[Kubbie]

Winter thaw: SI HT-18 "for-fun" build while I wait for a UXL. Used HST-18 instead.

I am new to the DIY subwoofer concept, so I really want to build a sub myself. I have spent the last few weeks just reading up, so now ready to start a thread. Although I have funds to buy $1-$2k sub, I am really just doing this for the fun of it - I like to find new hobbies and this seems right up my alley.

I placed an order for an HT-18D2 a few days ago, and am on the waiting list for a UXL. (I think that UXL will get a horn build down the road), this SI sub will be my trial run.

I just downloaded WinISD and am mucking my way through it.
But in the true spirit of DIY (at least for me), I would like to design the enclosure, instead of a tested/tried 'n true design like a Marty or a flat-pack. I picked up a few 5'x5' sheets of 18mm (3/4") Baltic Birch (sitting in my van for now.)

Wish list:
- SI HT-18D2 Driver
- Approx 28" (H) x 24" (W) x 35" (L - or deep) format
- I have decided to go slot ported, front facing driver /port. (Driver/port on 28" x 24" front face)
- 18 to 20 hz tune?
- For power, I plan to buy a iNuke3KDSP.

I hope I can ask some questions here, I know I'm not breaking any new ground. Just a guy new at this.

1. Can someone confirm my parameters are good? I don't want to get too far modeling if I am off somewhere.

2. For modeling, what is the driver volume? From reading here, I learned I have to subtract the port, driver and bracing volume to get the net internal. I know in some cases it is negligible, but I would like to do the math 1st (again, for fun). I saw on their site Displacement: 0.25 ft^3, is this it?

3. Trying to understand end correction in WinISD under the Vents section. Will I get I get better performance if none of the vent walls share the enclosure walls?

4. Port - if tuning determines I need to make a 90 turn to fit, does anyone smooth the corner with a radius? Or am I splitting hairs and over thinking it?

5. I see most slot ports have a few vertical braces. Don't you have to treat that a 3 slot ports, not just subtracting the 1.5" for the two 3/4" supports?

I think that's enough for now. Also just downloaded Sketchup, but I'm struggling just drawing a square box, so I'm pretty far from drawing baffle cutouts, roundovers, ports or bracing... Just going to be paper and pencil drawings for now.

Rob

 

[kiwijunglist]

5. I'm sure that the math is the same for calculating both ways. You get the same end dimensions.

 

[nograveconcern]

2. yes, displacement
3. no, you get a shorter port with a shared wall so there is no benefit to not sharing a wall. This is a benefit as you want to keep port length under 36-40" to keep the resonant frequency out of the crossover range. Add 1/2 of port height to the effective length for each port end with a shared wall.
4. You can. It's easy enough to 1/2" round over the 90 degree angle to keep it from making noise. You don't have to do anything to build up the inside angle, but it's up to you.
5. Port surface area goes up and likewise friction of the air in the port so yes, it is treated as multiple ports. Subtracting 1.5" will result in a port that is too short.

This site simplifies calculating minimum port area and slot port length
http://www.decibelcar.com/apps/108-slotcalc.html

Mind the 1:8 ratio W to H ratio for port dimensions.

 

[LTD02]

"Approx 28" (H) x 24" (W) x 35" (L - or deep) format"


the thing is that all these designs that are roughly the same size and form factor will perform quite similarly.


for example, starting with something roughly like this form factor (ignore the precise measurements):

if you make the bracing 2.5" tall so that is the height of the slot port and use two spacers, and then run the slot along the bottom and then about 6" up the back, it's going to work great (port will measure about 42" down the centerline and have an effective length of somewhere around 45" or a little longer). net will be about 9 cu. ft. and tuning will be around 17hz +/- about 0.5hz.






1. Can someone confirm my parameters are good? I don't want to get too far modeling if I am off somewhere.


those are different from what I use which are from here: http://www.data-bass.com/data?page=driver&id=46



2. For modeling, what is the driver volume? From reading here, I learned I have to subtract the port, driver and bracing volume to get the net internal. I know in some cases it is negligible, but I would like to do the math 1st (again, for fun). I saw on their site Displacement: 0.25 ft^3, is this it?


for an 18", I use about 0.33, but with large ported cabs if you are off by a little, it makes no difference, so no need to be super precise.

