First subwoofer build - feedback needed

[66o_]

Hey!

I'm planning my first speaker build and I wanted to get some feedback on my plan. First a little bit of context.

The backstory
I have a pair of Genelec 8340 in ~12m2 room. The room has a little bit of treatment and my desk and speakers are not placed optimally - nothing I can do about it for now. As a result frequency responce of my genelecs has a very severe dip at 50Hz.

Since there's nothing I can do about the room placement I decided to throw money at the problem and got a 7350A. It's placed in the corner and after calibration it gives me pretty much flat response all the way down to low 20Hz. Nice.

Now what I quickly realized is that I don't really want my system go down to 20Hz - the room is not correctly treated and the decay at the very low end is huge. I spent a bit of time playing with an EQ and I figured out I'm the happiest with the entire system high passed at around 35Hz. This gives me decent low end without the room ringing like crazy.

I always wanted to get into speaker building, I'm a fairly competent woodworker and I'm comfortable with electronics. I still have a couple of weeks to return the 7350A. Looks like I finally found myself a project.

Subwoofer plan
For the first build I want to keep things simple and stick to the sealed box. Design requirements:
- flat down to 35Hz after room gain and EQ
- as thin as possible for an easy placement in the corner

For the driver SB29SWNRX-S75-6 looks pretty good: Qts 0.29, Fs 20 Hz, Vas 66 L. Qts is below 0.4 which from what I've read makes it suitable for the sealed enclosure. I don't think alignment (Qtc) matters that much in my case since the room will be ringing much longer than the driver but Qtc of 0.7 gives me a tiny 10L box, I'm more than happy to bump it to 20L for Qtc of 0.6 and better Vas/Vb ratio. All together I should get Fc ≈ 41 Hz, F3 ≈ 42 Hz which I hope after room gain (corner placement) and EQ will be enough to go down to 35Hz at modest SPL.

20L means I can get the box thin as well. The driver is 133mm deep, I'm thinking ~160mm internal depth - is that enough? 15mm birch plywood (I have a few sheets), double for the front panel and one or two braces.

I'm planning to power it with ~200W IcePower or Hypex module, whatever I find at a better price. EQ and crossover will be done directly in my RME interface.

How's my plan looking? Anything I got obviously wrong? As I mentioned this is my first build so any feedback more than welcome.

 

[SigbergAudio]

You don’t need a double baffle, add more braces instead. And this will be relatively rigid to begin with since it’s quite small. So if you make it thin, four of your six panels will be pretty small.

Internal depth is fine probably, I’d leave a distance of at least twice the diameter of the vent in the pole piece.

 

[Elgrosso]

Seems like a pretty good plan. Maybe you could measure the current sub near-field and then in different places to get an idea of the room gain per spot, before deciding for the final box Q/volume.
And if you have multiple placement options, with the money saved could you even build 2 or 3?
Great way to smooth the decay.

 

[66o_]

You don’t need a double baffle, add more braces instead. And this will be relatively rigid to begin with since it’s quite small. So if you make it thin, four of your six panels will be pretty small.

Internal depth is fine probably, I’d leave a distance of at least twice the diameter of the vent in the pole piece.

Great tips, thank you!

Seems like a pretty good plan. Maybe you could measure the current sub near-field and then in different places to get an idea of the room gain per spot, before deciding for the final box Q/volume.
And if you have multiple placement options, with the money saved could you even build 2 or 3?
Great way to smooth the decay.

Can you elaborate please? How room gain per spot will influence the final Q of the subwoofer?

 

[Flaesh]

I'll add an illustration to my thread now.

 

[Elgrosso]

Can you elaborate please? How room gain per spot will influence the final Q of the subwoofer?

Sure, the idea is to optimize it for the room & placement more than for an universal optimum response.
Depending of the room size and final spot the response will gain a lot, could be too much, or it could just fill a missing low end of a too small box for ex.
Evaluating the room / boundary gains would give an idea of the response to simulate and then build.
(With this driver maybe there’s not a huge difference between 10 and 20L, but that’s the idea)

So measuring one sub near-field or outside to get its response.
Then measuring it in each potential place in the room, to read the differences.
That’s the variables you can use in simulation for the sub target to get flat (or other) once in the room.

For the multisubs you probably know, it could tame the 30hz mode, if you need that low.
If not it would still smooth the decay for the entire sub range.

 

[66o_]

So measuring one sub near-field or outside to get its response.
Then measuring it in each potential place in the room, to read the differences.
That’s the variables you can use in simulation for the sub target to get flat (or other) once in the room.

