Spreadsheet for Folded Horn Layouts...

I finally got a chance to look at this. Yep, it seems like an error. The height of the back panel was being calculated incorrectly. I did a quick fix (see attached), but came across another minor issue - when you enter dimensions, they're used as external dimensions rather than internal ones (so the external dimensions will be larger in each direction by 2*panel thickness).

I've decided to rewrite that section of the workbook anyway, and will upload the complete version when finished (a few weeks, as I'm converting it from a stepped to a continously expanding horn, and including some calculations for bracing).

You KNOW I didn't read this! I just clicked on the attachment.
 
I know I'm being anal, but why is this TQWP not 825.8cm2 (16" x 8") thru the whole 304.80cm (10ft) pipe? Under Initial Parameters, h should = 81.28cm (8" x 4 x 2.54cm). Are those parameters internal or external?

Did you do the optimization routine? D should be 0, not 0.12.

I'm removing the optimization routine from the new version of the spreadsheet, because it leads to confusion. Instead the spreadsheet will spit out the equivalent S1 to S4 parameters for HornResp input.
 
h 86.99cm = D -0.10
h 87.01cm = D 0.33
h 87.00cm = D 0.12 = blue TH!

h = 81.28cm = 8" x 4 folds x 2.54cm
w = 40.64cm = 16" x 2.54cm
d = 76.20cm = 2.5ft x 4 folds x 30.48cm(12" x 2.54cm)

Why do you prefer the right fold over the left fold?
 

Attachments

  • foldjpg.jpg
    foldjpg.jpg
    207.1 KB · Views: 837
Did you do the optimization routine? D should be 0, not 0.12.

I'm removing the optimization routine from the new version of the spreadsheet, because it leads to confusion. Instead the spreadsheet will spit out the equivalent S1 to S4 parameters for HornResp input.


Well, presenting a new spreadsheet - BOXPLAN-THAM. This spreadsheet allows you to optimize the layout a "THAM" style TH before cutting any wood. The spreadsheet gives you all the panel dimensions required. Basically you enter the parameters for the desired TH, then adjust the panel sizes until predicted result (in blue) is a close match for the "guide" curve.
 

Attachments

  • boxplan-tham.zip
    30.8 KB · Views: 287
Spreadsheet for "SS15" type folds

I've attached a spreadsheet that I've put together to assist with designing the layout of THs using a folding technique that's similar to the "SS15" TH design. I used this spreadsheet to design and build a new TH that fits a bit better in my car's trunk and addressed any bugs I came across along the way, so it's pretty much bug-free at this point.

To use:

1. Enter the dimensions of the subwoofer driver in the "Driver Dimensions" table

2. Enter your preferred box dimensions in the "Box Dimensions" table. If you are countersinking the driver (which I strongly suggest for this type of build), use the "driver offset" box to enter how deep the driver is going to be embedded into the baffle. If you're going to mount it on top of the baffle, enter "0" for the driver offset.

3. The spreadsheet models a dual-expansion TH like the SS15. If you want the change in expansion to occur a bit earlier in the horn, enter "x" for the "S3 at 3rd fold" option.

4. Under the "HornResp" params section, enter starting values for S1, S5 and L12.

Now, under the "Horn Design" section, enter the thickness of each panel under the "P" column, then start fiddling around with the horn expansion options and the panel lengths (basically all of the options in blue text). Keep your eye on the "Horn Expansion" graph. Basically your goal is to end up with a horn expansion that is as smooth as possible. The spreadsheet will automatically calculate the parameters that you'll need to enter into HornResp to simulate your TH design. If you don't like the results, adjust the TH layout accordingly, following the process above.

As the process is an iterative one, it may take awhile to come up with something acceptable.

Some notes:

1. The spreadsheet uses the "Advanced Centerline" method to determine the length of the horn. The measured results of the built vs. simmed horn suggest that this method produces results that are accurate to within 3%.

2. The spreadsheet assumes butt-joints and square cuts for all panels. It also does not calculate the dimensions of any braces (I'll add that in a later version). If you want to make angled cuts and use a different joining technique, it shouldn't be that difficult to take the panel dimensions from the spreadsheet and adjust them accordingly. And you'll need to design your own braces.

3. I was a bit skeptical about using 0.5" (12mm) ply for building horns, but it seemed to work out Ok for this test build, with certain conditions, like using stringers to make sure that the ply doesn't end up splitting when screws are being driven into it!

4. From my observations after having built THs for the same driver using two folding techniques, I'd say that the "SS15"-type fold is superior to the THAM-type fold in terms of end results, but the THAM-type fold is easier to build. With the THAM-type fold, a lot of attention needs to be paid to the bracing at the S1 and S3 points to reduce panel flex - much more than in the case of the SS15-type fold. This panel flex can cause the response at the upper end of the TH's passband to sound muddy if it is not properly addressed.
 