3. Trying to understand end correction in WinISD under the Vents section. Will I get I get better performance if none of the vent walls share the enclosure walls?



nobody understands end correction. in the original marty post, I linked up several instances of various theories and measured builds. there is no putting it together. for some adding 1/2 port height to each end measurers really close. for others with slot ports, that misses the mark by about 10-15%. in my head I use 1/2 port height added to each end then an additional 0.5 hz. then calling it within +/- 0.5hz seems to be a reasonable mystery variance to note.

4. Port - if tuning determines I need to make a 90 turn to fit, does anyone smooth the corner with a radius? Or am I splitting hairs and over thinking it?


some do, some don't. I haven't seen any direct comparo (same build but rounded internal vs. not). lots of successful builds not doing it. probably more important with small ports as port velocity goes way up.

5. I see most slot ports have a few vertical braces. Don't you have to treat that a 3 slot ports, not just subtracting the 1.5" for the two 3/4" supports?



for practical purposes (after considering how the end correction changes), they are probably the same.


good luck with your project.

 

[Kubbie]

those are different from what I use which are from here: http://www.data-bass.com/data?page=driver&id=46

I changed the driver parameters in WinISD, so I am modeling the box two ways:

1. Slot port
2. 2 round 6" ports.

With limited free time, I just started to cut up some of the plywood. (Didn't really happen since I have no pics.) It's going to be a slow build. I know I am going a bit backwards by cutting the exterior up 1st without a final design.

My exterior will be is 26" (w) x 28" (h) x 35" (d)
Interior (with double 3/4" front baffle) is 24.5" x 26.5" x 32.75" = 12.3 ft³ gross.

Assume:
.33 ft³ for driver
.5 - 1.0 ft³ bracing
1.3 - 1.75 ft³ porting (round vs. slot)

So depending on design, I should have about 9 to 10.2 net volume. Box is tuned to 18 Hz. Entered 1,100 watts under the signal tab.

I prefer the look of round ports. With two 6" ports at 42" each, does this port velocity look fine? Here is my problem - the 6" flared Precision Ports are 9 1/4" outside diameter. It gets tight in the front for both the SI-18 and the two flared ports, will I be ok with noise if I just use the straight port tubes? (I read somewhere that at 6" round, the port noise becomes less of an issue.)

At 42" long, I will have to make some turns to make the round ports to fit the box, I didn't see 6" 90's on parts-express, I would have to see if I can get 6" pvc to work. If that is tough to do, I am not married to the round ports and will go slotted.

Any opinions?

 

[LTD02]

the port velocity seems ok so far. may want to check how it looks when you step up to the uxl (if you plan to put more power on the uxl).


as you note, it gets really tight trying to put the driver and two of those on the face of your cab.


not sure where one would find a 6" 90 degree bend to work with the p.e. 6" port. actually, I'm not even sure where one would find a large radius 6" pvc elbow (maybe someone else can help with that?).


even a small roundover on the front will help. turbulence comes on fairly low with no roundover. http://www.subwoofer-builder.com/port-flares.htm

 

[LTD02]

not sure how much cutting has been done, but going a little taller works with those ports. the increased volume would shorten them up too, so maybe no bend would be required simplifying things greatly.


something like this?

 

[Augerhandle]

... 3. Trying to understand end correction in WinISD under the Vents section. Will I get I get better performance if none of the vent walls share the enclosure walls?



nobody understands end correction. in the original marty post, I linked up several instances of various theories and measured builds. there is no putting it together. for some adding 1/2 port height to each end measurers really close. for others with slot ports, that misses the mark by about 10-15%. in my head I use 1/2 port height added to each end then an additional 0.5 hz. then calling it within +/- 0.5hz seems to be a reasonable mystery variance to note....

Found this in the WinISD Help Menu

Port terminology: Physical/acoustical length, end correction
If you have read articles on port design on the Internet, you have probably stumbled on various terms involving ports. This document is to explain some of those terms.


--------------------------------------------------------------------------------

Effective radius/diameter
If your port has some other shaped cross-section than circular, then we'll need to calculate effective radius or diameter. Effective radius is radius of the circle, which has same cross-section area than non-circular object. For ports, only inside area contributes to port area, not it's walls. For calculating effective port radius/diameter, use following formulas:
effective radius = sqrt(S/pi) and
effective diameter = sqrt(4·S/pi),
where sqrt means square root.