Got it! All right, I think I have all I need to start the project. I'll order parts and report back once I start putting this thing together

 

[CDMC]

Did you model the power limits of the sub? Make sure your amp has enough to reach the mechanical limits of the sub, or you are leaving capability on the table. I am of the camp, if you can do 2-3 small subs, a lot of your issues may resolve, as a lot of the slow decay you are seeing is because of your room modes being from a single sub and listening location. Spread them out with multiple subs and the decay issue lessens.

 

[66o_]

Internal depth is fine probably, I’d leave a distance of at least twice the diameter of the vent in the pole piece.

@Sigberg Audio the driver is finally here and I measured the diameter of the vent - it's 55mm. Double that and if I follow your recommendation it'd have to add 110mm to the depth of the subwoofer which makes a bit deeper than I hoped.

You mind explaining a bit more where that rule of thumb comes from? What am I trading off by leaving less space behind the driver?

BTW I had a look at your Inkognito subwoofer - very nice construction, I may draw some inspiration from it

Did you model the power limits of the sub? Make sure your amp has enough to reach the mechanical limits of the sub, or you are leaving capability on the table. I am of the camp, if you can do 2-3 small subs, a lot of your issues may resolve, as a lot of the slow decay you are seeing is because of your room modes being from a single sub and listening location. Spread them out with multiple subs and the decay issue lessens.

It's my first speaker build ever so I'm building one sub for now but if that goes well I'll likely do one or two more. I've got a ICEpower 300AS1 to pair with SB29SWNRX-S75-6 (200W max power rating) which should be more than enough. The amp was pretty much the same price as the 200W version so I thought the extra power won't hurt (unless I forget to set the limiter and blow the driver )

 

[SigbergAudio]

@Sigberg Audio he driver is finally here and I measured the diameter of the vent - it's 55mm. Double that and if I follow your recommendation it'd have to add 110mm to the depth of the subwoofer which makes a bit deeper than I hoped.

You mind explaining a bit more where that rule of thumb comes from? What am I trading off by leaving less space behind the driver?

That's a very large vent for a 10" driver. So not really a driver that's super suited for your purposes in that regard. It would be better to have one that didn't have a vented pole piece at all. But it seems to be a flared opening so the actual diameter further in isn't that wide. You don't need that much distance, but I would consider at least equal distance as the vent diameter, so 55mm. Still a bit more than you wanted I guess.

 

[66o_]

That's how the vent looks like. I measured 55mm inside the vent. Too bad the pole vent diameter wasn't anywhere in the datasheet.

I think I'll build a test cab with 60mm clearance behind the driver which should still give me ~20cm deep box (mouthing depth for this driver is only 120mm), crank it loud and listen to any chuffing. I'll report back once that's done.

Thanks for all the tips so far!

 

[66o_]

I made a bit of progress on the subwoofer. The mdf panels are cut to size and I managed to get the opening for the driver cut using my trim router. Sketchy but it got the job done which is what matters.

I'm using 16mm mdf for the test cab and the next step is to decide on bracing. Any advice on this front? There isn't that much space in the enclosure and the panels are pretty small so I think the box should be fairly stiff as is.

For the glue up - is regular wood glue ok? Should I seal the box after it's glued with anything?

 

[Elgrosso]

Nice, box is finally not that deep, you could even mount it flush.
I’ve made a shallow enclosure too, leaving 2” to breath and it’s fine, but it’s a woofer not a sub so it does not need as much.

Maybe one brace in the longitudinal axis, with a large cut for the woofer?
Not exactly in the center to spread resonances, and not to cover the the vent.
Should be enough with no panels larger than 20cm unsupported.
It could be fine as it is too, in fact if you plan to build 3 you could use this as a test.

Wood glue is perfect, if your cuts are not great (but they look good) you can try PU glue as it will fill some small gaps while expanding a bit. For piece of mind you can apply a bit of glue with the finger on all interior edges.

 

[fivenp]

Any advice on this front? There isn't that much space in the enclosure and the panels are pretty small so I think the box should be fairly stiff as is.

Sigberg Audio Inkognito: Building a "design" subwoofer that actually sounds good

@carlob mentioned that my subwoofers looked interesting and suggested I created a thread to share a bit more information, including of course objective measurements. I'll start out with a more subjective / informative post to share a bit about the product and concept, then I'll add an additional...

www.audiosciencereview.com

This could give some inspiration regarding bracing approaches in a flat enclosure

 

[66o_]

Nice, box is finally not that deep, you could even mount it flush.
I’ve made a shallow enclosure too, leaving 2” to breath and it’s fine, but it’s a woofer not a sub so it does not need as much.

