Coming back to life! Question about 10" woofer position in a 3 way speaker

[EdL]

Your use of 2 RS270's begs the question:

Why not seal one of them and port the other?

[fjhuerta]

Yeah, I'll use the SB29. There's no reason to use the DQ25 or the 27TAFN - dual woofers and midranges will undoubtedly result in very low distortion levels, and the SB29 has the lowest distortion figures of the three tweeters I thought about using.

EdL, I *think* (not sure) that a ported / sealed combo *may* result in bass cancellation due to phase issues near the tuning point. I've never tried it, but an educated guess would be not to do it. So it's either ported or sealed all the way.

Tomorrow is still routing day..

[fjhuerta]

Memos to myself:

1) Next time, wear earplugs or something.

2) Even when measuring, routing, measuring again and routing, I should measure once again. There was a quarter of a circle that still was 1 mm. short. I don't know how it happened.

3) Small mistakes can be corrected with the Dremel cutting circle jig and lots of patience

DONE!!! Now to design the midrange enclosure and the bass one.

[cjd]

Looking good! :T

C

[Silversmoky]

Nice! :T Those are going to be beautiful. I like how the Tang Bands look next to the SB29 tweeter!

[fjhuerta]

Thank you all! Yes, I think they look pretty good; the Tang Bands are my favorite looking driver of them all.

While I design the midrange enclosure (which should be pretty easy), I'm wondering about "slimming" down the appearance of the speaker via their finish.

I thought about three choices:

1) Maple veneer + matt black baffle, black grilles.
2) Polyester black all around, black grilles.
3) Polyester white all around, black grilles.

While the maple veneer seems to be the most conventional choice, I wonder how they'd look in ultra-gloss white, with the black and titanium drivers... I'm thinking the answer would be "very cool". Since my living room is black leather, they'd look fine.

Hmmm...

[Silver1omo]

mmm, why not something like satin black with a white maple baffle?

[fjhuerta]

The problem is, I don't want to add veneer to the baffle because the drivers are pretty flush right now. If I veneered it they'd be sunk... I can't do much to it, except paint it.

So, does anyone know what colors "slim down" things?

[ultrakaz]

Can't you use a gasket for the drivers the same thickness as the veneer to keep them flush? I'm not saying veneer is better, just that it is still an option if you wanted a real wood finish. Painting without a pro style sprayer and booth can be a tedious time consuming endeavor, especially if gloss on a large surface.

Vertical stripes and darker colors tend to slim things down. Perhaps, a contrasting center banding of 2/3rds the width of the baffle running from top to bottom will keep the overall profile slim.

[fjhuerta]

Got a package from Parts Express - we've got grilles! I've never assembled a set, but they seem pretty easy to do. Also, I've decided to add 1" spikes to the bottom, so I'll need to very slightly redesign the baffle...

[fjhuerta]

Not much was done this week-end. I did the pilot holes (is that the way they are called?) for the drivers and got some cosmetic screws for the tweeters. I already drew the midrange enclosures, and cut the grilles.

I got the Parts Express kits. They say they are made for big speakers, PA systems and vintage stuff. I agree. I couldn't believe the size of the frames, they are really huge. And I think at least parts of them are made of fiberglass. Sturdy stuff.

After cutting and measuring, I finally got my grilles. They'll look fine, although

1 I still can't find a 5/8 bit, which is needed to install the grill to the baffle, and
2 I haven't yet glued the cloth to it.



I may have someone else finish the grill, since I have no clue as to how to properly fix the cloth to it.

[Paul W]

If you're after a 5/8 drill bit, Harbor Freight has a set of drills from 9/16 through 1" on sale for $34. HF drill set link

Very handy. I've had a set for 20 years and used the 3/4 and 1" earlier today.

[fjhuerta]

I found the bit yesterday, and the grilles are installed! It's a pretty nice kit, but definitely built for big speakers. They need 4 5/8" x 5/8" holes in the baffle... good thing I had a 1 1/2" baffle, or else it wouldn't have worked at all.

