Need some help with my B&C 18SW100 design

Chris Houghton

Freshman
Jan 26, 2013
2
0
1
Orange County, CA
I'm trying my hand at designing some dual 18's based on the B&C 18SW100 drivers but I think I might be in over my head. The design criteria for me is something that will pack into a 90" wide truck on at least two dimensions. For this I settled on 1/4 pack x 1/2 pack. I'm also trying to get solid performance in the mid to low 30's. These will mostly be used for 4-5 piece pop/rock bands in small venues and small outdoor stages for community type events. More and more of these type of events are incorporating DJ's these days so I'm sure I'll have a fair amount of that to deal with too. Finally, I'm taking this opportunity to learn. I realize I can just buy something that has already been thoroughly R&D'd with tuning and processor settings sorted out. I realize that is often the cheaper route as the time involved in the process is often not worth it in the end. I know this going in! With that out of the way, here's what I have so far.

In a somewhat backwards but necessary move, I decided on two of my three dimensions before even looking at a model. I built it up in Sketchup so it would be easier to measure internal box volumes. It also helped me visualize port depths and bracing.

While I'm not overly attached to this B&C driver, I chose it because it's readily available to me. I'm open to other suggestions. The port design I chose is similar to the eighteen sound DIY plan or Clair BT218 where I am using large triangular ports that use the cabinet as a port wall. I don't know how to model the effects of this in software yet. I've been using Bass Box Pro 6 and according to it's model, I have a current Fb of 33.51Hz and an F3 of 43.58Hz. This seems wrong to me. I assume it's an error on my part.

Also of concern is the cone displacement at around 50Hz. It seems quite high to me. I have to believe this driver can perform better than this so I must have my tuning or box volume wrong.

Since my design is a dual 18 with two independent chambers, I found it easier to model just one chamber. Is that a problem?

Can someone tell me if I'm on the right track? Should I just build one and see how it performs in the real world?

I understand what I'm looking at is a raw modeled response with no dsp but I'm not used to looking at it that way.

I've attached all the relevant Info I could including the Bass Box Pro 6 file, the Sketchup model, the B&C spec sheet, and a PDF with the specs and measurements that bass box spits out, plus some renderings for those that don't want to download the zip file.

Thanks in advance!
 

Attachments

  • Sub Design Sheet.pdf
    1.3 MB · Views: 44
  • 18SW100_8Ω.pdf
    406.6 KB · Views: 15
  • BB6 & Sketchup Files.zip
    69.9 KB · Views: 16
Hi! Here's some comments/feedback as someone who's built a double 18" subwoofer long these lines.
  1. Overall Box Tuning: Your "target" volume and tuning for the 18SW100 should probably be ~33Hz and ~200L (7 cu.ft.). These are B&C's numbers and they simulate well. I'd tune a little higher myself.
    1. You are indeed correct to model it as each chamber individually.
    2. 22.5" x 45" is a perfectly acceptable dimension to restrict yourself to, but you're going to need to make it at least 30" deep, probably more like 32" or more.
    3. Interestingly, you have created a cabinet in BBP6 that is the correct internal volume before the port is calculated and drivers are added. This is a weakness of BBP6, as the port in simulation is nearly always significantly longer than you need. If you can put in a correction factor, K, then you should put in about 1.7. Somewhere on this forum is the figure I'm using for this, but I can't ever find it, so I'm referencing my printout. I wish WinISD did this, without glitching and refusing the input, but it's free so oh well. The reduced internal volume in the sim is causing most of the garbage you're seeing in your freq. response.
  2. Port Tuning: Your port should be even bigger than this.
    1. Given a port a little bigger than this, several of us have arrived at about 10" or so for our port depth. Personally, I do mine a little bit longer, but it's all personal preference and prototyping.
    2. Again, port length in simulation is never equivalent to the real world. It's always wayyyy longer than it should be, especially in BT218 like designs where boundaries play into the port as much as they do.
    3. Triangular ports are kinda garbage. Just saying.
  3. Cone Excursion: Anything besides a K20 or bridged PL9.0PFC can't break the large format B&C subwoofer drivers.
    1. Your cone excursion is representative of the real life scenario. I assume that you are simulating this with the quoted Xmax on the B&C spec sheet. Forget it. Use Xvar, and remember that you have 28.5mm of excursion before damage. That should calm your fears.
Alright, I don't think I've given too many secrets away that you couldn't have figured out by spending an excess of money on plywood for a prototype and then kicking yourself.
 
Hi! Here's some comments/feedback as someone who's built a double 18" subwoofer long these lines.
  1. Overall Box Tuning: <snip> If you can put in a correction factor, K, then you should put in about 1.7. Somewhere on this forum is the figure I'm using for this, but I can't ever find it, so I'm referencing my printout. I wish WinISD did this, without glitching and refusing the input, but it's free so oh well. The reduced internal volume in the sim is causing most of the garbage you're seeing in your freq. response.

