Straight Line Array SB65WBAC25-4

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
For a floor to ceiling array I'd at least try it unshaded first. For a finite array shading makes a lot of sense.
I had plans for frequency dependent shading but started with an unshaded array. Even though I'm still curious how it would work out I never bothered with it as the results were very pleasing even before using Fir filters.
The frequency dependent shading would still allow me to use the array full range on the bottom end. Who knows, maybe some day I'll find the time :).
 
Notice the discussion and photos about the internal brace/nutplate. The nutplate provides an excellent way to attach the baseplate to the array towers. A substantial baseplate was essential for my arrays and will be for your straight array.

Mounting and attaching the SB65 drivers pose challenges which need some thinking to resolve before you cut wood. The SB65s are tiny!

Thanks for the information Jim Griffin, I've read over your threads and they are filled with number of good ideas. I'm planning on building a test baffle to sort out the driver mounting and planned on chamfering the front and back of the driver cutouts due to the increased baffle thickess. I might move to 0.5" since everyone says it should be good enough. The base mounting is a great idea, not sure I have the skill to replicate that though.........

I was also thinking of pressure mounting the top of the array to the ceiling with some foam or adjustable top plate. I'm allowed to wreck the ceiling as long as I fix it:). Will definitely be using the crimps and will take your advice on using tweezer to help with the attachments. I will be taking my time with the assemble since I'm not looking forward to redoing ANY part of this speaker!
 
For a floor to ceiling array I'd at least try it unshaded first. For a finite array shading makes a lot of sense.
I had plans for frequency dependent shading but started with an unshaded array. Even though I'm still curious how it would work out I never bothered with it as the results were very pleasing even before using Fir filters.
The frequency dependent shading would still allow me to use the array full range on the bottom end. Who knows, maybe some day I'll find the time :).

Great thread on the making of your LAs! Very informative:D.

With the 5x7 setup I was thinking of chambering the drivers in groups of 5 with terminals to the rear to allow for shading later on without having to rewire the array. Then I realized I wouldn't be able to get my hands into the chambers to attached the wires to the posts. I would have to build 7 access chambers in the rear for this. I might do a 1 chamber test build to see if it will be too much work or not. I do like the idea of being able to work on the array at various points if there is trouble but..........Also based on the corner array threads, your build and absorbing the mirror workings of the ceiling and floor/reading the Murphy corner array I think I will try no shading first. Also I do not have the skill to build a CBT physically so maybe electronically if so inclined at a MUCH later date.

The test build points to xrk971 suggestion of making separate enclosures and mounting them altogether but I'd like to stay with 1 single enclosure. It will work better with my intended setup. I think pipe clamps will be a solution for me if Chris and Dave do not have a distributor for a 4x8 oversize piece of BB or the cost to ship a few pieces in is prohibitive.

Going to read Jims threads again regarding driver mounting. Thanks for all the suggestions, the more the better. Hoping to have the drivers by the end of next week:)
 
mlee - whereabouts in Canukstan are you located? There are several sources of "Baltic Birch" type plywoods in the lower BC mainland area where we're located, and the material is actually available in 48x96"' sheets (129x2438 mm), as well as the goofy 60" (1524mm) square sheets.

Windsor Plywood carries the latter in 3mm increments from 3 to 25mm nationwide, I believe, but the former is a bit more restricted in distribution. The material that was used in the Planet10 flat pack kits was 15mm Maple face veneer (Multi-ply) by Murphy Plywood out of the States, and as far as I know, only available in BC through E Roko Distribution. The platform core is European sourced, but due to the extra processing for the finish veneer faces, is quite a bit more expensive than the "shop" grade of same sizes. For those, Reimer Hardwoods, Upper Canada Forest Products and perhaps some branches of P J White Hardwoods may stock material they'll likely refer to as either "Euro" ply or Baltic. Google those names to find distribution locations that might be near to you.
 
chrisb many many thanks for the information! I'm located in the lower mainland BC so according to E Roko Distribution website, they are located in Burnaby which is easy access for me :). I'll give them a call on Monday when they open and see what my options are. Only 1 source, geesh, that makes it much harder for us DIYers!
 