I'll send the baffles to my local furniture maker this Thursday. I still haven't decided sealed or ported? Ported or sealed? Time's almost up...

[lhwidget]

I found the bit yesterday, and the grilles are installed! It's a pretty nice kit, but definitely built for big speakers. They need 4 5/8" x 5/8" holes in the baffle... good thing I had a 1 1/2" baffle, or else it wouldn't have worked at all.

I'll send the baffles to my local furniture maker this Thursday. I still haven't decided sealed or ported? Ported or sealed? Time's almost up...

Port it Seriously, I think you can try both. What's the system Q if you seal the ported volume?

[fjhuerta]

I believe sealed is around 0.5.

I've finally decided on the outer dimensions: 114 x 37 x 70 cms. It's a huge box. The midrange inner box is 45 x 22.5 x 15 cms.

The outer box, once the inner box is substracted, ends up at 200L. The inner one is a 6.5 L enclosure (sealed). So, I'm going ported on the bass.

The enclosure will be veneered in Zebrano wood (at least as far as today, may change my opinion tomorrow). The baffle will probably be, too.

Now - I'm disappointed about the Parts Express grilles. They are far too bulky, and there's not enough friction to support them. Furthermore, on a project this size, the cloth ends up distorting the frame, so after a while one of the ends of it comes loose from its socket. I'll go with the usual wooden frame magnetic grille.

My local furniture maker has a couple of good ideas. He insists on no frame, because he thinks the drivers will look great with the Zebrano wood. I agree, but my son usually invites his friends over, and they are not as careful as he is regarding speaker cones (they are 4 year olds, so I understand it!). I do need frames.

I also found a couple of old PE spikes I'll reuse for this project. I loved their looks back then and I think they'll look great on this tower.

[lhwidget]

I agree about going with the ports, if the system Q is 0.5, the bass will probably sound a bit thin...

Too bad about the grill frames, they look like a neat system. I wondered if the cloth might pull the frames in with a big/long grill design.

[fjhuerta]

It has been some time, but things are going well. The enclosure was a bit of a nightmare. It's extremely heavy and complex on the inside, due to the reinforcements. So the furniture builder is taking quite some time to finish it.

The enclosure will cost me $700, including all materials and finish. It's being done in Zebrano laminate, and yes, even the baffle is laminated. I'll build some gaskets in order to make the drivers flush.

I expect the cabinets to arrive here on Tuesday.

I have to wonder how I'll move them around...

[cjd]

Wait till you get drivers and crossovers in there.

[JonMarsh]

Looks like you're having too much fun, Javier- but as is often the case, slow work takes time!

I have to wonder how I'll move them around...

do you have a plan? A rough idea of target weight? I'm sure we're all curious here to see them after you get them home...

[Face]

I use a hand truck with roll bar padding on it to move my large Tannoys around.

[fjhuerta]

Oh boy. These things are really heavy...

Well, for me at least. They are close to 130 pounds. Without any drivers or crossovers.

And they are HUGE. My wife gave me that look that said.... "yeah, you said they were big, but THIS?"

But, wow... are they pretty. The guy who did them was proud of them, now I can see why! ;x( He even added the magnetic grille, but sank the magnets behind the Zebrano wood. So I can run them with or without grilles without the ugly looking holes I had before!

Pics... just check out their size when compared to my current speakers...









I also think I found a name for them.... Papalotl. It's "butterfly" in Náhuatl, an aztec-derived language. The name "Papalote" (or "kite") also comes from it. We had just come back from the Papalote Museum in Mexico City, and loved every second of it (it's a kid's museum) earlier the day I got the speakers. You can see the logo of the museum on the blue bag on one of the sofas.

[Face]

Forget monkey coffins, those are...eh never mind.

Very nice! I hope you plan on building a platform under your couch to raise it up.

[Silversmoky]

Very nice Javier! Can't wait to see them with the drivers installed. That is some pretty veneer!

