New build thread using InDIYana '11 doorprize

[DanP]

I'm hoping to have these puppies ready for InDIYana '12, and I think I'm going to start a personal tradition of using whatever doorprize I win one year in a build that I bring the next year. So what did I win?...

A pair of Audax TW025A0s, which I've heard sound very nice, but I haven't auditioned them yet. They'll make up the top end of a super-slim TMWW with the Fountek FR88EX in the middle and a pair of ND105-4s on the bottom.

The woofers will be housed in about 18L tuned to 40Hz in a box that is 39"h x 5"w x 6"d (ID) with a 9" tall mid enclosure on top of that. Unless someone has a recommendation on port placement to deal with quarter-wave issues, my plan was to put both woofers as high as possible in the bass section of the cabinet with the port a couple inches underneath them and then heavily stuff the bottom half of the box to beat down the pipe resonance.

Overall dimensions will be 49"h x 6"w x 7.25"d with a 3/4" solid wood baffle. I'd really like to do walnut, but that's going to take a trip to the sawmill and I've already got some oak in the garage. We'll see how much time I can find between now and April 13th.

I used rabbet joints on this build for the first time ever and the added time was well worth the reduced stress when it came time to assemble the pieces. I will definitely be using this strategy on future builds. I've attached a few pics of my progress so far as well as the simulation in its current iteration. The mid was quite easy to deal with but the tweeter has some weird FR stuff going on with it and the woofers impedence peaks needed to be beat down before the LP would work, so this is not a low-part-count crossover. I think I'm at 15 components right now, but I'll try to knock that down a little during voicing with real in-box measurements. 3-ways suck.

I'm taking suggestions on EASY finishes for the sides, back, top and bottom. I don't have the patience nor the time for 10 coats of paint with 72 hours of drying and 2 hours of sanding in between.

Dan

[Pete Schumacher ®]

I'm hoping to have these puppies ready for InDIYana '12, and I think I'm going to start a personal tradition of using whatever doorprize I win one year in a build that I bring the next year. So what did I win?...

A pair of Audax TW025A0s, which I've heard sound very nice, but I haven't auditioned them yet. They'll make up the top end of a super-slim TMWW with the Fountek FR88EX in the middle and a pair of ND105-4s on the bottom.

The woofers will be housed in about 18L tuned to 40Hz in a box that is 39"h x 5"w x 6"d (ID) with a 9" tall mid enclosure on top of that. Unless someone has a recommendation on port placement to deal with quarter-wave issues, my plan was to put both woofers as high as possible in the bass section of the cabinet with the port a couple inches underneath them and then heavily stuff the bottom half of the box to beat down the pipe resonance.

Overall dimensions will be 49"h x 6"w x 7.25"d with a 3/4" solid wood baffle. I'd really like to do walnut, but that's going to take a trip to the saw mill and I've already go some oak in the garage. We'll see how much time I can find between now and April 13th.

I used rabbet joints on this build for the first time ever and the added time was well worth the reduced stress when it came time to assemble the pieces. I will definitely be using this strategy on future builds. I've attached a few pics of my progress so far as well as the simulation in its current iteration. The mid was quite easy to deal with but the tweeter has some weird FR stuff going on with it and the woofers impedence peaks needed to be beat down before the LP would work, so this is not a low-part-count crossover. I think I'm at 15 components right now, but I'll try to knock that down a little during voicing with real in-box measurements. 3-ways suck.

I'm taking suggestions on EASY finishes for the sides, back, top and bottom. I don't have the patience nor the time for 10 coats of paint with 72 hours of drying and 2 hours of sanding in between.

Dan

Get some spray sanding primer to put down a reasonably smooth surface and sand it lightly. Then spray a base coat of your color, followed by a textured paint. If you get one with speckles and don't want them, just paint one last coat of the base color, and then clear over that to keep it from gathering dust as easily. Hides the seams reasonably well without too much time involved.

And of course, there's always Formica.

[Wolf]

I'm glad I'll get to hear those! I think you chose drivers very well. Is the tweeter physically larger than the mid? Just curious.

Can't wait to hear these!

