Car Audio Classifieds! banner

1 - 11 of 11 Posts

·
Registered
Joined
·
60 Posts
Discussion Starter #1
Just a quick note: If you've followed my (short) post history at all, you'll know me as the guy that comes straight out of left field with a crazy design idea about once every two months or so. This post is no exception. Be prepared.

Also, and this might be hard to believe, but I have every intention of building the thing I'm about to describe, even though I, admittedly, don't have much experience building subs or any idea of how I'm going to build half of it. Consider this a rough draft of the design.

Anyway, on to the idea behind it:

There are two main voice coil configurations used in the design of subwoofers: overhung, and (less commonly) underhung. Both work well for musical reproduction, but neither are optimized for pure SPL applications where it isn't necessary to reproduce sound accurately or without distortion. Yet they are being used exclusively in pure SPL applications; despite all the technological progress that has been made, companies continue to come out with these 25kW, 100 pound, SPL monsters that use voice coil configurations designed for accurate musical reproduction.

What's wrong with using overhung and underhung designs when you don't care about distortion, you might ask? Well, each design is incredibly inefficient in it's own way.

Let's say you had a sub with an overhung VC design, with a wind height of 2.5" and a top plate thickness of .5". That means that, at any one time, only one fifth of the VC is in the gap, which means you are only utilizing one fifth of the power flowing through your voice coil. The other 4/5 of the power is being transformed directly into unusable heat. This also means that the resistance of the coil is 5 times higher than the part you are actually using.

Underhung designs aren't any better. Let's say you had an underhung VC with a wind height of .2" and a top plate height of 1". At any one time, the VC is only occupying 1/5 of the gap, which means you are only utilizing 1/5 of the motor's total force.

My whole point is, if you are going to throw out the requirement that voice coil motion be linear, why only use linear designs? Because now we're just talking about any design that makes something move forward and backward around 60 times a second as fast as possible.

My idea? Well, there's two parts. The first is: What if we merged the two designs? To do that, you would need a coil with a wind height exactly as tall as the top plate thickness. Now, normally, with just one of these coils, you wouldn't get much travel. None of it would be linear, and it would be hard to predict how it would move in response to a signal. But, and this is the second part of my idea, what if we had multiple coils like that? Not in parallel or in series, but individual coils that are activated one by one, as the voice coil moves through the gap.

Imagine, if you will, a voice coil with a really tall former height. At the bottom of this former is not one, but 8 .5" tall coils, all as close together as possible. The second coil from the bottom is wound over the bottom one's leads, the third wound over the second's as well as the bottom's, etc. Above these coils are the spiders, in their usual spot. The top spider has a single set of leads, to which one of the leads from each coil is connected, and to which one of the outputs from the amp is connected. Above them is a 4" section of former coated with a thin insulating layer of plastic. Glued on top of this layer of plastic are 8 .49" bands of copper, the bottommost band soldered to the other lead from the bottommost coil, the second band from the bottom soldered to the other lead from the second coil from the bottom, etc.

Touching these bands is a straight, thin, copper bar that has all the properties of a light spring. This is the brush, the opposite side of which is connected to the other amp output. I've already figured out what the optimal width of the brush should be: it's (the thickness of the top plate)*(sqrt(2) - 1).

Paired with a motor with a top plate thickness of .5" and quad 1" slugs, this would produce an SPL-only sub with a one-way xmax of 1.75" (and i'm using the old definition here. none of that 70% BL bs) and a one-way xmech of 2".

Assuming each coil is 1 ohm, and the amp being used to power it is putting out 5000 watts, when the sub is at max efficiency (when one of the coils is in the exact center of the gap), the sub would be putting out as much power as an 8 ohm subwoofer with the same wire gauge, motor, and wind height would be putting out at 40,000 watts. When the sub is at minimum efficiency (when a coil is as far out of the gap as possible without the next coil being activated) the sub would be putting out as much power as that same 8 ohm subwoofer would be putting out at 20,000 watts.

If any of you are interested in what that same figure is, averaged with respect to time, the formula is: 5000*8*average(abs(sin(theta)/2)+1/2) from 0 to pi, which, according to WolframAlpha, is 32,732 watts. So, this design is over 6 and a half times as efficient as an equivalent overhung design.

Obviously, the big advantage to this design is greater efficiency. And greater efficiency means that less energy is being wasted as heat, which means you can put more power into it without frying the voice coil. So not only are you getting more bang for your buck on a per-watt basis, but you're also getting a higher maximum wattage.