[fjhuerta]

I've been slowly working on the speakers. 1 ~ 2 hours a day. I finally managed to finish the midrange enclosure. It measures 17 x 45 x 14 cms, and it's completely sealed from the rest of the enclosure. It's fully lined with 1" foam and completely filled with polyester.

Those Tang Bands are really wonderful-looking drivers; probably my all time favorite. The SB Acoustics tweeter fits very neatly between them.





Tomorrow, if I'm lucky - impedance plots!

[fjhuerta]

Since I'm at a very boring meeting, I have time to think... heh.

Here's where I'm at. I like the Duelund curves. So far I've only built a single speaker using it (the Zaph ZDT 3.5 with a different filter) and I was happily surprised at how the design managed to blend every driver perfectly.

So, a Duelund could sound like a natural filter for the speakers.

Trouble is, I know the RS270 starts behaving badly at 1 KHz. The Tang Band and the SB tweeter don't worry me at all, the W4's break-up occurs far too high and the tweeter is a champ. But I don't know - if I aim for a 300 ~ 350 Hz crossover point, will a 20 - 1,000 Hz bandwidth be enough with a Duelund filter?

Time to start reading again...

[fjhuerta]

Some more food for thought...

1. Crossover topology: Quasi - Duelund vs. 4th order L-R. 4th order is tempting. I know phase will be easily aligned. It's relatively easy to design. Component count is decent. BUT... and this is a big one (lol): I can't seem to get a decent enough impedance out of the woofers / mids, since they both drop to 3.x Ohm at certain points. And I cannot do a series connection on the mids, since their efficiency is too low to cross them over to a pair of RS270s in parallel. Quasi - Duelund means 2nd order, and as far as I've been able to model means slower slopes (obviously) and less aggressive impedance plots. Which leads me to...

2) Inductors: If I use iron cores, efficiency goes up. I reckon I could get away with 4 dBs of BSC and end up roughly with a 90 dB/W speaker, but even with a 2nd order slope, impedance *may* reach 3.1 or 3.2 Ohm on a good day. Also, the saturation issue with the iron core worries me - I'm spending so much money on this project, I don't want to try to save some dollars on a stupid thing like that.

3) FR on the woofer. I know for a fact (MarkK's pages) that there's some funky business going on at 700 Hz, even though PE posts don't show it, on the RS270s. I don't know if 300 Hz 2nd order will be enough of a roll-off.

4) Crossover component quality - should everything be poly? Are 10W resistors enough=

I'll be doing some reading on Duelund style crossovers, but I studied Zaph's big SB12.3 pretty closely, and this is what I found.

1) 90 dB with 2 RS270s is easy, he got 91 out of a single 12" SB woofer after considering floor bounce, etc. That will be my aim.
2) Iron core inductors on the woofer are OK.
3) Electrolytics are OK in shunts.
4) The big SB 12" driver has serious issues at 550 Hz, and he got away with a 250 Hz 2nd order crossover.

I think I won't have any issues with my design, all things considered. I *think* I'll go with iron core inductors on the woofers in order to get as much bass out of them as possible. Once I have real life SPL / Z data I'll see if I can get away with, say, a 3.4 Ohm impedance throughout the bass (my self-imposed lower limit). If I can't, "back to the drawing board" I'll go (and I'll probably use air cores).

[cjd]

RS270 breakup is, I think, not the worst I've heard - however, you may find you really don't want it 2nd order. I'd give them a listen full range for a while, then with something to roll off the top a bit, just to get a better ear to what's going on.

Mine are ~90.5dB give or take (un-padded tweeter) so quite possible.

Love the box work. This is going to be a really nice setup.

And yeah, that impedance is going to go low. It just is. You can hear the benefit of an extra solid amp behind them no matter what, IMO, so probably an OK thing in the grand scheme of things.

[Paul W]

Are both you guys seeing the same 700Hz bobble that shows up in Mark's measurements?