Maybe one brace in the longitudinal axis, with a large cut for the woofer?
Not exactly in the center to spread resonances, and not to cover the the vent.
Should be enough with no panels larger than 20cm unsupported.
It could be fine as it is too, in fact if you plan to build 3 you could use this as a test.

Wood glue is perfect, if your cuts are not great (but they look good) you can try PU glue as it will fill some small gaps while expanding a bit. For piece of mind you can apply a bit of glue with the finger on all interior edges.

Thanks @Elgrosso. I ended up sealing all the joints on the inside with pva glue as per your advice. For bracing I decided to use a couple of offcuts to stiffen the baffle right below the driver (see the pics). The box is done now and looking pretty good if I may say so.

I also just had an idea: I can prepare a bunch of 4-8mm thick pieces of plywood and attach them one by one behind the driver. That would effectively decrease the distance behind the driver without changing any other params of the enclosure in any significant way. So that way I could determine the real minimum distance I need.

This could give some inspiration regarding bracing approaches in a flat enclosure

Beautiful build. I'm not sure I need corner braces (I can't probably fit them in the upper corners) but that brace with circular part looks very clever. Not sure yet how I'd cut it using a router but I'll experiment a bit.

I also realised I forgot to order threaded inserts and any form of speaker posts so I'll have to wait few more days before hearing my sub for the first time. Bummer.

 

[Elgrosso]

Cool it is mounted flush, and you even rounded the edges.
Do you have a DATS to measure the impedance of the system? It would help you to fine-tune depth, bracing and filling.
Filling not absolutely necessary but it would smooth out the traces, and decrease the Qtc a bit.
But all this depends of how much you will push it of course, just as is would already be good I’m sure (most commercial ones are not as solid).

Threaded insert not needed in my view, if you don’t plan to unscrew it 20 times direct into the mdf is fine.
For the speaker posts, maybe some recessed ones? Always a pain to have posts protruding in the wrong spot when you want to push the box on the wall. (also just a small hole for the wires + filling with hot glue is good for a prototype)

 

[66o_]

Threaded insert not needed in my view, if you don’t plan to unscrew it 20 times direct into the mdf is fine.

Oh I wish I listened to you. I've got a bag of threaded inserts today and decided to give them a go. What a nightmare. The mdf bulged around the insert when I installed it and the entire thing looks less than solid.

I've installed 4 before giving up. I'm going to get some t-nuts next and try them on the other 4 mounting holes. So glad I'm discovering all of this on a test cab

Also note to self: install t-nuts before gluing up the box next time.

 

[SigbergAudio]

Beautiful build. I'm not sure I need corner braces (I can't probably fit them in the upper corners) but that brace with circular part looks very clever. Not sure yet how I'd cut it using a router but I'll experiment a bit.

Those pictures are from 2020 I think, so this doesn't look exactly the same anymore, but it worked quite well back then too. You get a quite rigid cabinet by default when 6 panels are quite small as well.

 

[66o_]

T-nuts for the win! Hammering them in was no longer possible because the box is already glued up so I pressed them using a normal woodworking clamp. So much easier and stress free compared to those bloody threaded inserts. The driver is properly mounted now.

Next job: connector. I'll go for a speakon. I wasn't sure if I wanted to build my own faceplate amp out of the ICEPower module but I finally decided against it. Turns out ICEPower 300as1 can power another 300W amp module, add a dsp board and I'll have a complete solution for two subs in one enclosure. Also this will be a cool project on its own.

Those pictures are from 2020 I think, so this doesn't look exactly the same anymore, but it worked quite well back then too. You get a quite rigid cabinet by default when 6 panels are quite small as well.

That's my not-backed-by-science feeling too. The largest panels are 33x45cm, seem stiff enough event at 16mm thickness. I like the idea of regular thickness front panel backed with a circular brace with supports like in your design, I may give it a try in the final version of the speaker.

 

[66o_]

It's making noise! I finally cobbled together a few cables, hooked it up to my soundcard and did a few measurements (in room) in 0-100Hz range using REW. Then applied 3 filters to EQ out the worst room modes and got this response in my listening position.

Then I took another, full range measurement to figure out how to match levels of the sub and my main speakers and finally measured this:

I'm sure there's a lot of room for improvement but I'm quite happy already. Now that I've got a calibrated mic I'd like to measure the sub alone, but I need to do more reading about how it's done properly.