DIY Music (non-HT) subwoofer conundrum

haraldo said:

Well looking at these measurements of one PB13-Ultra measured in a wide-open-space...
Darn if there's very much compression going on here.
And < 10% distortion at 105db @ 20 Hz
http://www.avtalk.co.uk/forum/index.php?t=msg&th=22950&start=0&rid=0&SQ=0

If the Kappa Perfect 12 VQ in the ported cabinet is even better than this, I'm stunned.....



Thanks!, but still probably easier than some of the very complex sub cabinets.....

Click to expand...

The PB13 has about a 17 percent larger surface area(13" driver vs. 12"), but a lesser port cross section. You can see an almost 6dB(that's a 50 percent output reduction) loss at the highest SPL trace around the PB13's port tuning. The 12" Kappa VQ enclosure I recommend will likely output at least equal output, if not more, around the 17-25Hz range. Though I expect above this range, that the PB13 would output a few more dB. The Kappa Perfect enclosure design that I recommend(and that avaserfi used) had a port cross section area that is about 30 percent larger in cross section area as compared to the PB13 Ultra, while the cone area of the Kappa Perfect is smaller - the total ratio of cone area to port cross section area is almost 50 percent more as compared to the PB13 Ultra. This will lead to substantially less port compression(I estimate barely any compression around port tuning). I can not comment on distortion, as the only distortion measurement of the Kappa Perfect drivers by 3rd party so far have been in sealed enclosures which cause higher distortion and far less LF output. But, when compared to the leading comparable high end sealed units, the Kappa Perfect trumped the Velodyne DD-12 in both distortion and ouput level - and nearly matched the DD-15. The DD line uses extremely high quality drivers coupled with servo feedback to reduce distortion, and the passive motor on the Kappa Perfects still produced overall lower distortion. However, these subs are used in small sealed cabinets, causing substantial SPL loss and increased distortion. Avaserfi already reported driving 20Hz up to 110dB as measured at the listening position, with no audible(to him) distortion. I also specified far more effective internal acoustic dampening material in the Kappa Perfect sub-woofer as is used normally, so that the unit can be used at higher frequency bands without coloration occurring, as well as extra bracing as compared to what is used standard, so that panel resonances would be shifted somewhat higher in frequency.

-Chris

Change of plans

WmAx,
You have thoroughly convinced me to use a ported system. That now leads me to more questions, I hope you dont mind!

I was thinking of using two ports for aesthetics, how does the math work on that? Do I add the diameters of the cross-sections together, or is it the length of the two ports I add together? Or am I completely misguided? If two ports introduces too much complexity I can definitely go down to one port.

I would like to just buy the flared port tubes off of PartsExpress if possible, I'm not that great of a carpenter and would like to keep my box as simple as feasibily possible, without damaging quality.

If you want go 'simple' build without sacrificing quality, it won't be by using tube ports. The tube ports will have to be much shorter than the slot port design I specified, there by severely limiting port cross section area. This will result in severe output loss/compression around the tuning frequency of the system. If you want to build simple, but not sacrifice quality, you will need to use passive radiators in place of the ports. A passive radiator serves the same function as a port. You will need to use a very long stroke 15" PR(not a normal grade one) or 2 x 12" PRs to not cause compression with the Kappa Perfect driver. This is going to add at least a hundred dollars plus shipping to cost of each cabinet. If you want some suggestions for PRs, let me know.

-Chris

Port

WmAx,
I'll attempt the rectangular slot port, that shouldn't be too difficult with sufficient planning. I actually have a relative who is a cabinet maker, who I will finalize my plans with later and possibly work with to make sure I dont mess up too badly.

As I have seen in Avaserfi's project he chose the MidQ insert at 3.95 ft^3, with a slotted port equivalent of 12.5" x 3" x 55.5". Would this be the optimum volume? I will do something very similar if so.

My knowledge of rectangular ports is very lacking. Im doing some Googling and other research here to understand it better.
My first problem, is how do I measure out the 55 inches? I have prepared a diagram to show you what I was thinking. Do you measure along these red lines? Or do you measure directly along the edges?
As you can probably tell I have a little ways to go.

(PS: Im only going to be building one subwoofer)

Headfoot said:

WmAx,
I'll attempt the rectangular slot port, that shouldn't be too difficult with sufficient planning. I actually have a relative who is a cabinet maker, who I will finalize my plans with later and possibly work with to make sure I dont mess up too badly.

As I have seen in Avaserfi's project he chose the MidQ insert at 3.95 ft^3, with a slotted port equivalent of 12.5" x 3" x 55.5". Would this be the optimum volume? I will do something very similar if so.

My knowledge of rectangular ports is very lacking. Im doing some Googling and other research here to understand it better.
My first problem, is how do I measure out the 55 inches? I have prepared a diagram to show you what I was thinking. Do you measure along these red lines? Or do you measure directly along the edges?
As you can probably tell I have a little ways to go.

(PS: Im only going to be building one subwoofer)

Click to expand...