As to easy finishes- you could dye a can of polyurethane, and just give it some coats. That would be really easy. KoolAid may work.

Later,
Wolf

[DanP]

Yeah, the tweeter is 100mm and the mid is about 90mm I think. Both of those finishing ideas sound reasonable, but the dyed polyurethane sounds stupid easy, so that's prolly what I'll do. I've got high hopes for these as well.

Dan

[mzisserson]

Heheh cool... I like off-beat designs and this one looks neat! Great tweeter by the way!

[DanP]

Four more weekends until InDIYana! I've been sneaking out to the garage when I can trying to finish these up before the get together. At this point I've got all the mounting blocks in place and all four removable baffles trimmed and mounted. I found these composite deck screws at HD that look really tidy on the baffle. No dealing with inserts and I didn't even recess the heads, so using them saved a lot of time.

Dan

[diy speaker guy]

The woofers will be housed in about 18L tuned to 40Hz in a box that is 39"h x 5"w x 6"d (ID) with a 9" tall mid enclosure on top of that. Unless someone has a recommendation on port placement to deal with quarter-wave issues, my plan was to put both woofers as high as possible in the bass section of the cabinet with the port a couple inches underneath them and then heavily stuff the bottom half of the box to beat down the pipe resonance.


Dan

This plan of action is probably not going to yield optimum response, but since it's a 3 way, if the xo point is low enough it might not matter much.

Without a simulation, I'd suggest putting the port as far as possible on the opposite end of the box that the drivers are on - in other words as close as possible to the bottom.

I just saw this now and it's probably finished by now, but if not I can probably do a quick simulation of the bass section for you. I'd need all the t/s parameters including Bl and Sd as well as the port info (diameter, length and preferred location if any).

[DanP]

This plan of action is probably not going to yield optimum response, but since it's a 3 way, if the xo point is low enough it might not matter much.

Without a simulation, I'd suggest putting the port as far as possible on the opposite end of the box that the drivers are on - in other words as close as possible to the bottom.

I just saw this now and it's probably finished by now, but if not I can probably do a quick simulation of the bass section for you. I'd need all the t/s parameters including Bl and Sd as well as the port info (diameter, length and preferred location if any).

Just in time diy_speaker_guy. Even though the weather in Michigan has been beautiful lately, I still haven't managed to cut any holes yet, and I would love you to model this for me.

Here's the specs according to the manufacturer (Dayton ND105-4):

Re 3.7 ohms
Le 0.78 mH
Fs 53.8 Hz
Qms 7.16
Qes 0.59
Qts 0.55
Mms 7.2 g
Cms 1.2 mm/N
Sd 51.5 cm^2
Vd 20.6 cm^3
BL 3.9 Tm
Vas 4.5 L

Internal dimensions of the enclosure are 5"W x 6"D x 39"H with a 2"ID x 5"L port. Woofer 1 will be located at 36.5" from the bottom and woofer 2 at 32".

My current plan is to put the port at about 26", but like I said, that's only so I can fill the bottom with enough stuffing to kill the pipe resonance. I'd much rather use that resonance if possible. Thanks!

Dan

[isaeagle4031]

Look up Chris Roemer. He has modeled these drivers to death and has hands-on T/S numbers on them.

[diy speaker guy]

Drivers are in series.

Both of the following simulations use .5 ohm resistance in series with the driver to account for amp and cable resistance as well as an added passive series crossover inductor. Depending on the components you use your actual series resistance will vary.

The first picture shows your original plan. Since you wanted drivers in the top and stuffing in the bottom I had to model it upside down. I included all user inputs so you can verify they are correct. It's only shown with .25lb/ft stuffing, more stuffing would smooth it out more but it's not an optimal situation. Shown at xmax (I think) - I'll talk about this more later.

http://i43.tinypic.com/15nu0br.jpg

The next pic is the changes I mentioned, keeping the woofers and the port far away from each other. I modeled this one right side up, since the stuffing is at the top. Also shown with .25 lb/ft stuffing in the first half of the line. This looks a bit better.

The port is down 36 inches from the top (internally) with the drivers and stuffing up in the top half of the line.