[fjhuerta]

Yep, I'm still not convinced about the 2nd order roll-off. I've been reading Jon's excellent post on Duelund crossovers:



I *think* I can get away with this. Was the 270 clean to 1 KHz? I think it is - not as clean as a paper or poly woofer, but it gets there (at least according to PE's LMS files). I did some tests, and



It can do 1 KHz. It all goes downhill from there, but I don't think it really matters.

It's not exactly 2nd order, it's faster. This is 2nd order.



The schematic:



And the impedance: 3.4 Ohm minimum with iron core inductors! :T



I've used the KHF Tool on LSPCad and the Parts Express LMS files before when simulating woofer response, and the model always matches real life measurements pretty closely, so I'm hopeful a Duelund crossover, 1KHz center frequency with an Aleph of 2*sqrt(2) will work with the TB W4-1337sd. It can certainly do the 175 - 6,000 Hz required bandwith without any sort of issues. I just hope phase lines up correctly...

[fjhuerta]

This is a Duelund filter with an Aleph of 2.82. I got the number from Steen Duelund's site - he recommends it if you can get a midrange driver that is up to the task. A pair of W4-1337sd's can do it, I'm sure.



-18 dB points for the midrange are at 160 and 6,300 Hz. The woofer rolls off fast enough, I think, and the SB can certainly do 1KHz at -18 dBs. This is my target for this project.

[fjhuerta]

You know what? This speaker would be AWESOME with active crossovers... two mini-DSP's and 3 stereo amps... yeah, I measured them, and $200 in active crossovers surely beats whatever $$ amount the passive crossover needs.

Now, if only I had 3 identical stereo amps

[cjd]

You know what? This speaker would be AWESOME with active crossovers... two mini-DSP's and 3 stereo amps... yeah, I measured them, and $200 in active crossovers surely beats whatever $$ amount the passive crossover needs.

Now, if only I had 3 identical stereo amps

Won't be cheaper in the end, no way no how.

Also I disagree with two statements you made: 1, the RS270 break-up is NOT listenable above 1kHz and your filter will control it somewhat well enough, so , may do OK. I'm a bit concerned that it's riding the break-up peaks along the reference line, and that your measurements are somewhat smoothed and the real harshness will still be peaking close to the -20dB range at 3kHz (fairly audible).

2. Paper or poly have totally different problems in this range. Not necessarily harsh, so more listenable, but I find something else goes away.

Nothing ventured, nothing gained. Give it a whirl.

[fjhuerta]

Hi C!

The graphs weren't smoothed, they are the LMS files Parts Express had for download a long time ago.

I already took some measurements and modeled using real data. Things were pretty close to simulations, except for the break-up. Somehow, I don't think the off-axis aspect of it was reflected correctly. Nevertheless, the topology works well.

So far, I've optimized for FR and phase. There's a bit of the mild W4-1337s break-up at the upper end, but whenever I get rid of it, phase stops tracking correctly between the midrange and the tweeter.

There's no smoothing on the woofer measurements. I did model for diffractive baffle step, so it's there. I didn't consider the port output at this moment. The tweeter and midrange do have a 1/12th octave smoothing, because I took my measurements at 2.2 meters, and there was quite a bit of noise in them.

The schematic.

Crossover points at 410 and 2,500 Hz. Topology: Duelund, Aleph = 2.828, Fc = 1 KHz.



There's absolutely nothing out of the ordinary, it's just that I decided to model the whole thing in a different way. It's a 2nd order electrical on the tweeter, a quasi-fourth order on the woofer with Rs across the Cs in order to shape the curve to a 2nd order, with an electrical trap for the break-up. The midrange is harder to describe - I'd say it's a first order on the high-pass, and a 2nd order on the low-pass.

I tried to keep component count to a minimum.

FR.



90.5 dB/W efficiency, using iron-core inductors on the woofers Yep, the TB's wouldn't make it while wired in series, but in parallel they reached the goal without any issues. The tweeter was loafing around, its efficiency is close to 95 db/W.

