Car Audio Classifieds! banner

I just built my first custom sub. How do I figure out how much power to throw at it?

3K views 29 replies 14 participants last post by  Jeffhulbert50 
#1 ·
Well, my first "from scratch" build went off without a hitch, and just 5 hours after the parts came in the mail, I got it all clamped up and drying in the kitchen. (Fun fact: the guys over at fixmyspeaker.com have never built a quad leadwire 2.5" spider before, or at least up until a few days ago. They thought I must have made a mistake ordering the damn thing.)

Overkill? Probably. But my build's full of overkill, so what's a little extra?

Anyway, after it's all said and done, there's still one question that remains unanswered: How much power will the damn thing take? I was planning on hooking it up to my amp and increasing the gain until I got scared, but there must be a more scientific way of finding that out.

Lets get down to the nuts and bolts of it:

Im using an Arc Audio Flatline motor, which has double stacked 6.7" dia ferrite magnets and a .51 inch top plate. Conveniently, the bottom plate is also milled out a fair bit, which means the bottom spider will contact the top plate about 3/16" before the voice coil ever contacts the bottom plate, so bottoming out and ruining the voice coil is impossible with this design.

Here's the link to the voice coil I'm using.

As mentioned before, I got quad leads on this thing, so that's not gonna be an issue. Soft parts (to the best of my knowledge) don't matter much, but they're all from fixmyspeaker, so they're top notch, if that means anything.

Anyone got an idea on this one? If you need more info about anything, just say the word. Thanks.
 
See less See more
#3 · (Edited)
Honestly, without getting all Einstein, here is only one way to find out what kinda power your driver can take.
And I'm afraid we gone need some pics or there will be consequences ...
 
#7 ·
OP TO THE RESCUE!!!



Unfortunately, this is the only non-blurry photo I took mid-build. Sorry. I was just so focused, that I just didn't even think of do***enting what I was doing.

But hey, to make it up to ya, I ran out to the other room and took a few photos of what it looks like now:








Literally all the welding clamps I could find in a 5 block radius went towards clamping that surround. And my god, what a sexy, sexy surround it is.


Any idea what stiffness and how many spiders you have?
Did you use a spacer if more than 1
I have 2 8" medium compliance spiders. Used a .25" spacer. By the way, as far as I can figure out, there's no good way to install spacers. Unless you're literally building the thing in the basket from the ground up, spider by spider, you really have to get your hands dirty to get those things glued right in between the spiders. That was probably the hardest part of the build, by far.


the issue with that much lead is mass. there heavy and restrict movement when the spider tries to move. they basically far exceed the thermal power limits of the coil, hence why they didn't understand why you wanted that.
Restrict movement? Even with PSI's fancy unobtanium thread, hand stiched a billion times around each lead for maximum durability and mobility? You know, the thing they're constantly bragging about on their websites?

In all seriousness, though, these things are stitched in SOLID with a good amount of slack between threadings. If your talking about the effect extra weight has on the spider, yeah, I can agree with you there, but as far as actually pulling the spider and restricting it from stretching, I can't see that happening.

mechanical limit is when the movement exceeds something that restricts its movement. if the surround allows so, the coil hitting the top plate would be the limit. the linearity of the spiders will create distortion at high throw once the leads start to hinder smooth travel.

thermal limits will only matter if the power applied is being put to use via movement of the cone. lets say the cone moves a max of 1" each way but you keep adding power. at that point the power is simply wasted and increases the coils temp causing rise in resistance.
Believe me, I understand that completely. But knowing a ballpark figure of how much power you can expect to handle is still very useful and comes in handy pretty often when you're first setting things up- for example, I need to know whether I should wire the sub in series and bridge two amps at 1 ohm each to get 3K watts maximum RMS, or wire in parallel with one amp at .5 ohms for 2300 watts, or wire in series with one amp at 2 ohms for 1000 watts. So yeah, I get that more power =/= louder, but I still need to know.

And just for future reference's sake, here's the breakdown on the excursion limits: xmax, one-way linear: 1"; one-way non-linear: 1.5"; absolute max mechanical excursion, one way: 1.9375".
 
#4 ·
the issue with that much lead is mass. there heavy and restrict movement when the spider tries to move. they basically far exceed the thermal power limits of the coil, hence why they didn't understand why you wanted that.

theres two power limits. thermal and mechanical.

power handling doesn't equate to its wow factor or worthiness. especially if it doesn't use the power for gaining output.

mechanical limit is when the movement exceeds something that restricts its movement. if the surround allows so, the coil hitting the top plate would be the limit. the linearity of the spiders will create distortion at high throw once the leads start to hinder smooth travel.

thermal limits will only matter if the power applied is being put to use via movement of the cone. lets say the cone moves a max of 1" each way but you keep adding power. at that point the power is simply wasted and increases the coils temp causing rise in resistance.
 
#6 ·
Any idea what stiffness and how many spiders you have?
Did you use a spacer if more than 1

That coil should take around 1000rms fairly easily
But I think you will reach mechanical limits before/mechanical problems if your spiders are too stiff

It should be a great sub
 
#9 ·
Literally all the welding clamps I could find in a 5 block radius went towards clamping that surround. And my god, what a sexy, sexy surround it is.


