Help designing my first sub

[dtb]

Hi guys,

I've been spending a lot of time reading and playing with WinISD but I need some pointers before I can really go any further.

My setup is quite basic/modest I guess, in my lounge I have a htpc with soundcard going to a 2ch gainclone amp. This feeds some some two way floorstander speakers, 6" woofers, 36L transmission line design, rated about 90db and down to about 35Hz +/- 2db. Use is probably 50% regular tv, 30% movies, 20% music.

I have always been quite happy with the bass output but have occasionally thought about adding a sub to fill in the lower frequencies, I expect it would help with music a little and add quite a bit to movies.

So far I've mostly been modelling along the lines of a single 12in DA DCS305-4 with the DA SPA250DSP although I'm starting to lean towards the Reference RSS315HF-4. I'm thinking of the DSP plate amp as I figure it will be useful as I don't have a receiver so it will let me do things like EQ the sub and highpass the floorstanders. Does this make sense and is it a worthwhile consideration?

I don't have much experience with subs and hooking them up, if I decided to build a second identical sub is that still possible to hook everything up correctly with this amp. I'm guessing the RCA cable would be split into each sub and then the HF out from one sub would go to the 2ch amp for the floorstanders?

This amp's subsonic filter has a minimum at 25Hz vs the normal 250W plate amp at 20Hz. When I'm modelling and add these filters the 25Hz has a much more marked effect on the frequency graph leading me to think it is much less desirable. Is this a significant concern?

When modelling should I be aiming for the flat frequency response first and then add the subsonic filter and leave it at that? I've noticed with the filter applied I can then play further with the box size and tuning frequency to return the graph to a more flat response, is this the right way to go?

Along similar lines I experimented with some EQ to improve the graph although this can cause issues with excursion and air velocity but is this still a worthwhile method to tune things while modelling and before building the box?

This is probably enough questions for now, I appreciate any help and answers to my questions. I guess some general advice regarding my setup would also be useful, my original thinking was having floorstanders I could potentially lowpass a single sub relatively low avoiding localisation issues while placing it out of sight behind the seating position. However maybe highpassing the floorstanders would be beneficial allowing the sub to take on more work, I see many people often recommend multiple subs so maybe a couple smaller 10" could work better.

Anyway sorry if this has turned into a wall of text and as I say any help is greatly appreciated.

[dtb]

I'm starting to lean towards the 10" RSS265HF-4, it models quite similar to the 315HF but in a smaller box but losing out a couple dB. Does this look ok? Box is 3.2cf tuned to 22Hz (in Red). I also modeled a suggestion from Chris Roemer I found of 2.25cf @ 25Hz, is there a benefit of a freq curve like Chris's I don't know about? Chris's post here http://techtalk.parts-express.com/fo...59#post1277659



My room is relatively small I guess but an L shape as below with dimensions in metres, living room feet is about 14.5 x 24.5, dining/kitchen about 21 x 9.5. Walls are brick with tiled floor. Would this room act like a small room or large, or maybe a bit of both? TV and speakers are on the wall marked 4.6.





Like I say I'm leaning towards the 10", the enclosure is smaller, it's cheaper to ship the sub + amp to Aus, it still extends pretty low and I can add a second one down the track or possibly even stretch for two now.

[wogg]

I think that RSS265HF-4 will do great. I'd also love to try that DSP plate amp. For your setup I'd run the soundcard from your HTPC straight to the plate amp, then the line outs on the plate amp to your gainclone amp. That way you can setup a proper high pass on your mains to match your sub, and do any EQ corrections for the whole room. The subsonic filter can be turned off, you may not need it at all. Technically it protects the speaker from infrasonics, but in reality there's very little material you could play that would even threaten the driver. The DSP gives you the option of playing with the settings easily at any time.

Edit: I looked at the block diagram closer, the plate amp won't apply EQ to the high pass line outputs so you'd still need to correct that elsewhere if needed. You could easliy run EQ on the HTPC for that if you want.

[dtb]

Thanks for the reply wogg I'm glad to see I'm on the right track. I did notice on your site you used the RSS265HF-4 before, the one in your profile pic right?

I've worked up a design for the enclosure in Sketchup, I tried a few things but decided to go for something tall with a square base with the port up firing. It's probably bigger than I'd really like but it seems there's no avoiding this if you want to go this low.



I'd appreciate any thoughts on the bracing. I copied something I did with some bookshelf speakers, the enclosure is using 18mm MDF and I cut this into some lengths roughly 18mm wide and then cut that up into the lengths I needed to build up the brace, gluing to the enclosure piece by piece. Maybe not the best way to do it but doable given the tools I have and my rudimentary wood working skills.

Do you think this would provide sufficient bracing? The baffle will be double thickness 18mm. To my mind this should do a good job on the five remaining sides but you guys experienced in this may say otherwise.

[Chris Roemer]

It's very common just to use a few dowel braces (3/4" to 1"d) run between opposing panels. The outer edge of a "window" brace doesn't do all that much.
Nothing wrong w/a vent out the top. When I do that, I'll attach some hardware cloth (maybe 1/4" screen) over the inside end of the port, so if anything falls (or gets accidentally dropped) in there (by a kid), you'll be able to get it out fairly easily.