Impedance.


OK, so impedance is closer to 3.1 Ohm in the bass. No way I can raise it without messing the FR, not at this point. Not bad everywhere else.

Bandpass.



As smooth as the TB is, doing a simple bandpass filter with the minimum amount of components is hard. Nothing to notice here except for the fact that I didn't compensate for the upper mild break-up. I'm not sure I'll do it at this point.

Notice the big RLC network - absolutely necessary in the lower end, IMHO.

Highpass.


Again, a simple 2nd order filter with a big RLC network.

[fjhuerta]

Lowpass.



Yep, there's the slight irregularity at 700 Hz and a slight peak after 1 KHz. Nothing to worry about, IMHO. Break-ups are controlled.

Transfer functions.



Smooth and unexceptional. I love working with the W4-1337 and the SB29 - they are so smooth, there's not much work needed with them. The RS270 plays low, so the break-up will hopefully not be an issue.

Simulated off axis response.



Again, smooth.

It has taken me around 8 hours to get this far with the crossover, but considering I'm basing most of my work on stuff I already did (the Quasi-Duelund ZDT3.5 with the RS180, RS52 and Vifa titanium dome) it has been far easier than the first time around.

The only thing I still don't know is if I'll hammer out the mild break-up of the W4. IF I do, phase is misaligned. I don't know. At -22 dB or so it may not be any issue, since Zaph said...

This driver's breakup node above 10kHz does not propagate as harmonic distortion lower in frequency, meaning you can cross it over as high as you want.

Cool... isn't the W4 lovely?

[Silversmoky]

Looking very nice Javier! You have really put the work in on this! For what it's worth I have a tower design that I used the W1337's as mids crossed over very similarly to what you have here and they have that same mild breakup at about 20 db's down. I notice no harshness and they sound very smooth yet detailed to me. But, I might not have the golden ears that some of the talent has here, so take it with a grain of salt! Can't wait to see these done. Good job.

[fjhuerta]

OK... here's a new variation on the crossover.

I tried lower A values. Phase issues were big. The lowpass had issues - the break-up was evident and hardly controllable. The highpass was fine, but the bandpass really needed more parts in order to work correctly.

I tried higher A values and things started getting better and better. I settled on A = 3.3.

The topology is the same, values are different. The only reason I did it is because the W4 is the centerpiece of this design; it's small enough to work higher up in frequency, and it definitely has the needed bandwidth.This one is A = 3.3. The midrange works a bit harder (320 Hz - 3,200 Hz x -over points), but really - no big deal.

The new -18 dB points are 140 Hz and 7,400 Hz. Still pretty much within the capabilities of the driver.

Impedance is a bit wilder, but on the safe side. Phase tracking is at least equal to A = 2.8. But I could add an RLC shunt that takes care of the mild W4 break-up... so now the midrange is squeaky clean. And phase tracks perfecly well; better than my last design.

I'd like to know if anyone sees any problem with the design; if there aren't, I'll work it a bit more and build it. I think a 4" driver can do 3.2 KHz without beaming - off axis simulations seem to point this out. At 15°, there's no phase cancellation, either. Sweet.

Here it goes.

FR



Lowpass



Highpass



Bandpass



Z

[fjhuerta]

Schematic



Transfer function



Off axis



The off axis measurement is a real one I took at approximately 15° (with such large cabinets, exactly measuring 15° and moving the speaker would have been kinda difficult!). There's no phase cancellation whatsoever and FR is pretty flat. I also took a measurement at 30° - 40° and it's equally well behaved.

So, things are moving forward, although a bit slowly. I'll keep on working with this crossover for a while and post the definitive version when I finish it.

[fjhuerta]

Looking very nice Javier! You have really put the work in on this! For what it's worth I have a tower design that I used the W1337's as mids crossed over very similarly to what you have here and they have that same mild breakup at about 20 db's down. I notice no harshness and they sound very smooth yet detailed to me. But, I might not have the golden ears that some of the talent has here, so take it with a grain of salt! Can't wait to see these done. Good job.

