Mini cum laude 3-way project thread

PE's "magna cum laude" project is fairly popular, and for good reason. Those things can play loud! The downside is that they are quite large, making it difficult for some people (like me) to justify and/or construct, and also fairly expensive which can be intimidating to first time builders. My solution was the "mini cum laude," a more budget oriented project with a similar layout (WWMT) and smaller footprint. The basic design uses dual SD215 8" DVC subwoofers (vented), a dayton PA130 as the midrange, and the hivi SD1.1 tweeter for the top end. It has an F3 of about 30hz, and a max volume of over 110dB (with about 100 watts) before woofer xmax becomes an issue. At no extra total cost, you can also use dual SD270 10" dvc subs sealed, which will give you another 3dB of sensitivity at the expense of a bit less extension (48hz F3), but with the floor and room gain I doubt this will be too big of an issue. In fact in smaller rooms, I'd wager the 8" version sounds too boomy due to small room gain. Also, since I have a feeling many will use this as part of a HT setup, if placed on either side of the TV and next to a wall the bass will be further magnified, making the sealed a better choice in this situation.

Cost for a pair if these is under $250!

Here's a rendering showing overall dimensions, layout, bracing etc. Since I did this one, I decided to build the midrange chamber onto one of the sides, so imagine the mid and tweeter shifted and moved over toward one of the sides. This simplifies construction and lowers baffle diffraction effects. Box is approx. 2.25 cubic feet after bracing and the mid chamber are taken into account, overall size is 11"X16"x36" which can be cut from a single sheet of 3/4" birch ply or MDF if you're careful about the cuts. Mid is open-backed t-line similar to the mini statements. I intend on heavily damping it , the lack of a rear wall is to reduce lower frequency standing waves, not to vent midrange into the room.

[IMGDEAD]http://www.flickr.com/photos/[email protected]/2189344342/[/IMGDEAD]

So I posted about this and did some sims and renderings, and a few people also wanted to see a more hifi version using a better mid and tweeter. Well there are hardly any reasonably priced midranges with high sensitivities, save the RS150, which seemed like it might work. For the tweeter I immediately chose the seas 27tdfc. I've used this tweeter several times and it simply can't be beat for the price. Low FS, low distortion, and $30. These two driver choices keep the sound quality high but the price pretty low. A bit (ok, a LOT) of crossover simulation later...

This simulation is for the SD270 woofers sealed, deeper extension can be had if using the SD215 vented (30hz F3), but at the expense of ~3dB of sensitivity. I'll have a simulated crossover for this version soon, but here's the SD270 version:

[IMGDEAD]http://homepages.nyu.edu/~msa286/laudehifixo3.gif[/IMGDEAD]

Price for drivers and XO parts is $350/pair, which is pretty cheap for a WWMT. This assumes using the jantzen p-core inductors for large values, NPE caps for large values, and poly caps and 18 guage inductors for the smaller values. I have a parts list I will post soon, everything but the tweeters (madisound) can be ordered from parts express. I know the parts count could be slightly reduced by being smarter about the notch filters, but it was flatter this way and not all that much more expensive so I think I'm going to leave the overall topology untouched.

I'll be building all four versions of this design, and I'll post my process and progress in this thread. I'll also be getting down and dirty with a measurement setup, so I'll be able to see how the results stack up against the simulations!

Please don't hesitate to offer any comments of suggestions, I hope this design can become popular to those with a somewhat limited budget or workshop, but who still crave a large, full sound.
 
Yes indeed I must have made a mistake with the URL for the rendering, I can't seem to edit the first post...
Here it is again:
2189344342_b85985c53d_o.jpg
 
I think I'm going to ditch the PA130 as the midrange in favor of something with (likely) lower distortion and a smoother response curve, at least for the vented 8" version where I don't need the extra sensitivity. The problem is mostly that the PA130 has a pretty large dip around 1khz and it's made worse on this baffle size, so although really cheap it might not be the best choice here.