[billfitzmaurice]

The bracing that spans from side panel to side panel looks good, but perhaps spaced a bit wide. My preference is for no more than 8 inch spacing, preferably less. The spline braces on the panel faces don't do anything useful. There's no connection from the baffle to the back, which you should have. As for the double thick baffle, a single brace connecting the center of a single layer baffle to the back has the same effect as doubling the baffle and back thickness. More braces have even more effect, so all you end up with using a double baffle is unnecessary weight.

[dtb]

Ok thanks guys that's very helpful. The double baffle seemed to be pretty common so I figured it must be necessary but if I can add in one or two braces front to back to get the same effect that will save a lot of weight as you say. Would it make sense to have the woofer a bit higher and then I could have a brace at the mid point and possibly one at 1/4 height? Thinking about it this seems a good way to go, three braces side to side and front to back at 1/4, 1/2 and 3/4 height with the top front to back brace ending in the middle.

I'll redo my sketchup model, having the single baffle it's no longer square and while it's really not a big deal I can be a bit pedantic on some of these little details

[billfitzmaurice]

Double baffles are typically used in manufactured boxes because MDF is cheap, and the labor required to install braces isn't. It's best to place the driver so that the braces are spaced as evenly as possible. Another option is to place braces where the driver mounting screws can pass through the baffle into the braces.

[wogg]

Thanks for the reply wogg I'm glad to see I'm on the right track. I did notice on your site you used the RSS265HF-4 before, the one in your profile pic right?

Yup, that's correct. My box is a bit smaller, tuned to the low 30's. It would work well in a room, but its home is in a car where cabin gain gets me way below 20. I like your box design but that top port would make me nervous about losing stuff in there. Have you calculated the port length yet?

[dtb]

All right I've worked out the below. Front to back and side to side the spacing for the braces is 186mm (~7.5"), top to bottom it's 154mm (~6"), judging from what you've said bill I'm assuming this should do a good job. The edge of the driver to the sides is about 78mm (~3"), would you suggest an extra brace at that point on either side running to the 3/4 height, side to side brace?

I like your box design but that top port would make me nervous about losing stuff in there. Have you calculated the port length yet?

Yeah the port length is about 21" using a 4" flared port so that's kinda what's necessitated the tall box design wanting to keep things fairly simple with just a single straight length. I'll follow Chris's suggestion with some wire mesh, maybe near the top if I can figure something out, then I wouldn't have to reach far to pluck things out I can of course make it down firing and may well go that way but I'd have to add that 4" clearance adding more to the overall height and build some sort of feet with my rudimentary wood working skills.

[billfitzmaurice]

It looks good except that there's no connection from the center of the driver to the back. That can be done, but it's a lot more practical to put one brace from either side of the driver to the back, and to keep them as close as possible to the driver have them at the driver mount holes. Also, there's no need to have the braces intersect as shown. They can, but it takes a lot more work than to shift them to one side of the center upright.

[dtb]

Yeah I had thought about that just putting in each brace individually, something I probably would have done come build time.

The small difficulty will be the extra braces on either side of the driver as the port will interfere on one side. I guess the top side to side brace I will brace to the back with a single piece as in the sketchup model and then have the two driver braces go to that top side to side brace.

I guess the other option is go with the down firing port with one or two bends avoiding where the driver brace would go. The plan is to use pvc pipe so some 90 degree bends should be no big deal but then I'm not sure how to work out the correct port length?

[Chris Roemer]

Since the "screen" MIGHT create a little noise, I prefer it down in the box (bottom end in this case) rather than near the top. It's really for "emergency" use only.
My feeling about bracing front to back (near the driver) is that it's not THAT critical, since the driver frame adds rigidity to the baffle, and the distance from the driver edge to the baffle edge isn't that great. If you want, you could skip the F-B brace below the woofer and instead use a pair placed near, say, 4 o'clock and 8 o'clock (when viewed from the driver front). Prob'ly move the brace below there up a bit to compensate.

If you use an elbow, its effective length is the same as a centerline drawn through the middle of the curve.

[billfitzmaurice]

The plan is to use pvc pipe so some 90 degree bends should be no big deal but then I'm not sure how to work out the correct port length?

Measure it on the center line.

[dtb]

Thanks a lot guys, I think I should be good now to work out something that will work with the bracing and port once I've built the enclosure.

I've tried to calculate the internal volume as best as I can, I used an online calculator for the driver and did some calculations for the amp, bracing and port myself. Playing around in WinISD am I right in thinking that the volume isn't super critical? It seems if I'm off a couple of litres it doesn't have much effect on the frequency response or the port length.

I've tried to find info about tuning the port length, is it correct that I can use a tone generator and find the point where the driver stops moving and that will be the tuning frequency? I figure good practice would be to start with the port a bit longer than what WinISD suggests, measure and then shorten if needed?