Thank you! I do think the breakup won't be noticeable, but as you can see I managed to get rid of it... I wonder if my design doesn't have any issues because I think it could work...

[fjhuerta]

Oh yeah... how did I measure the woofer? It went pretty easy... I'll do a post on the subject. I liked how in the end the simulation looked pretty close to real life measurements. As C said, it's a great sanity check.

[cjd]

PE smoothed. Your later data shows the dip I knew was there, but seemed missing in the first graphs.

Also, the latter option is really looking like it should work quite well on the ear - the bobble ~700Hz is workable, and you've cleaned out the stuff that's just harsh and painful quite eloquently IMO.

My wife is my sanity check on "listenable" - her limit on bad stuff is VERY low (there's a reason we watch most movies at no more than ~75dB nominal - louder is just too loud for her most of the time, though I can creep up on 80dB and more her ears are ringing at the end, just like they do when getting out of the car after a drive)

You know things are heading the right way when far MORE of what you really want to do (w4 grunge management) just works along with everything else.

[fjhuerta]

OK, 18 crossover iterations later, I think I'm done.

I'll be using A = 2.8, so that the midrange isn't working that hard in the lower end. According to Jon, the unit would be optimal at 300~400 Hz, and Zaph's measurements point out that although the unit is very good, it doesn't have the lowest distortion profile. So, A = 2.8 will cover these bases.

Also, on the Heavy Metal Testing thread around here (I think), a post is reprinted from Darren Kuzma where he points out at irregularities above 3 KHz coming from (I think) vibration issues. So I'd rather avoid those, and, once again, A = 2.8 does the trick.

I did work out the phase issues between mid and tweeter, too. And with 2.8, impedance never drops below 3.1, while perfectly supressing the woofer break-up modes.

So that settles it.

I'd only have one more question... I'm deciding between iron core Erse inductors on the woofer (a 3.9 mH) or air core 14GA's. I'm getting close to 1.5 dB (!!!) more of output at 70 Hz or so and below with the iron cores. What would you do?

[fjhuerta]

Done!

After modeling 23 different x-over combinations (no, I'm not kidding, either!), I arrived at what I think is the definitive one. Phase is OK, impedance won't drop below 3.1 Ohm (3.4 was very optimistic or only possible with air core inductors, which dropped efficiency by 1.5 dB!), but most importantly, the drivers are being used where their distortion is lowest, not one of them is being forced by excessive travel, and their break-ups are avoided or trapped with shunts.

Better yet, the crossover is relatively cheap, so to say. $210 USD, with poly caps in series everywhere, electrolytics on shunt circuits, air core inductors everywhere except for the first one on the woofer, and Dayton resistors (doubled up at the woofer shunts and also where used in series with the mids).

One new feature I had never tried before. I think I talked about it in this thread. I like the BBC dip, mostly because I use my speakers in a room with hardwood floors and walls. But I always end up being curious about how a "flat" speaker would sound when used with "audiophool" material.

Enter the "Adaptative BBC Dip" mod.

It's a really stupid thing, but it'll work. Here's the general idea.



Circuit topology is the same, except that I re-ordered components. Check out the 3 10 Ohm resistors in parallel connected to the midrange circuit. My plan is: R12 is not going to be soldered to the rest. I'll run one wire from the generator (+) leg to a "BSC terminal plate" mounted just above the input terminal plate. The return wire from the "BSC terminal plate" will be soldered right to the C6 input, effectively creating a parallel element to R13 and R9 . I'll then use a dual banana plug



With the third 10 Ohm resistor (R12) soldered between the legs. If I plug in the banana plug with the 10 Ohm resistor, I'll have no BSC.