I've mocked this up using the RS125 and HiVi SD1.1
This is about $300/pair, $50 more than the PA130 original concept but a lot smoother. The 8" drivers have lower sensitivity than the 10" drivers so we're at 86dB here as opposed to 90dB. The crossover point here is fairly high between the mid and woofers (450hz, a.k.a 2.5foot wavelength), but phase alignment is excellent. I'm not going to be able to go lower than 350 any way I slice it, will I see any benefit with a 350 xo rather than a 450? The woofers and tweeter already have very little interaction.
[IMGHTTPDEAD]http://homepages.nyu.edu/~msa286/laudedvc8rs125.gif[/IMGHTTPDEAD]
* Vented impedance not modeled here, assume a standard vented twin peak below 100hz...
 
Sure John,
I was thinking about an open back transmission line mid for two main reasons. The biggest is to eliminate the standing waves caused by a typical closed back cabinet, as the distance from the driver to the rear wall (15.25"), is smaller than the lowest wavelengths it will be producing, so a standing wave would be created that would lead to either a peak or dip in the vocal range, which is the last thing I want :). This is typically what leads to "boxy" sounding mids, as the rear energy re-enters the cone delayed in time and leads to muddy sound/

This open line is intended to be lined and filled with some damping material, I'll have to actually experiment and see what the right amount is. The goal is to damp the rear energy, but still let some escape for an ambient "open baffle" type sound, which should lead to larger sounding vocals and strings.

As for the RS125 lowering the SPL curve, the limiting factor with the 8" version is the woofers, which are 86dB. I won't be using the RS125 for the 10" version as the RS150 is a better match for sensitivity. I'll still try the PA130 version (I'm going to build a whole set of removable baffles so I can try all these combinations), but I want to make sure I have a backup if that doesn't quite work out as a midrange.
 

MaVo

Member
2006-01-07 12:40 pm
just for fun, you could make one of the woofers open baffle, so that the rear radiation off the woofers will cancel out, making your bass directional with a cardioid radiation pattern, which would also be the dest description for your mids, as you want to damp the rear wave. but that would seriously decrease your max spl.
 
I don't know if this will help, but instead of just leaving the back open, what would be wrong with using an aperiodic vent, that is adjustable. When you take measurements, and changing the opening you may get different readings, allowing you to hone in on the best flow through.

The only negative to this would be the extra cost to the project. However, only the person performing the experiment really needs to use one, until the best open area was determined.
 

tinitus

diyAudio Moderator R.I.P.
2005-11-24 1:47 am
Interesting project

Have you seen this new TB 5", looks ok

http://www.tb-speaker.com/detail/1208_03/w5-1611sa.htm

About the standing wave thing with a closed back mid chamber
I think that mounting the driver in a tube is almost the worst thing you can do...and I dont think you will avoid standing waves...think of a BR port, and as far as I know it produces a standing wave/ressonance that moves back and forth...allthough its open both ends

To minimise reflected sound I believe its good to have plenty of air around the back(sides) of mid driver...its closed back enough as is, in its raw form ;)

If you want it open, I would suggest to make mid chamber shallow/flat, but with the open "slots" at the sides, merely like a attached baffle, and no sides :cool: :)
 
tinitus said:
Interesting project

Have you seen this new TB 5", looks ok

http://www.tb-speaker.com/detail/1208_03/w5-1611sa.htm

About the standing wave thing with a closed back mid chamber
I think that mounting the driver in a tube is almost the worst thing you can do...and I dont think you will avoid standing waves...think of a BR port, and as far as I know it produces a standing wave/ressonance that moves back and forth...allthough its open both ends

To minimise reflected sound I believe its good to have plenty of air around the back(sides) of mid driver...its closed back enough as is, in its raw form ;)

If you want it open, I would suggest to make mid chamber shallow/flat, but with the open "slots" at the sides, merely like a attached baffle, and no sides :cool: :)

"T", that TB is so new it has not gotten out to any distributors as far as I can tell. At least it's not on any of the places I look. If you have any idea, please let us know.