I have clamp rings so you don't need so many clamps when you recone.

I have 2 8" medium compliance spiders. Used a .25" spacer. By the way, as far as I can figure out, there's no good way to install spacers. Unless you're literally building the thing in the basket from the ground up, spider by spider, you really have to get your hands dirty to get those things glued right in between the spiders. That was probably the hardest part of the build, by far.




Restrict movement? Even with PSI's fancy unobtanium thread, hand stiched a billion times around each lead for maximum durability and mobility? You know, the thing they're constantly bragging about on their websites?

yes. a lead wire isn't without resistance. they add a lot of weight and strain to move them when bunched together. ask anyone that designs subs. has nothing to do with the thread or how many wraps they sew them with.

In all seriousness, though, these things are stitched in SOLID with a good amount of slack between threadings. If your talking about the effect extra weight has on the spider, yeah, I can agree with you there, but as far as actually pulling the spider and restricting it from stretching, I can't see that happening.

it does and will effect efficiency . you basically doubled the resistance of the leads normally used.

Believe me, I understand that completely.
But knowing a ballpark figure of how much power you can expect to handle is still very useful and comes in handy pretty often when you're first setting things up- for example, I need to know whether I should wire the sub in series and bridge two amps at 1 ohm each to get 3K watts maximum RMS, or wire in parallel with one amp at .5 ohms for 2300 watts, or wire in series with one amp at 2 ohms for 1000 watts. So yeah, I get that more power =/= louder, but I still need to know.

And just for future reference's sake, here's the breakdown on the excursion limits: xmax, one-way linear: 1"; one-way non-linear: 1.5"; absolute max mechanical excursion, one way: 1.9375".
 
#12 · (Edited)
How did you arrive at your throw numbers? Xmax should be driver movement within ideally 70% BL and xmech should be mechanical limits.
I ran a full magnetic simulation of this motor in FEMM and I still don't know what "70% BL" means. I just used the formula (wind height-top plate thickness)/2. Since nearly all of the magnetic flux lines go through the top plate and into the pole piece, I figured it was a pretty accurate rule. As for how I actually measured it, well, calipers and carefully.


Wire it to .5 on one amp and be happy
You can always turn the gain down
/thread. Probably the best option, I'll give you that.

That was a mistake. The Flatline motors from Ultra are not very powerful, with any kind of stiff spiders or surrounds they end up very high Q.. especially with aluminum coils. It would not surprise me if that sub has a qts higher than 1.
Well, worst comes to worst, and the speaker is worth more to me rebuilt than built, I can always save on having to buy more parts by using the hairdryer trick on the rubber glue to remove the surround/spiders, and nitromethane RC fuel as a solvent to remove the CA for everything else. It'd be a delicate, 8-12 hour job, but it could be done.

But in the mean time, we'll see how she plays. Not loud enough? We'll turn up the gain, rinse n repeat until we're either satisfied or she goes up in a ball of smoke. I'm sure it'll be fine.
 
#13 · (Edited)
After much trial-and-error, I have found a way to use FEMM (AKA Finite Element Method Magnetics, a free program you should totally go out and download) to calculate this motor/coil combination's BL, which the manufacturer does not provide and which cannot be found anywhere else. Just a note before we begin, BL is the product of the average field strength in the gap between the top plate and the pole piece and the total length of the wire in that gap, so it's proportional to the inverse of the width of the wire used
(less width = more turns of wire in a given area) - meaning that it would be higher with higher resistance coils or coils with a shorter wind height. Seeing as how I'm using a D1 coil with a 2.5" wind height, don't expect a high BL. The actual number we should be paying attention to in order to rate the strength of the motor is the average B in the gap; after all, if I instead chose to use a D4 coil with a 5/4" wind height, I would end up with a BL 4 times higher than what I currently have.

I'll be posting screenshots of FEMM on a step-by-step basis to show you guys how I get the numbers I got, and to hopefully guide those who want to do the same thing. It'll make sense to those of you who are familiar with the program. To those of you who are not, well, just trust me on this, and even if you don't, hopefully you'll enjoy the pretty colors anyway.

Alright, well, on to the mathy stuff:


Step 1:

Before we do any analysis, we must first make a model in the program that has the same dimensions and is made of the same materials as the real motor we are simulating. If you don't already know how to do that, FEMM has some great do***entation that takes you step-by-step and teaches you how to do it. I highly recommend learning how to use this program; the learning curve is not steep, and it's super useful. Anyway, here's my model:



Step 2:

Add a couple lines connecting the side of the top plate to the side of the pole piece. Add a block label to both this new block and the space below it, label them both the same thing you labeled the surrounding air before. We're adding this new block just to simplify things down the road.



Step 3:

Mesh, simulate, then view result:



Step 4:

Select gap between top plate and pole piece, select integrate -> Integral of B over block, press OK:



Step 5:

Ignore the z-component and the fact that the r-component is negative. We only care about the absolute value of the r-component. The value of the r-component is in Tesla*Meter^3, which tells us is that the field in the gap is equivalent to a uniform field with strength 3.45916 * 10^-5 T occupying one cubic meter. But the volume of our gap is much less than one cubic meter, so the equivalent uniform field strength must be much greater. Think of it as being more concentrated.