If I leave nothing plugged, I'll have a nice 2 dB BSC. If I replace the 10 Ohm resistor, I'll have any value in between. No need to take out the X-over boards. Cool, yet stupid



This is the circuit impedance:



And the Xfer function:

[fjhuerta]

Bom

Here's the crossover BOM:
Item Qty. Price Total
Dayton DNR-1.0 1 Ohm 10W Precision Audio Grade Resistor
Part Number: 004-1
2 $1.25 $2.50
Dayton DNR-10 10 Ohm 10W Precision Audio Grade Resistor
Part Number: 004-10
6 $0.98 $5.88
Dayton DNR-6.0 6 Ohm 10W Precision Audio Grade Resistor
Part Number: 004-6
4 $0.98 $3.92
1.0uF 100V Non-Polarized Capacitor
Part Number: 027-320
10 $0.30 $3.00
15uF 100V Non-Polarized Capacitor
Part Number: 027-344
4 $0.74 $2.96
50uF 100V Non-Polarized Capacitor
Part Number: 027-354
4 $1.06 $4.24
150uF 100V Non-Polarized Capacitor
Part Number: 027-364
4 $2.68 $10.72
Dayton Audio DMPC-0.68 0.68uF 250V Polypropylene Capacitor
Part Number: 027-408
2 $0.87 $1.74
Jantzen 0.35mH 20 AWG Air Core Inductor
Part Number: 255-030
4 $3.59 $14.36
0.51mH 14 AWG Perfect Layer Inductor
Part Number: 266-325
2 $14.16 $28.32
4.5mH 18 AWG I Core Inductor
Part Number: 266-564
2 $10.17 $20.34
Erse Super Q 3.9mH 16 AWG 500W Inductor
Part Number: 266-920
2 $19.56 $39.12
Dayton Audio DMPC-5.6 5.6uF 250V Polypropylene Capacitor
Part Number: 027-425
2 $2.50 $5.00
Dayton Audio DMPC-7.5 7.5uF 250V Polypropylene Capacitor
Part Number: 027-429
2 $3.04 $6.08
Dayton DNR-0.51 0.51 Ohm 10W Precision Audio Grade Resistor
Part Number: 004-.51
2 $1.25 $2.50
Dayton DNR-0.82 0.82 Ohm 10W Precision Audio Grade Resistor
Part Number: 004-.82
2 $1.25 $2.50
Dayton DNR-5.1 5.1 Ohm 10W Precision Audio Grade Resistor
Part Number: 004-5.1
2 $1.25 $2.50
Dayton DNR-7.0 7 Ohm 10W Precision Audio Grade Resistor
Part Number: 004-7
2 $1.25 $2.50
Jantzen 0.55mH 20 AWG Air Core Inductor
Part Number: 255-038
2 $4.58 $9.16
Jantzen 0.65mH 18 AWG Air Core Inductor
Part Number: 255-236
2 $6.54 $13.08
Jantzen 0.025mH 18 AWG Air Core Inductor
Part Number: 255-198
2 $2.81 $5.62
Dayton Audio DMPC-50 50uF 250V Polypropylene Capacitor
Part Number: 027-443
2 $11.38 $22.76

[fjhuerta]

Here are the final curves. I hope Steen Duelund is proud. I sure know the W4-1337s is one of the very few drivers that can be used this low and this high... A = 2.82, Fc = 1000 Hz.







I ordered the parts from Parts Express. I'll be picking them up next week, when I'm at Disneyland So, crossover construction will start February 14th, as well as final tuning of the enclosure. I can hardly wait.

[Duelund Audio]

This is a Duelund filter with an Aleph of 2.82. I got the number from Steen Duelund's site - he recommends it if you can get a midrange driver that is up to the task. A pair of W4-1337sd's can do it, I'm sure.



-18 dB points for the midrange are at 160 and 6,300 Hz. The woofer rolls off fast enough, I think, and the SB can certainly do 1KHz at -18 dBs. This is my target for this project.

Actually, he recommended A=4, but the demands become even harder to fulfill, he only used heavily modified or self built drivers for these.

Otherwise, very nice thread. :T