6.5" + 1" low distortion design

Hi All,

I am looking to improve my living room setup (detailed below), and have a proposition I’d like to pass by you all to check I’m barking up the right tree!


Current setup

I currently own a pair of Void Mycro 6s (http://voidacoustics.com/uploads/Mycro-6-manual_1.pdf). They just about satisfy my SPL requirements, but are far from Hifi. I have modded them by filling ports and extra wadding - but still fall a little short.

Bass is handled by my dual 8” PPSL (http://www.diyaudio.com/forums/subwoofers/303976-sorely-tempted-ppsl-3.html), filling in below 120Hz. Power is a Lab Gruppen FP2400Q, providing 380w per Mycro 6 and 1280w to the 8” PPSL. DSP is handled via a DCX2496, with a Focusrite Saffire DAC.


Proposal

I would like to build pair of replacement speakers for the Mycro 6s that will, in approximately the same package, exceed their SQ - while maintaining SPL output. Budget approx £450.

I aim to do this by a new cabinet, that improves on several areas of the Mycros:

BMS 4524 compression driver, crossed at 2kHz.
⁃ Elliptical (?) horn, W90, H40 dispersion (?). 3D printed in ceramic.

B&C 6NDL38, crossed to sub at 120Hz
⁃ Sealed cabinet (4.5L), Q=0.707

⁃ Stiff cabinet (18mm MR MDF)
⁃ Zero parallel cabinet walls (heavily trapezoid, slanted top)
⁃ Flat front with tapered edges, minimising baffle step / effects

⁃ Active crossover. I have a spare TAS5630 I can use for tweeter (100w @ 0.1% THD @ 8R)
⁃ EQ through REW


Crossovers

BMS4524 was chosen due to its ideal 2kHz crossover, and extended response (20kHz+). Also heard great things about it on the interwebs.

HF response of the B&C is fantastic, but LF is a little on the edge. I propose to cross it at 120Hz to the sub, but by letting its natural roll-off do the work, with a 70Hz 2nd order to protect excursion @ 200w. This sims the same as a ported version of the same with a 120Hz 2nd order crossover. (The sub will have a 120Hz LP)


Summary / questions

⁃ My original plan was for a sealed 8” driver, but struggled to find a nice 1” that would cover the extended range to prevent beaming (1.3kHz+). Is there anyway around this or is there any need? 6.5” + 1” seems to be a nice sweet spot.
⁃ I have chosen Elliptical horn due to its constant directivity. My listening position shifts alot in the horizontal plane, hence the 90x40 dispersion. Is this still the best choice? I was toying with Tractrix, but is a little to directional maybe.
⁃ Would MTM make sense / be worth the extra space?
⁃ Is the natural roll-off a suitable method for crossing the LF of the 6.5”? Or will I run into phase weirdness?
⁃ Any other driver suggestions? I have looked primarily at BMS, B&C and Beyma offerings.
⁃ Will this sound a marked improvement better!? (Or do I need to go further…)


Thanks everyone who made it this far. I am looking forwards to your responses :D
 
Sounds like a really nice project. My 1st thought was to suggest an 8" to give the mid some headroom.

Maybe lower your xo point if you're really worried about the 8" beaming. I'm running 15's crossed to 2" CD's and even at really loud levels if I turn off my subs and the 15's there's really not much coming out of my horns so I reckon for home use you'll get away with a lower xo point.

If you'd prefer to stay with the 6" driver rather than MTM I'd suggest trying a TMM with the lower mid run as a 0.5 way.

If you're crossing to the sub at 120Hz I'd be inclined to just add a 2nd order HP at 120Hz and use a 4th order LP on the sub.

Rob.
 
For a smooth transistion at 2kHz from the midrange to a 90H x 40V horn, the polar pattern of the midrange should approach 90-degrees at 2kHz. Naturally, this midrange should not show any signs of cone break-up around 2kHz. The goal to "get in clean" and "get out clean" around 2kHz often requires BW3/BW3 crossover circuits for both constant power and rapid attenuation of any midrange cone break-up, or non-linear "SPL blips" common in compression drivers. At first glance, the Faital 8PR200 looks decent, also B&C 8PE21

For matching M-T physical time alignment, there are a few short horns which might come close. You want a horn with an entry angle similar to the exit angle on the BMS 4524 compression driver. *suspect 14-degrees*
.
 