Measuring along the red line would be correct. Be sure that you account for the displacement of the vent and the driver in your design.

Good suggestion, I hadn't factored in the size of the driver (but I did remember the port)!

My box will have to be 5.26 ft^3 if you include both the port and the basket displacement.

If you dont mind me asking, what is the relationship between the tuning frequency of the port and box volume? Do the resonant frequency of the box and port have to match up? Also, if by having a larger sized port (Width x Height), would I be able to cut down on length?

(Also WmAx, if you happen to know some good PR's offhand that would be nice, but dont go and spend any time over it, PR's would be my backup plan)

Headfoot said:

Good suggestion, I hadn't factored in the size of the driver (but I did remember the port)!

My box will have to be 5.26 ft^3 if you include both the port and the basket displacement.

If you dont mind me asking, what is the relationship between the tuning frequency of the port and box volume? Do the resonant frequency of the box and port have to match up? Also, if by having a larger sized port (Width x Height), would I be able to cut down on length?

(Also WmAx, if you happen to know some good PR's offhand that would be nice, but dont go and spend any time over it, PR's would be my backup plan)

Click to expand...

The larger the port surface area, the longer the port gets and vice versa. Increasing port surface area without increasing length will increase the tuning frequency of the enclosure. Decreasing the port surface area without decreasing length will lower the tuning frequency.

Headfoot said:

Good suggestion, I hadn't factored in the size of the driver (but I did remember the port)!

My box will have to be 5.26 ft^3 if you include both the port and the basket displacement.

If you dont mind me asking, what is the relationship between the tuning frequency of the port and box volume? Do the resonant frequency of the box and port have to match up? Also, if by having a larger sized port (Width x Height), would I be able to cut down on length?

(Also WmAx, if you happen to know some good PR's offhand that would be nice, but dont go and spend any time over it, PR's would be my backup plan)

Click to expand...

There is no particular relationship between tuning frequency and port resonant frequency. The mass of the air in the port is the weight on the end of the spring. The F3 is the combination of the driver resonance and and box resonance, that produces the lowest F3 with minimal ripple. The port is an unstopped open pipe whose resonant frequency is the speed of sound divided by twice the length of the port. The sub needs to be 24 db down at port resonance, or the port resonance will color the sound, as the port will speak as on open organ pipe.

I would counsel against a passive radiator. If you want a back up I would use the low Q insert and the alignment I gave you, and use standard tube vents.

However if you have good facilities the mid Q insert with the slot vent should give you a good sub.

Thank you both very much! With these last few pieces of information I feel like something just "clicked".

I began designing the box in a CAD program the other day and then writing it out on paper, and I ran into difficulty with trying to manage all the variables correctly, as my knowledge was more limited at the time. Now with this new information and a good deal of Googling, I feel a little more secure on slot ports. I've decided to restart from the ground up.

So far I have decided:
Total Box: Width will be (Slot width + 2*(thickness of board or 3/4"))
Depth will be about 14"
Height will be determined after doing the rest of the mathematics.
Box volume will be 5.26 ft^3, although I still need to compensate for the volume of the bracing, as I have not sat down and drafted the bracing yet. In the end we can probably estimate 5.3 to 5.4 ft^3 minus port, basket, bracing, and the back of the plate amplifier to equal 3.95.

Now here is the (hopefully) final problem I face: I'm not sure if my current tuning frequency and box volume is optimum for my purposes of pure music reproduction. (3.95 ft^3 and 19.5 hz). It was simply what I could find existing material on.

The tube port configuration you gave me earlier is now the official backup plan. (Great suggestion by the way!)
========================================
I would like to sincerely thank all of you for the help, you've been extremely patient and helpful. I can't believe I've come from only knowing that speakers make sound to building my own subwoofer. It's really a great feeling and it would be a very distant possibility if it weren't for all of your input.
========================================
And finally, I'm going to get my hands on a relatively nice digital camera to take pictures of the build. Also the build is planned to go down on the end of the week of May 19th. Im getting my wisdom teeth removed earlier in the week, making that time to polish off the designs.
PS: Im not sure if math and Vicodin will go together too well. I'll keep you posted

WmAx said:

If you want go 'simple' build without sacrificing quality, it won't be by using tube ports. The tube ports will have to be much shorter than the slot port design I specified, there by severely limiting port cross section area. This will result in severe output loss/compression around the tuning frequency of the system. If you want to build simple, but not sacrifice quality, you will need to use passive radiators in place of the ports. A passive radiator serves the same function as a port. You will need to use a very long stroke 15" PR(not a normal grade one) or 2 x 12" PRs to not cause compression with the Kappa Perfect driver. This is going to add at least a hundred dollars plus shipping to cost of each cabinet. If you want some suggestions for PRs, let me know.

-Chris

Click to expand...

Will a properly designed ported system outperform the one with passive radiators? both being designed close to optimum...

haraldo said:

Will a properly designed ported system outperform the one with passive radiators? both being designed close to optimum...

Click to expand...