The first graph in this pic is frequency response at xmax, 12 watts. You didn't give an xmax spec but IIRC it's 4mm.

The second graph shows excursion (xmax rms = 4mm x .707 = 2.8mm)

The third graph shows port velocity at xmax. For reference, the top line in this graph, 0.064, is equal to about 22 m/s. 0.03 is about 10 m/s.

Lots of people say this driver sounds good well beyond xmax though. If you push it much past 12 watts as shown the port noise is going to be a serious problem. A larger port diameter or a lower tuning (longer port of the same diameter) would help but unless you want to retrofit a folded slot port there's not much you can do about it now inside the confines of the woofer box though. A flared port would help in theory, but since a flared port needs to be longer for the same tuning and there's nowhere to put the extra length you can't do that either.

One final note - your port is 5 inches long and the internal depth of the enclosure is 6 inches. That doesn't leave much room for the internal end of the port to "breathe" unless your baffle is very thick. If you restrict the airflow too much around the port it will lead to even more port noise.

The port diameter is going to limit the performance potential of these speakers (if the nd105 indeed sound good well past xmax).

http://i43.tinypic.com/3584g78.jpg

[DanP]

Very cool, diy - thanks for the model. There doesn't seem to be any downside to putting the port at the bottom and you've certainly shown a downside to putting it at the top, so bottom it is. Now I wish I had cut the baffles so the bottom one could accommodate the port. That way I could spin it upside down and listen both ways to see how much of a difference there is.

The only thing that seems a little odd to me is that 0.25lb/cu ft in the top half of the line only turns out to be a little over an ounce of stuffing - is that really all that is recommended or should I try different amounts? Should I try to taper the stuffing gradually from woofers to port? It wouldnt be hard as there are 4 "sections" separated by window braces.

I'm aware the port is a little too small and a little too close to the back wall, but in my experience, it will only make an audible difference on a handful of songs, which is an acceptable compromise to me.

Dan

[diy speaker guy]

You can use as much stuffing as you want (need). I show stuffing at .25ft/lb because that's how I model it. I get the response as it's flat as it's going to be with just enough stuffing to knock down the worst of the mess, and only then add more stuffing.

More stuffing will really smooth out the responses I've shown in both simulations. You can use as much as 1lb/ft in the first half of the line. Keep the stuffing away from the bottom of the box/port area.

If you want I can try to optomize the port location for smoothest response, I haven't bothered yet. The drivers are not necessarily in the optimal location within their enclosure either, but overall it's certainly not bad the way it is in the second simulation. Improving it is possible but it's nitpicking.

I can show you what it looks like with any amount of stuffing you like. Let me know.

[diy speaker guy]

This is the same thing as the second simulation but with .75lb/ft.

http://i39.tinypic.com/29v1ffl.jpg

[diy speaker guy]

One more pic. This shows the port 26 inches down from the top (not to be confused with what I modeled in the first simulation, which was 26 inches from the bottom). The first graph is with .25lb/ft stuffing, the second is with .75lb/ft.

Basically you can put the port anywhere you want from 26 inches (from the top) down to the bottom. You can pick either of the locations I've shown or if you need to you can pick a new location within the specified range without harming the response too much.

I could probably get it flatter but I'd have to change the driver position and there's probably not much point anyway.

http://i44.tinypic.com/14nh4ie.jpg

[Paul K.]

Which version of Martin's software are you using? If it's 2_09_08 or later, you can change the scaling of the axes in the graphs for better resolution. I particularly like to do that for the Y-axis to show F3 better as well as the response ripples. Just thought I'd ask in case you didn't know.
Paul


One more pic. This shows the port 26 inches down from the top (not to be confused with what I modeled in the first simulation, which was 26 inches from the bottom). The first graph is with .25lb/ft stuffing, the second is with .75lb/ft.

Basically you can put the port anywhere you want from 26 inches (from the top) down to the bottom. You can pick either of the locations I've shown or if you need to you can pick a new location within the specified range without harming the response too much.

I could probably get it flatter but I'd have to change the driver position and there's probably not much point anyway.

http://i44.tinypic.com/14nh4ie.jpg

[diy speaker guy]

Sure, I can show my graphs in better resolution in the future if it would help.