Attachments

  • boxplan-ss-4-2.zip
    59.7 KB · Views: 179
  • 20131229-phantasmo-1.jpg
    20131229-phantasmo-1.jpg
    242.6 KB · Views: 621
  • 20131230-phantasmo-1.jpg
    20131230-phantasmo-1.jpg
    209.7 KB · Views: 742
  • 20131230-phantasmo-2.jpg
    20131230-phantasmo-2.jpg
    99 KB · Views: 591
3. I was a bit skeptical about using 0.5" (12mm) ply for building horns, but it seemed to work out Ok for this test build, with certain conditions, like using stringers to make sure that the ply doesn't end up splitting when screws are being driven into it!

It does work!

And if in doubt you can cut the head off of a finishing nail and leave the pointed side out of the chuck in a drill. Use the nail as a small diameter drill bit to make a pilot hole. IT will definitely not split then.
 
It does work!

And if in doubt you can cut the head off of a finishing nail and leave the pointed side out of the chuck in a drill. Use the nail as a small diameter drill bit to make a pilot hole. IT will definitely not split then.

Ok, I'll admit that I cheated a bit, so I still remain a bit skeptical, LOL. Only the side and internal panels for this build were done with 0.5" ply. I used 0.75" ply for the top, back and bottom panels (where I believed I'd have the most problems with panel flex), and the baffle is actually made up of two sheets of 0.5" ply bonded together. And this box is just 133 liters net, so it's not large as far as pro audio subwoofer boxes go.

I still can't imagine using just 0.5" ply alone to construct subwoofer boxes any bigger than this, especially for larger and/or more capable drivers than the Dayton PA310. And I'm still waiting for anyone who's done so to post an impedance curve for their build - that will quickly tell me whether or not panel flex is an issue :).

However, even though butt-joining definitely takes a lot longer with 0.5" ply because of the additional steps that need to be taken to avoid splitting the ply and basically just keeping the panels straight during assembly, the end result is a lighter and smaller box.
 
Panel Flex is a product of lack of proper bracing if you use 12mm material.

That's true for all material, not just 12mm material.

The removable panel on my POC #1 TH is 17mm ply. It's 45cm by 60cm and fastened to the box by about 20 screws around its perimeter. Unbraced, it flexed so badly that the expected upper peak in the impedance curve was almost eliminated. It took a wood center brace, two 3mm thick metal braces and another layer of 17mm ply bonded to its top to reduce the flexing to an acceptable amount, and even now it will vibrate off any item placed on top of it when I'm playing the TH at very loud volumes.


I make almost all my horns out of 12mm plywood. Or prototypes out of OSB. They are extremely stiff.

Stiffness is a relative term. I thought the panel I mentioned above was extremely stiff too (it feels like an 11 Kg concrete slab now), but guess what - it still flexes :(!

Finally, FWIW, lighter panels should actually be better - it should help to remove the panel resonances out of the subwoofer's passband. However I remain a bit of a "doubting Thomas" concerning the stiffness of the thinner ply, and still want to see the impedance curves of those 12mm ply subwoofer horn builds :)
 
Some more pictures of the TH that I built using the spreadsheet.

The first picture shows a quick and dirty FR I took this afternoon. Red=measured response, brown=predicted response, blue=response with 30Hz 48db/oct and 120 Hz 48dB/oct filters applied.

The second shows its size when compared to one of my "Blastorama" tops.
 

Attachments

  • 20131231-phantasmo-1.jpg
    20131231-phantasmo-1.jpg
    144 KB · Views: 555
  • 20131231-phantasmo-2.jpg
    20131231-phantasmo-2.jpg
    197 KB · Views: 497
That's true for all material, not just 12mm material.

Finally, FWIW, lighter panels should actually be better - it should help to remove the panel resonances out of the subwoofer's passband.

Pretty much, even concrete flexes a bit if excited with a low enough frequency.

True, though it also absorbs acoustic efficiency, so only a good plan if the alignment is under-damped enough, which means it's larger, ergo all things considered, better to go stiffer via heavily braced light construction with 'light' being relative to the acoustic pressure it's likely to 'feel'.

Your experience, measured results Vs simmed is clear proof as to how much acoustic pressure even a small [T]H can modulate/control a driver at low power.

GM
 
How does it sound in the ride????

Hah, it wasn't designed to be used while *in* the ride, LOL. Once the car's rear door is closed, the mouth is effectively blocked, and if it's turned around, the port will be partially blocked by the rear seats.

I suppose I could fold one seat down and try it out - given the measured frequency response, and my experience with running the previous TH in the same car, I expect that it will sound like a very efficient 38 Hz vented box.