No. Both being designed close to optimum, they will perform nearly identical. However, some drivers require a port of such long length in combination with high cross section area, that the right port would result in a resonance in the passband. In these select cases, a passive radiator would outperform the port system.

-Chris

So that means building an optimal subwoofer with passive radiators are basically easier (and more expensive) than a ported system.

You would choose to build the ported system because it's not that expensive, is that right?

haraldo said:

So that means building an optimal subwoofer with passive radiators are basically easier (and more expensive) than a ported system.

You would choose to build the ported system because it's not that expensive, is that right?

Click to expand...

That is one reason(lower cost). The other reason is that I prefer the look of a slot port to an additional cone. However, if you mount the PR on the bottom, cosmetic appeal would be a non-issue.

-Chris

I definitely appreciate the lower cost of the slot port.

CAD Format

I am going to "build" the box in a CAD program tommorrow, and it can output to a huge number of formats, more than it would be reasonable to list. I would like to post it here to make sure that everything is gravy on the build. I have access to a full version of AutoCAD to print the final version with.

Which (if any) format would you all prefer?

From browsing the other threads I have noticed that the size of this slot port will cause it to resonate at ~120hz. (http://forums.audioholics.com/forums/showthread.php?t=38634&page=5)

The amp I am going to purchase has a built in crossover. It can be adjusted from 40hz to 120hz and it is 4th order (24db/octave). It also has a subsonic filter that can be either 12,16,20,25hz though this is unrelated to the port resonance. Will this built in crossover function well enough to eliminate the resonace?

Also this amp should be able to remove the peak, it has a notched parametric eq. In this box though, where would the peak move to? My proficiency in WinISD is not great.

So -- Will this crossover be sufficient, and where will my peak move to in a box with the same volume and port size as Avaserfi's (3.95 ft^3, 12.5" x 3" x 55")

Headfoot said:

From browsing the other threads I have noticed that the size of this slot port will cause it to resonate at ~120hz. (http://forums.audioholics.com/forums/showthread.php?t=38634&page=5)

The amp I am going to purchase has a built in crossover. It can be adjusted from 40hz to 120hz and it is 4th order (24db/octave). It also has a subsonic filter that can be either 12,16,20,25hz though this is unrelated to the port resonance. Will this built in crossover function well enough to eliminate the resonace?

Also this amp should be able to remove the peak, it has a notched parametric eq. In this box though, where would the peak move to? My proficiency in WinISD is not great.

So -- Will this crossover be sufficient, and where will my peak move to in a box with the same volume and port size as Avaserfi's (3.95 ft^3, 12.5" x 3" x 55")

Click to expand...

If your port is 55 inches long, then the port resonance will be 120 Hz. So the driver needs to be 24db down at 120 Hz. So if your fourth order crossover is set to 60Hz, the driver will be 24 db down at 120 Hz. If you crossover above 60 Hz you will have to notch out the port resonance. There will also be even harmonics of the port resonance at 240 Hz, the second harmonic, and the fourth harmonic at 480 Hz. If you suppress the fundamental the even harmonics will also be suppressed. Since the port is an open pipe it will only radiate the fundamental pipe resonance and the even harmonics.

I hope this answers your question. I think you were planning to use the mid Q alignment. Is that correct? I have not modeled that alignment, but from memory, I believe others have. If you want me to model that alignment I can.

Thank you!

You have answered my question fully, and I hadn't realized that the harmonics would be introduced as well. This gives me some nice options. If (when I find the model of this) there are no significant peaks then I could cross it over at 80 and PEQ -12 dB at 120hz, letting me crossover higher if I want that. Otherwise I can crossover at 60hz as you suggested and then use the PEQ to remove any sort of room gain I may find.

I believe its been modeled before... I'll do a search and edit this post accordingly. Otherwise if it isnt too much trouble (I've already made you do a whole ton of models! ) I would appreciate one specifically on this box.

EDIT:
Here's the only model I could find with just a real quick search (Its almost 1:00 am!), from Avaserfi's thread on his build. Although this doesnt specify the other goodies, like port velocity, SPL @ 400.
http://forums.audioholics.com/forums/showthread.php?t=38854&highlight=avaserfi

EDIT 2:
Volume computations are almost complete... I'm going to draw up the bracing this weekend or early next week, build the box in the CAD program and see what flaws we can find and fix.

My modeling of Infinity Kappa perfect 12 VQ mid insert vented with slot vent

Here is my modeling of the Kappa perfect 12 VQ mid insert vented with slot vent. I have played with this to get the smoothest response and have a shorter slot vent with a length of only 50.19 inches. This gets your port resonance up to 132 Hz. Your vent air velocity is only 12m/sec
I'm pretty happy with this alignment. See what you think

Thank you! This will be even better than before! I'll re-write up the new information sometime early this week as soon as I have time to work on it.

This weekend I spoke with my relative who is a cabinet maker, and we talked the plans over. We came into a question --

What kind/brand of sealant would be best for sealing the inside of the box (since it is made out of MDF)