[DanP]

One more pic. This shows the port 26 inches down from the top (not to be confused with what I modeled in the first simulation, which was 26 inches from the bottom). The first graph is with .25lb/ft stuffing, the second is with .75lb/ft.

Basically you can put the port anywhere you want from 26 inches (from the top) down to the bottom. You can pick either of the locations I've shown or if you need to you can pick a new location within the specified range without harming the response too much.

I could probably get it flatter but I'd have to change the driver position and there's probably not much point anyway.

http://i44.tinypic.com/14nh4ie.jpg

Graph 2 looks mighty tasty and the added stuffing doesn't seem to have killed the output too much. I'll probably start with .75lb/cu ft and if the bass sounds anemic I'll back it off a bit. I'm hoping to get some holes cut and drivers mounted this weekend. Thanks again!

Dan

[diy speaker guy]

If you want the FR graph without added resistance (so you can add the inductor resistance in PCD instead) and impedance graph so you can use this in PCD I can put that up too.

[DanP]

If you want the FR graph without added resistance (so you can add the inductor resistance in PCD instead) and impedance graph so you can use this in PCD I can put that up too.

I appreciate it, but I'll WT3 the woofers once they're mounted.

[diy speaker guy]

I was bored so I played with this a bit more. The graphs are still 5db/division just so you can easily compare with what's already been posted. This is probably what I would do with your box, feel free to take it or leave it.

http://i43.tinypic.com/wv22rt.jpg

I moved the drivers down to 10 inches from the top (the centerpoint between them). The port is at 22 inches down from the top (internally). The port is 10 inches long for a new tuning of ~31 hz. It can do 10 watts at xmax.

The lower tuning helps the port velocity by both moving it down in amplitude (although that's at least partly due to the lower power handling) and also down to a lower frequency so it is less of an issue. By all accounts the driver has lots of clean throw, might as well use it.

I've also included a diagram of how to implement the extra port length while still remaining true to the simulation. By folding the port like this you have lots of room for the extra port length and also have space to use flared ports if you want.

The graphs show response at both .25 and .75ft/lb stuffing as well as excursion and port velocity.

[DanP]

Making progress. Mounting the Fountek driver is no joke. It's only got about a 1/4" wide flange and the mounting holes needed to be drilled out to accommodate even a #6 screw. The back of the baffle needs to be chamfered severely to allow for any kind of air flow as well.

Dan

[Wolf]

Making progress. Mounting the Fountek driver is no joke. It's only got about a 1/4" wide flange and the mounting holes needed to be drilled out to accommodate even a #6 screw. The back of the baffle needs to be chamfered severely to allow for any kind of air flow as well.

Dan

Is that Red Oak 1x6? I've done that before! :D
Lookin' pretty good!
Later,
Wolf

[DanP]

Is that Red Oak 1x6? I've done that before! :D
Lookin' pretty good!
Later,
Wolf

It is red oak, but not 1 x 6. These are actually 6" wide and 1 x 6 is only 5.25" which for some reason still catches me off guard when I go to HD to buy stuff like this.

Luckily I have a saw mill about a half hour from my house that sells rough sawn lumber in widths from 6 - 12" up to 3" thick in just about any species you can think of. The price is about 1/2 that of the big box stores even after you have them plane it flat for you. I really wanted to do these speakers in walnut, but I already had the oak on hand and time is in short supply around my house.

Dan

[DanP]

Just in time. I got them buttoned up last night, but only got a couple hours of listening/voicing time on them. They sound quite good as is, but I think I can make them sound a little better and/or shave off some crossover components after InDIYana.

I'm very impressed with the Fountek driver and we all knew the Daytons could hold their own in the bass department. The verdict is still out for me on the Audax tweeter. It has a nice sweet sound, but the combination of its naturally rising top end and the diffraction peak on such a narrow baffle left me with what amounted to a 7dB dip in the middle of its range to deal with. By the time it was all said and done, I'm still +/- 2dB from 100 - 10,000Hz, but it's a little peaky on the upper end.

Attached is the final simulation from in-box driver measurements - I didn't get a chance to take final system measurements yet.