To find the average B of the gap, first find the volume of the gap by, again, selecting the gap, then selecting Integrate -> Block Volume:



Step 6:

Now, divide the TM^3 figure you got in step 4 by the volume you got in step 5:

3.45916 * 10^-5 T*M^3 / 5.75668 * 10^-5 M^3 = .601 Tesla

So, your average B is .601 Tesla.

Step 7:

To find your L, you'll either need a picture of or the actual voice coil to inspect. Count the number of turns in the first layer, multiply this number by the number of layers to find the total number of turns, then multiply by your top plate height, then divide by your winding height. I counted 43 turns in the first layer, so, the calculation would look like this for me: 43* 4 layers *.5"/2.5" = 34.4 turns of wire in the gap. To find the length of the wire, take the diameter of your voice coil in inches, and multiply that by .0254 to get the diameter in meters, then multiply by pi to get the cir***ference in meters, then finally multiply by the number of turns in the gap to get the total length of the wire in the gap. For me, that would look like: (2.5 in * .0254 M/in * Pi) *34.4 turns= .1995 M/turn * 34.4 turns = 6.8625 meters.

Step 8:

To find the BL, just multiply your B times your L!

For me, that would be: .601 T * 6.8625 M = 4.124 TM = 4.124 BL

I think we know why Arc Audio never published the BL. Christ. I didn't even know a subwoofer with a 4.1 BL could even play (that is, assuming it plays when I install it.). Please, I'm begging you guys here, please find whats wrong with my math. I just can't accept the idea that my brand new sub is a goddamn 4.1 BL lemon.
 
#27 ·
Something is strange with your model, there looks like a weird boundary condition that's causing a problem with the circuit. Did you modify and use the woofer.fem model that came with femm?

I think with the route you went to specify a boundary around the air in the gap, the B it calculates from it is going to be the ***ulative for that area and not a good representative of what the coil sees. I say that because .601T is really high for a motor that size - as an example, a larger 3" TC9 motor with a 1/2" top plate only has about .5T

The way I've always done it is to leave the air in the gap open without a boundary, draw a line through the gap, and then just plot the B along that line as a graph. It gives a better representation of what the coil will see, and you can look at the symmetry of the B field in the gap (important). Here's an example of this with one of my models of a modified Destroyer motor. Notice the difference in flux flow between our models.

 
#17 ·
Holy shit guys, I just installed it and played it for the first time. First off, against all odds, math, and even the laws of physics, this thing plays LOUD. Id say about as loud as 1 1/2 12" Fosgate P3s. I actually had to open the trunk and make sure this was the only sub playing because I frankly did not think it was after hearing it for the first time. I really can't understate how much math has lied to you and I both- it may have a BL of 4, but it bottoms out at ~1000 watts and like 4 inches p2p excursion.

And that's not even the most amazing part! The sound quality, oh god the sound quality- this thing plays LOW. And I mean, the lows you can't even hear, only feel. The lows my old RF P3s can't even touch- and they're SQL woofers.

I ****ing love this ****ing thing.
 
#21 · (Edited)
I meant "bottom out" in a much different way than it's usually used- I didn't mean that the coil was bumping into the bottom plate, as that is impossible with this design. Instead, I meant that the bottom spider was bumping into the top plate. It's also a type of bottoming out, I guess, but not as dramatic or damaging. Still makes a clicking sound, I assume because the rigid, ceramic-like glue on the bottom of the spider is what's making contact. I should have probably been more clear about what I meant.

Why would you think your calculated BL would tell you the entire story? The physics say once a subwoofer is installed in a vehicle it's far too difficult to approach it analytically. It's not that math or physics lie, it's our limited ability to model extremely complex calculations, hell look at the 3 body problem. :). Glad you like it and keep up the experimentation and exploration!
I called it glad your happy
Thanks guys
 
#20 ·
Why would you think your calculated BL would tell you the entire story? The physics say once a subwoofer is installed in a vehicle it's far too difficult to approach it analytically. It's not that math or physics lie, it's our limited ability to model extremely complex calculations, hell look at the 3 body problem. :). Glad you like it and keep up the experimentation and exploration!
 
#22 ·
Well, it's not hard to beat Rockford P3's, haha
 
#23 ·
No, but in a world of instant gratification, it is nice to see someone slow down and learn. RF put some R&D behind their product...Didn't they?
I liked learning about the FEMM.
Continued success sir.
 
#25 ·
Good point, my only RF experience past the 90's is the Power T600-2 that runs the sub in my work car. I got it in trade for some work I did, zero experience with the new subs.
 
#26 ·
RF does a lot more R&d on there higher end then you guys give credit for. just like many brands, they let china taylor the lower end with a few pointers from the customer. look at the new drivers they came out with. there like 20 members here that actually know whats special about it. is it worth the cost, maybe not, but atleast they didn't cookie cutter the hell out of it like every other overnight company
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top