-max rather than their listed X-max; which I'm thinking I can easily do by limiting the available power.... basically utilize a 1,000 watt amp instead of the ~~1,600-1,800 watts that S. Linkwitz's handy spreadsheets indicate are neccesary to utilize a full 33mm of X-max (Yes, this still is crazy amount of power, considering I'm using 25 watts of chip amp power, or looking at 8 watts of SET power).
I'm trying to figure out how to have a linear Cms value, would limiting the Xmax also limit the non-linearities due to extreme excursion on the suspension?? I'm assuming maximum non-linearity is present at maximum excurision.
Any suggestions on how to measure Cms non-linearities for a non-engineer? I do have engineer friends, so might be able to utilize some more professional equipment.
The special amplifier of ACE-bass (Audio Pro) has the effect of linearising Cms, or rather having an artificial C dominating over Cms, such that the total C becomes linearised.
Ron E said:
Could you provide reference for that? I think by reading the whole article of where that came from, I really could learn alot. (Currently trying to purchase more books, as reading is alot easier to do than get shop time to build proto-types... so if you have suggestions for books, please pass them along).
nunayafb said:
I'm interpreting this as: In a smaller box, the internal air is a stiffer spring, and so would dominate the 'parallel spring' analogy. Where as when the air is not that 'springy', suspension would dominate (which could occur in very large boxes, or ported boxes).
Which, would seem to be supported that Fb is much larger than F3 for smaller boxes, than larger boxes... as the air suspension is much stiffer, which makes it more difficult to resonate at lower frequencies...
Which... also seems to indicate, that in small boxes, the thermal expansion of the air could really cause problems if I'm dissapating alot of energy into heat.... hmm, non-linear response over a long period of time due to heating effects... something to think about.
Thermal expansion of air is only a problem if the box is 100% sealed. Any porosity or microscopic holes will allow the pressure to reduce to the current atmospheric pressure. But yes, small sealed boxes can have cooling problems with high powered coils.
Be careful with the spring analogy, because it is just that an analogy. The restoring force(springiness) is due to the increased pressure on the inside of the box due to the reduction in box volume, acting on the cones surface area.
As forr mentioned this is actually very linear in the range which we are dealing with, and can be modeled by press1*vol1=press2*vol2.
press1= atmospheric, vol1=box volume, vol2= (vol1-Sd*excursion)
and press2=press1*vol1/vol2,
gotta run, hope this is complete enough...
Cms is a function of the surround and the spider??
Correct me if I'm wrong... but Cms is only a function of the Surround and the Spider?? I also realize that the surround is an air sealing mechanism.
Could someone build a sub with those linear force springs (looks like a tape measure) as the spider to make the forces trying to return the cone to zero position more linear?
(Yes, I do realize that there are some engineering difficulties, as you're having to deal with +/- excursion, along with having the springs only excerting force to return the cone to zero)
Correct me if I'm wrong... but Cms is only a function of the Surround and the Spider?? I also realize that the surround is an air sealing mechanism.
Could someone build a sub with those linear force springs (looks like a tape measure) as the spider to make the forces trying to return the cone to zero position more linear?
(Yes, I do realize that there are some engineering difficulties, as you're having to deal with +/- excursion, along with having the springs only excerting force to return the cone to zero)
From
http://www.musicanddesign.com/driver_box.html
---Turning off the suspension nonlinearity and only considering the air compliance nonlinearity shows that the air nonlinearity can generate both even and odd order HD.---
Can someone explain this generation of odd order HD ?
http://www.musicanddesign.com/driver_box.html
---Turning off the suspension nonlinearity and only considering the air compliance nonlinearity shows that the air nonlinearity can generate both even and odd order HD.---
Can someone explain this generation of odd order HD ?
You want the nonlinearity of Cms to balance the nonlinearity of the Bl system, otherwise you get oilcanning. The spider also serves to center the voice coil. You could try using a slide of some sort and weak springs, but I can almost guarantee it would rattle...
forr said:
I wonder if this is because the cone is not equidistant to every point of wall in a sealed box. I wonder if a spherical subwoofer (sealed) would have air non-linearity.
But yes, I wonder about Air nonlinearity and the generation of odd order HD. Do you have any thoughts as a basis.... I have only the one possible thought above.
Thx
Ron E said:
454Casull said:
That's exactly what I'm thinking 454Casull. Wondering about non-linearity for those drivers with the extremely flat BL curve, such as the published Adire Tumult curves.
Personally, I think the Tumult would be excellent for some active compensation (servo accelerometer for example), because of the BL curve. A paper I was reading, sorry no reference at the moment, seems to indicate thermal non-linearities and BL non-linearities dominate Servo-subwoofer compensation failures. With some good thermal management (still thinking of that), and a flat BL curve, I was wondering what other things would affect sound quality... and came up with Cms
Now wondering about applying servo compensation to something that sort of indicates that it would make it more prone to oscillation.
- Status
- Not open for further replies.