Questions? Comments? I'd love to hear it. And just give me a day or two, and I can draw out designs if my ultra-wordy description didn't cut it.
 

·
Registered
Joined
·
1,085 Posts
Xmax is seldom an important consideration for spl subs. All the early spl champs like the CV stroker and the early DD subs were evenhung, which is where the winding height is the same as the top plate thickness, and is a great method of getting motor force and efficiency.

The split coil idea has been tried with some of the RE xxx subs and works great for linear bl, but is not at an advantage for pure spl because of the heavy coil.
 

·
Premium Member
Joined
·
11,704 Posts
My idea? Well, there's two parts. The first is: What if we merged the two designs? To do that, you would need a coil with a wind height exactly as tall as the top plate thickness. Now, normally, with just one of these coils, you wouldn't get much travel. None of it would be linear, and it would be hard to predict how it would move in response to a signal. But, and this is the second part of my idea, what if we had multiple coils like that? Not in parallel or in series, but individual coils that are activated one by one, as the voice coil moves through the gap.

Questions? Comments? I'd love to hear it. And just give me a day or two, and I can draw out designs if my ultra-wordy description didn't cut it.
I know someone who has done a design like that -- but it was controlled via optical sensors / digital controls.

Worked very well from what I heard but didn't go into production :thumbsup:
 

·
Hurtin' Feelins' errrday
Joined
·
12,531 Posts
Where's Jeff??!!
 

·
Hurtin' Feelins' errrday
Joined
·
12,531 Posts
Agreed..

Maybe later when I got some time.
 

·
Fi Enthusiast
Joined
·
343 Posts
well that was interesting, i have no knowledge to pass but am tuned in for the explanation lol
 

·
Registered
Joined
·
1,270 Posts
Moving mass will be an issue since it will have nearly as much coil as an overhung design, but you won't have the same power handling. There are trade-offs for every design.

Since you're interested in different motor topologies check out XBL^2, LMT, and split coil. Some examples of cool designs to check out are Fosgate T3, W12GTi-MkII, and DSS Tridents

I also suggest you read through some engineering and/or physics books. It's very important to know what formulas to apply and when. You tend to oversimplify or use incorrect formulas in your calculations in most of your posts. Formulas are only good when they are put into the proper context and physical limitations are reconciled. This one for example your results should be in Volts squared not watts and I'm unsure about how you handled the AC phase angle in this particular instance to comment.

So how did you figure 6x more efficient than an overhung design? If you want to try it; go for it. Experimentation is good for the soul. :)
 

·
Registered
Joined
·
60 Posts
Discussion Starter #9
Yeah... The problem isn't so much that I don't know as much as I should about engineering or physics, it's more that im one of the rare few that likes math but is terrible at it. Something about not thinking things out completely before moving on to another facet of the design.

Anyway, you're completely right. I screwed that up in several different ways. The actual formula involves a modified triangle wave that's hard to model mathematically but easy to find the average/integral graphically/geometrically. So you just gotta trust me that the actual figure is around 5.4 times as efficient.

I'm around 6 shots into the night (not an alcoholic, just celebrating the new year) so I have no idea if that made sense. I'll definitely check out that shit later when I sober up. As far as actually building the thing goes, I'm stuck on finding motor parts for the thing, as well as sourcing really thin 3" aluminum pipe to make the VC out of. Anyone have any idea where I can purchase whole 3" motors or 3" motor parts btw?
 

·
Registered
Joined
·
1,085 Posts
You're not going to find aluminum pipe thin enough to work for a former, but pop or beer cans work well for that. You're going to try and wind your own coils, though? As someone who's tried that before - good luck!
 

·
Registered
Joined
·
1,270 Posts
What do you mean by efficient? Efficient use of flux or overall driver efficiency? I'm just skeptical of your ~6x efficiency calculation... I know I come off as busting your balls, but working in STEM fields has taught me sometimes its best to be critical of designs early on for myself and others. Passing a design through a gauntlet early on results in less time and money wasted when potential flaws of a design are addressed early on. I would love to see you complete your design and be fairly happy with it. (I doubt you will see the results you're expecting though)

Also something to think about is how to make your design durable. Your design will have a lot of possible fail points due to it's complexity. It would suck to build it and it just go pop. You don't happen to have access to a machine shop do you?
 
1 - 11 of 11 Posts
Top