Attachments

  • Faital 8PR200.JPG
    Faital 8PR200.JPG
    150.8 KB · Views: 637
  • 1in shallow horns.jpg
    1in shallow horns.jpg
    151.5 KB · Views: 717
  • Narrow baffle.jpg
    Narrow baffle.jpg
    237.8 KB · Views: 655
Last edited:
BMS sells a 1" horn model 2193 which they recommend for the 4524 compression driver. SO, we can assume the fiberglass 2193 matches the exit angle of the 4524. Looks like ~$50. In the USA there are several ABS plastic waveguides/horns for $10-$15 which include polar pattern measurements you can study.
 

Attachments

  • BMS 2193 horn.jpg
    BMS 2193 horn.jpg
    124 KB · Views: 857
Hello RobWells, LineSource. Thanks for your replies!

8" was my initial plan, with a crossover at 1.8kHz or so. As I started looking into the beaming issues I went off the idea for dispersion reasons. However, your confidence is giving me confidence that this is not such an issue after all :)

The remaining issue is the 1" CD. With a horn supporting lower frequencies(1.4Khz seems like a sensible XO), it would likely switch to the BMS 4544, which supports a happy 1.3kHz XO. However, I'm not sure if that is specced by assuming the full SPL being used for the driver? (IE, will the 4524 be happy running at always <-6dB at a low XO?). If so, the smaller CD may be a little sweeter - but I understand that the HF will start to be limited off axis by the larger horn.

An Oblate spheroid horn seems to be a good choice in terms of dispersion. It makes sense now that of course the polar response of the CD+LF should match in all axis - keeping that power response constant! Physical time alignment is no issue, the DCX can accommodate for that handily.

Both the 4524/4544 have 20deg exits. I'll do some scouting around to see what horns are available in that range. Otherwise I am more than happy to 3d print one. I have difficulty accessing CNC routers at the moment, so keen for a round profile I can do with a hand router !!

The next big challenge is narrowing down the 8" choices. I was happily limited for the 6.5"s.. The Faital looks a good start, 18sound have some interesting options too. And Beymas catalogue of 8"ers is verging on ridiculous... Any more pointers are appreciated :p

Cheers,
 
If it's of any help I used to run my Beyma CP380M's crossed at 1.2K with no problems at all. (reference level home cinema) The recommended crossover frequencies are for PA use using the full output of the driver. At home I doubt we would use more than a couple of watts.
 
I like the direction this is taking; I've been planning a very similar build: 7-8 litre cabinets, 6" woofs (probably Faital 6FE100 - I like their drivers, they out-perform the price bracket) and a 1" compression driver with OS waveguide.

I was planning to post a design/build proposal thread tonight, but spotted this and decided to throw in 2cents worth here :)

The new compression driver which has really caught my eye is the Eminence N151M-8, a steal at CAD$59 from qcomponents.ca.
JBL M2 for The Poors
See posts 453 -454

I don't know the exit angle of the N151M-8, nor the entry angle of the waveguide I'm looking at: the Denovo EOS6 from DIYSoundGroup: Denovo Audio EOS6 waveguide DIY Sound Group
so, those are details to be determined for suitability. I guess if you're truly hard-core DIY and want to make your own waveguides, I have to admire that. I think the EOS-6 is worthy of a look, for US$10.75 though.

I just took the manufacturer's published graphs for the drivers, extracted frd and zma files with FPGraphTracer. Then I added boxmodel data, baffle step, and extracted minimum phase with Charlie Laub & Jeff Bagby's FRD Response Blender spreadsheet. Then, within an hour, using Bill Waslo's XSim, I had modeled a passive crossover that shows +/- 1.5dB from 100Hz-20kHz.

I know real measurements, polars have to be made, but as a proof-of concept sketched on the back of a cereal box it looks pretty darn good.

Then, after I'd done all that, I discovered the B&C 6FHX51:
B&C Speakers
I was once bitten by the Altec 604-8G bug, so the co-ax approach has its charm too.

Pardon my rambling, don't want to hijack the thread, i just hope my comments are food for thought.