For example, you have square port which has 100 square inches of cross-section area, what is effective diameter and radius?

Now, just plugging the values to above formula we get:
radius = sqrt(100/pi) = 5.642 inches
diameter = sqrt(4·100/pi) = 11.284 inches.

Of course, diameter could have been easily obtained from radius by just multiplying radius by two.


--------------------------------------------------------------------------------

Acoustical length
First, we need to discuss what is the idea behind port, and how it works. Basically, port forms a mass-spring resonance circuit with enclosure volume compliance. Helmholtz was first to discover such a resonator, hence it was named after him. Mass in port resonates with compliance of box air. By tuning the enclosure, we chose mass in port, so that resonates at desired frequency. As our port will have some area, we can calculate, how long this "tube" should be to have desired mass. This length is so called "acoustical length" of our port.


--------------------------------------------------------------------------------

Physical length
Unfortunately, effect of the port doesn't stop at the end of our port. Moving mass extends somewhat beyond of the port. Researchers of acoustics say that radiation impedance has reactive component, mass. So if we make our port length equal to calculated acoustical length, we'll find out that tuning frequency is lower than what was expected. We can make good approximation to this extra mass by adding this radiation load mass from desired port mass and then calculating required port length.

By default, WinISD assumes that you are going to mount your port above way. And therefore gives you physical length of your port assuming mounting like in above figures. See also FAQ 3.1.


--------------------------------------------------------------------------------

End correction
End correction factor is analytically or empirically determined factor, how much port extends beyond its physical ends. For free end, end correction is 0.30665 times port effective diameter. For flanged end, more analytical expression is available, 4/(3·pi) ~= 0.42441 times port effective diameter. Flanged end is calculated assuming that tube terminates to infinite baffle. Which is not exactly true, but.. Following table summarizes various port configuration types and their total end correction factors (refer to above picture to various end types):

Summary of end correction values Port configuration end correction 1 end correction 2 Total end correction
Two free ends 0.30665 0.30665 0.6133
One flanged and one free end 0.42441 0.30665 0.7311
Two flanged ends 0.42441 0.42441 0.8488

Physical port length is obtained by subtracting port effective diameter multiplied by suitable end correction value from port's acoustical length.

It doesn't seem to elaborate on whether WinISD does this calculation internally or just spits out the correction factor for us to subtract.

 

[Kubbie]

the port velocity seems ok so far. may want to check how it looks when you step up to the uxl (if you plan to put more power on the uxl). [/URL]

UXL will get a new box, so not worried about it for this one.

as you note, it gets really tight trying to put the driver and two of those on the face of your cab. [/URL]

Yeah, I drew it out on a piece of plywood full size 18 1/2" and two 6" flared Precision Ports and it just looks bad - way to crammed together, almost touching.

not sure how much cutting has been done, but going a little taller works with those ports. the increased volume would shorten them up too, so maybe no bend would be required simplifying things greatly.

Nah, all sized and some assembly started. Box size is as listed.
My exterior will be 26" (w) x 28" (h) x 35" (d)

something like this?

Yes! that would be much better. The 6" pvc just rounded over might work. But starting to lean towards the slot port - models almost exactly the same, and the 90 degree PVC elbows make assembly and bracing a bit more work.

not sure where one would find a 6" 90 degree bend to work with the p.e. 6" port. actually, I'm not even sure where one would find a large radius 6" pvc elbow (maybe someone else can help with that?). [/URL]

A local plumbing supply house has them, or just search Amazon for "1/4 Bend, Long Sweep Elbow, 6" Hub" - run about $40 each.

 

[Kubbie]

When done, I do want the box showing wood grain on the top and sides (stained, lacquer), but the front baffle a matte black. I also want a larger round-over on the sides.

So I came up with this: I had some scrap 8/4 poplar I planed down, and cut rabbets the depth of the ply. This allowed me to glue, screw straight through, as well as pocket screw from the back so I don't have any exterior nail or screw holes, and no expose ply when I round over. (pic is top panel to one of the sides)

It may look fragile, but with back/front on and bracing, it will be solid.

 

[More is Better]

A single 8" port should work better than 2x 6" ports even though it is 10% less area. The bigger the tube the more air you can move without port compression. Round ports are ideal for airflow but it's harder to implement than a slot port. You want to be interior port opening to be 1 port diameter or more away from any box walls too.