It looks great, but you never know.
 
Max:

Would you mind explaining in some detail the crossover network for the RS150/27TDFC version? I would really like to know what types of filters are implemented (lowpass, highpass, lpads, notch, zobel, etc) and how (6db, 12db, crossover points, etc). All the dirty details and why. I know this is a lot to ask, but I'd really like to try to understand how you arrived at the topology you chose. Thanks!
 

tinitus

diyAudio Moderator R.I.P.
2005-11-24 1:47 am
John L said:


"T", that TB is so new it has not gotten out to any distributors as far as I can tell. At least it's not on any of the places I look. If you have any idea, please let us know.

It looks great, but you never know.



Your right, but here we cant get Daytons either :clown:

But I suppose it would be possible to order a pair factory direct...actually if my finances were just slightly better, I would have them ship a pair to Zaph ;)
 
m0tion said:
Would you mind explaining in some detail the crossover network for the RS150/27TDFC version? I would really like to know what types of filters are implemented (lowpass, highpass, lpads, notch, zobel, etc) and how (6db, 12db, crossover points, etc). All the dirty details and why. I know this is a lot to ask, but I'd really like to try to understand how you arrived at the topology you chose. Thanks!

No problem! Here's a broken down schematic showing which parts do what, followed by an explanation of why I chose the layout.

[IMGHTTPDEAD]http://homepages.nyu.edu/~msa286/hifixoexplained.gif[/IMGHTTPDEAD]

Crossover points are 400hz between the mid and woofers (would like to move this down to 350hz), and 1800hz between the mid and tweeter. I chose 1800hz between the mid/tweet since it's a good balance of coherency and power handling. The 27tdfc can be crossed as low as 1400hz, but can only get to about 103dB because of xmax issues at that point (the tweeter is actually moving too much). Crossing higher allows the entire system to reach about 108dB before the woofer's xmax becomes an issue. The steep 4th order electrical slope on the tweeter (acoustic 24dB/octave) improves phase alignment with the midrange and minimizes interactions with the big woofers that could cause phase issues in the midrange. The steep slope was also needed because the midrange falloff had to be steep to get rid of the metal cone breakup, and ended up being a solid third order (18dB/octave) slope. There is a very tiny .025mh inductor after the capacitor in the midrange circuit to further reduce the metal cone breakup. I will instead try the zobel method I used in the RS125 version I posted later, as this seems to work even better and is cheaper... There is also an LCR notch filter on the mid which for all it's complex looks does some minor shaping to the woofer's curve. I think this whole section can be avoided with a bit more work on the other parts... And finally the tweeter has a contour filter to flatten out its slightly rising response, which also gives the low end of the tweeter a bit more presence. This can probably be left out, I've heard this tweeter without it and it sounds fine, but is a little on the bright side.

The woofers and the midrange are crossed at about 400hz, i'd like this to move down to 350hz which shouldn't be too hard to accomplish (change the cap going into the midrange to 55uf and the inductor on the woofers to 4.7mh most likely, maybe add a 2-ohm resistor after the big 8.2mh inductor). This ensures good physical integration of the drivers, which are within the 1-meter wavelength from each other to minimize lobing). The subwoofers have a zobel which flattens the impedance and reduces some of the breakup, and a .1mh inductor further kills the breakup. This has a shallow slope at first which then accelerates

And I guess that's about it?
Let me know if you have any more questions!
Max
 
OK I've cleaned things up a bit for the SD215/RS125/SD1.1 version. Crossover point lowered to 325hz while at the same time reducing cost (fiddling with the woofer-mid circuit, was able to optimize a few components). Impedance is a solid 6-ohms and has very little variation. I'm going to apply what I learned here to the RS150 version now, see if I can simplify it a bit and remove unnecessary parts, as well as lower the crossover frequency.
[IMGDEAD]http://homepages.nyu.edu/~msa286/minilauders125.gif[/IMGDEAD]