Does anyone know how to model this mlee straight line array design in hornresp?
The driver specs are downloadable: http://www.loudspeakerdatabase.com/Hornresp/SB_SB65WBAC25-4.txt
driver dimensions are 64x64mm
35 drivers
so about 225cm high
is there a way to model five drivers in a shared sealed volume versus each driver in its own sealed box?
If you can model it in hornresp please attach the hornresp Export File so I cam learn how its done. Cheers.
 
mlee - I've just retired from over 25yrs working in the commercial millwork/cabinet trade, during which time the company had the opportunity to buy from many product distributors, and "experiment" with quite a few trends in the sheet products category (anyone remember "WheatBoard?" :eek: ). I've also built at least 100(?) pairs of speakers and kits for Dave, so certainly have my own perspective / bias on the subject.

The "only one source" - in our geographic area - would refer specifically to the "cabinet" grade (i.e. whole piece Maple veneer face) plywood by Murphy of Oregon. The 5/8" ( 15mm) that we'd settled on as a compromise for a wide range of our builds and flat pak kits is not as high a volume demand item in the trade as the more commonly used 1/2" and 3/4", and it was only due to a couple of custom factory mill orders ( i.e 30 sheet minimum) through the Victoria office that Roko brought them in at all. They, and the other sources listed also carry "shop" grade Euro sourced birch faced plywood in 4x8ft sheets in the more commonly used 1/2" & 3/4" thicknesses. If you intend to apply any type of finish to the boxes other than paint or stain & clear coat- i.e plastic laminate or veneer, the much cheaper C4 grade birch veneer face product is more than good enough - but note that can be a bit of a crap shoot to try to stain or clear coat the raw birch faces, and a furniture grade paint job on plywood is a whole 'nother world of pain.
 
Founder of XSA-Labs
Joined 2012
Paid Member
Does anyone know how to model this mlee straight line array design in hornresp?
The driver specs are downloadable: http://www.loudspeakerdatabase.com/Hornresp/SB_SB65WBAC25-4.txt
driver dimensions are 64x64mm
35 drivers
so about 225cm high
is there a way to model five drivers in a shared sealed volume versus each driver in its own sealed box?
If you can model it in hornresp please attach the hornresp Export File so I cam learn how its done. Cheers.

I don’t think HR can model 35drivers in an array. I have done this in Akabak for Wesayso’s 25 TC9FD array. It shows the theoretical frequency response falloff that then requires EQ to make flat. The simulation matches quite closely to actual response.
 
Here is where the compromises set in :) Been doing some more reading and please correct me if I'm wrong but looks like a full ceiling to floor array might be better off NOT shaded while as a non-floor to ceiling array is better off as CBT.

That said I was originally thinking if sectioning sets of drivers in the array so I could delay and reduce the amplitude of the sections to create a simulated CBT from a straight array.

Audio Engineering Society
Convention Paper 5653
Implementation of Straight-Line and Flat-Panel
Constant Beamwidth Transducer (CBT)
Loudspeaker Arrays Using Signal Delays

This poses a problem with the 7x5 array in that the impedance of the sections will not play nice unless there are an even number of drivers in a section. The best compromise for this functionality would be to use xrk971 suggestion and do sections of 6 drivers with the last sections using a resistor and inductor to simulate the last driver impedance.

With the above in mind for 6 driver section and 6 section array, each section could be 6 ohms nominal impedance. This would result in a 6 ohm nominal impedance for the 6 section array vs 5.6 ohms nominal for the 7x5 array.

The 7x5 array seems nicer in that each chamber will be the same size and same driver make-up. Also does the CBT still work if there is another boundary 2 inches away from the top of the array? Seems like this might create more problems than it solves. Thoughts?

Throwing out the idea of a CBT I suppose if the array was chambered off then you should simulate a point source via shading or make a focused sectioned array at the listening position. Thoughts?

Any other configuration options that would increase functionality?
Thanks!
 