 

[Kubbie]

Found this in the WinISD Help Menu

...stuff, science...stuff...math, stuff, stuff...

Thanks,.I did not get the time to digest this yet, so that's all I hear for now. But I want to learn why I'm doing something.

A single 8" port should work better than 2x 6" ports even though it is 10% less area. The bigger the tube the more air you can move without port compression. Round ports are ideal for airflow but it's harder to implement than a slot port.

With my limited knowledge on this stuff, I would sound like Cliff Clavin on Cheers describing a picture of a double helix and the structure of deoxyribonucleic acid, then turning it sideways to explain rocket propulsion.

Add my extensive study of the post-graduate level fluid dynamics (mostly from the physics I learn from watching The Big Bang Theory), I did some Reynolds decomposition modeling and found that a cylindrical/tube port is most effective at producing a laminar flow, while the slot port introduces eddies and recirculation, resulting in a more turbulent flow.





No. I just like the look of round ports.

 

[More is Better]

found that a cylindrical/tube port is most effective at producing a laminar flow, while the slot port introduces eddies and recirculation, resulting in a more turbulent flow.

No. I just like the look of round ports.

That bit is correct, and 1 round port is more effective than 2 round ports that have the same total area as a cross section of the port/ports.


Even simpler... it's easier to blow through a 3" pipe than a 3" handful of drinking straws.

 

[Kubbie]

I haven't done any more work on the box, other than make a quick sketch.
I think this will work:
10.5 Cft net,
18 Hz tune,
3x 4" Precision ports, 27" each.
Signal 1100 W, H.P at 18hz

Seems to keep velocity reasonable.

 

[Kubbie]

I am split between two 6" round and three 4" round ports.
WinISD looks the same except port velocity.

The 3 port would look somewhat like the SVS PB-13 below.

Am I underported for an 18" driver by choosing the three 4" ports?

 

[More is Better]

That bit is correct, and 1 round port is more effective than 2 round ports that have the same total area as a cross section of the port/ports.


Even simpler... it's easier to blow through a 3" pipe than a 3" handful of drinking straws.

this




3x4" ports is 37.7 sqin
2x6" ports is 56.5
1x8" port is 50.3


Out of these three options, I would pick the 8"


http://www.subwoofer-builder.com/porting/index.htm


Would really want the 10" at 78.5sqin. or 12" at 113.1sqin

 

[Kubbie]

3x4" ports is 37.7 sqin
2x6" ports is 56.5
1x8" port is 50.3


Out of these three options, I would pick the 8"
...

Would really want the 10" at 78.5sqin. or 12" at 113.1sqin

12" gives me a 80" port, and 1st port res at 84 Hz.
Even a 10" is 54" long, and given my box dimensions makes it almost not possible without some funky design that has part of the port outside of the box. I'm trying to model and strike a balance between a reasonable port length and port speed.

My real question is: where do I find port compression problems in WinISD? I am new to this and trying to learn.

 

[dtsdig]

When done, I do want the box showing wood grain on the top and sides (stained, lacquer), but the front baffle a matte black. I also want a larger round-over on the sides.

So I came up with this: I had some scrap 8/4 poplar I planed down, and cut rabbets the depth of the ply. This allowed me to glue, screw straight through, as well as pocket screw from the back so I don't have any exterior nail or screw holes, and no expose ply when I round over. (pic is top panel to one of the sides)

It may look fragile, but with back/front on and bracing, it will be solid.

Nice job with your corners there! That was good thinking.

 

[Augerhandle]

12" gives me a 80" port, and 1st port res at 84 Hz.
Even a 10" is 54" long, and given my box dimensions makes it almost not possible without some funky design that has part of the port outside of the box. I'm trying to model and strike a balance between a reasonable port length and port speed.

My real question is: where do I find port compression problems in WinISD? I am new to this and trying to learn.

http://www.subwoofer-builder.com/flare-it.htm

 

[More is Better]

Winisd doesn't show you the problems. It will let you model a horrible sounding sub.


I modeled the build you are doing last night, If you can stretch one dimension to around 40"-42" a single 8" port looks like the ticket. You want to keep the end of the port 1 port diameter away from internal box walls.


If you want the port to get shorter, make the box bigger....go try it.


That last link is how you build the port flares too. As big as you can doughnut flare.