This one?
atoKTQw.jpg


If so, this spreadsheet should make the number crunching easier : https://drive.google.com/file/d/1tl8A4Fbkk07TmR7h2vyxVBQcDaCFOwri/view?usp=sharing
EDIT: NB all units are cm for linear and litres for volumetric.

HTH,
David.

PS, agree with everything else you said too.
 
Last edited:
  • Like
Reactions: Max Warasila
I have a dual 18 plan laying around if you're interested. It has a SB1000'ish layout, but different port configuration.
Only problem.. I haven't build one yet, so some adjustments would be necessary.

If you're interested, PM me.
 
Hi! Here's some comments/feedback as someone who's built a double 18" subwoofer long these lines.
  1. Overall Box Tuning: Your "target" volume and tuning for the 18SW100 should probably be ~33Hz and ~200L (7 cu.ft.). These are B&C's numbers and they simulate well. I'd tune a little higher myself.
    1. You are indeed correct to model it as each chamber individually.
    2. 22.5" x 45" is a perfectly acceptable dimension to restrict yourself to, but you're going to need to make it at least 30" deep, probably more like 32" or more.
    3. Interestingly, you have created a cabinet in BBP6 that is the correct internal volume before the port is calculated and drivers are added. This is a weakness of BBP6, as the port in simulation is nearly always significantly longer than you need. If you can put in a correction factor, K, then you should put in about 1.7. Somewhere on this forum is the figure I'm using for this, but I can't ever find it, so I'm referencing my printout. I wish WinISD did this, without glitching and refusing the input, but it's free so oh well. The reduced internal volume in the sim is causing most of the garbage you're seeing in your freq. response.
  2. Port Tuning: Your port should be even bigger than this.
    1. Given a port a little bigger than this, several of us have arrived at about 10" or so for our port depth. Personally, I do mine a little bit longer, but it's all personal preference and prototyping.
    2. Again, port length in simulation is never equivalent to the real world. It's always wayyyy longer than it should be, especially in BT218 like designs where boundaries play into the port as much as they do.
    3. Triangular ports are kinda garbage. Just saying.
  3. Cone Excursion: Anything besides a K20 or bridged PL9.0PFC can't break the large format B&C subwoofer drivers.
    1. Your cone excursion is representative of the real life scenario. I assume that you are simulating this with the quoted Xmax on the B&C spec sheet. Forget it. Use Xvar, and remember that you have 28.5mm of excursion before damage. That should calm your fears.
Alright, I don't think I've given too many secrets away that you couldn't have figured out by spending an excess of money on plywood for a prototype and then kicking yourself.

Thank for this awesome info! I have a few follow up questions for you.


Your "target" volume and tuning for the 18SW100 should probably be ~33Hz and ~200L (7 cu.ft.).

When you say target volume, is that the raw internal volume of the box or the volume after accounting for driver, bracing, and ports?


If you can put in a correction factor, K, then you should put in about 1.7.

I haven't seen a place to do this in BBP6. Is there better software I should be using?


Port Tuning: Your port should be even bigger than this.

Are you referring to the mouth of the port or the depth... or both?


Triangular ports are kinda garbage. Just saying.

What makes a triangular port less desirable than a round or square port?


Use Xvar, and remember that you have 28.5mm of excursion before damage.

Is this because the voice coil travels in two directions?


Thanks again for your help so far.
 
I ran it thought WinISD - B&C's recommendations looked perfect to me. You could go a little smaller - 150L and tune it a bit higher 35 - 37 Hz. This would reduce cone travel a little but the box would not go quite as low.

The advantage of a triangular port is how it braces the side walls of the box without adding any extra weight. From a fluid dynamic perspective there are better ways to design a port.
 
Thank for this awesome info! I have a few follow up questions for you.

When you say target volume, is that the raw internal volume of the box or the volume after accounting for driver, bracing, and ports?

Specifically, I mean Vb, the resonating body that remains after the volumes of the driver, port, and rest of the box has been accounted for.

I haven't seen a place to do this in BBP6. Is there better software I should be using?

I'm to cheap to have tried anything other than WinISD. BBP6 does the same math, just with some extra design tools on top of it. You could always model in detail using AkAbak, but that's so much pain for the same results 90% of the time.

Hornresp may have this functionality, but I don't recall. Peter Morris can probably speak on that off the top of his head. Or from the mouth if that's easier.

Are you referring to the mouth of the port or the depth... or both?

I'm referring to the port area. Try to get at least 50% cone area in port area. At least.

What makes a triangular port less desirable than a round or square port?

The corners cause turbulence and do not actually add port area as a result. For very short ports, I actually quite like them, but in general any acute angle will cause undesirable flow within a port. This is not my area of expertise, sorry I can't explain much further.

Is this because the voice coil travels in two directions?

No, the 28.5mm number comes from the quoted 57mm peak-to-peak excursion before damage number given by B&C. Xvar has to do with the repeated posts by Bennett Prescott saying it's a more representative number for the B&C drivers performance in most applications.

Thanks again for your help so far.

Sure thing!