Arrived!
 

Attachments

  • SB.jpg
    SB.jpg
    166.6 KB · Views: 498
.....Any other configuration options that would increase functionality?
Thanks!

I wonder if it possible to squeeze 36 drivers in with a floor to ceiling stack? I wonder if you'd like to remeasure the exact dimensions in mm, floor to ceiling at both speaker spots and then measure in mm the linear displacement of six adjacent drivers eg measured face down on a table please?

An 8 foot ceiling is under 2440mm and a 36 stack of adjacent drivers is only 36x64mm = 2304mm so theres a nice 100mm+ reserve for a useful margin of error for most ceilings.

Dont worry about needing much room for a base or top plate. A cunning design needs neither allowing floor to ceiling drivers in an elegant slender tower that can be stable via firm floor and ceiling coupling. The tower could be made from just one sheet of thinner plywood cunningly bent into a tear drop cross section with the complex curves offering vibration resistance and a narrow rounded front section offering minimal diffraction compromises.
 
I wonder if it possible to squeeze 36 drivers in with a floor to ceiling stack? I wonder if you'd like to remeasure the exact dimensions in mm, floor to ceiling at both speaker spots and then measure in mm the linear

Yes I will remeasure later tonight for thoroughness as 36 a side would make things easier. For thought exercise if 66mm centre to centre spacing is realised, frame edge to edge is 2374 mm = 93.46457 inches. Although I do have a sliding door for that room I might have significant problems standing the speaker upright. I think I'd like a little bit more clearance to be safe. Pretty sure my ceiling is only 8ft but maybe can squeek out another inch? Will measure in the exact spot the speakers will go.
 
The height of the spot the speakers will be installed is 95" so it I would be hard pressed to get 36 drivers in at spacing 66mm.

Attached is a comparison of size vs TG9FG. Also cutouts are 66mm CTC spacing and 2-3/8" cutout. There is a tiny bit of wiggle room at 2-3/8" cutout. Just enough to slightly move the drivers around. It is a tight fit and no room for cumulative spacing error.

These are small drivers and 3/4 front baffle thickness is pretty big so will probably go with 1/2 inch front baffle thickness and 3/4 everywhere else. Will have to route out a bit of the rear opening to ensure the quick connects fit and not rub on the sides under vibration.
 

Attachments

  • compare.jpg
    compare.jpg
    88.3 KB · Views: 450
  • IMG_20180816_180851.jpg
    IMG_20180816_180851.jpg
    52.6 KB · Views: 434
  • unloaded.jpg
    unloaded.jpg
    66.5 KB · Views: 451
Perhaps also consider a smaller diameter through hole but add reliefs for the basket webs (these extend enough to dictate a 2 3/8" diameter cutout) and the side connector tabs (these need 2 1/4" cutout to clear). If surface mounted on 3/4" thick material, the baskets would be blocked for their entire depth so the small speaker doesn't have much room to breathe.

If you are using CNC to cut the holes, then you can include reliefs, chamfers, and edge routing to mitigate some of these conflicts yet maximize the mounting landing pad. I'm on the side of reducing the through hole diameter to assure a good mount while adding reliefs where needed.

Of course if not CNC'ing the baffle you can use creative hand routing or a Dremel tool or manual filing to carve out clearance where needed.
 
Last edited:
Great ideas Jim Griffin. I will try the smaller cutouts tomorrow and clearing some material for the tabs. I will be using a router so maybe a template for clearing the wire tabs will have to be made. I will also move to 1/2 inch front baffle. The 3/4 when surface mounting seems like it is too thick.

Quick question, when mounting your drivers did you screw directly into the front baffle with wood screws or did you use bolts and nuts? Thank you for all the great information!!!
 
I used #6 size wood screws and mounted my SB65s directly to the 0.5 inch thick MDF baffle. No mounting with bolts/nuts for me.

The mounting holes are on 2 inch spacing around each SB65 driver.

I used 2 5/8" (66.675 mm) center to center mounting for my Modified CBT24 arrays.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.