My M8ta spin

[KJP]

Ive been working on these for almost a year? Two or three days a month for what seems as long as I can remember now..... :g>

I think I eluded to this project elswhere here. A GrahamT Extreme mt in some Jonmarsh Avalon klones with a semi constrained layer wrapped in a Sonus Faber and sandwiched in a Revel. Grahamarshavalonfaberevels?

I'm taking Jons baffels and attaching them to an enclosure made of 1/2 inch Birch ply. There are 5 full shelf braces, the walls are stuck with dynamatish adhesive backed pads and covered with a half inch layer of fiberglass under batting stapled in. Stuffed in that is a 3 by 26 inch Precision flared port. This enclosure is then slid into a shell of 1.5 inch MDF lined with two layers of 30lb roofing felt. The baffel will seal to the outer shell with neoprene strip.



and



for



From there the top half of the baffel will get covered with black leather, the bottom half and the shell will be painted with Duplicolor truck bed coating.

Next I'm going to bolt panels to the sides to allow grillcovers ala Revel

Image not available

and the grill drawn in

Image not available

At first the sides will probably just be Oak plywood with some cherry stain. I want to do them out of some nice polished marble slab but the budjet is getting harder and harder to get past the wife ;x( :brunette:

Some kind of footing system like the Kharma Exquisite Reference is also in the works. :T

These things are stupid heavy. I cant lift them. I dont have a scale, and I am kind of wimpy so.... ops:

I can play the woofers full range pretty loud with my ear on the sides and back and not feel or hear anything coming through. I think that was what I had hoped for, did I mention that I cant lift them?

Sorry about the lo Q cell phone pictures. I had a bunch of good photos that got stolen at the same time as my camera got stolen a while back..... :M

I just got this phone so I had to share.

I'll start the measurement process after the bulk of the finishing work is done and post my progress as i go.

Thanks to Jon for being more nuts than me. :B

[WillyD]

Those look badass.

[posix]

I beg you to re-think your baffle and "kidney"-shaped sides. your enclosure re-design kicks ass and I wouldn't want to see the geometric shapes spoiled by organic sides and grill. Just take the sides off and don't add anything else to it, it's THAT good on its own!
keep up the good work!

[Jed]

I beg you to re-think your baffle and "kidney"-shaped sides. your enclosure re-design kicks ass and I wouldn't want to see the geometric shapes spoiled by organic sides and grill. Just take the sides off and don't add anything else to it, it's THAT good on its own!
keep up the good work!

Agreed--- really nice work though. Can't wait to see it when it's finished.

[KJP]

Hm, I want grills, I dont really like the looks of the Avalon grills, I do like the looks of the Revels.....

Yup, I'm gonna do them kidneys and all. My wife said they look like Ears. Thats what I'm going with. Ears. Not kidneys.

Non of that is really in concrete at this point, if someone drew me a better idea I would love to see it. :T

[cjd]

Well...

the blobs look like something slapped on - not integrated, vaguely sloppy. Similar modification to the "Type Si R Super Turbo Ultra" stickers on stock Civics (though let us not forget the 12" exhaust tip)

If you like it, by all means... but it's not particularly good design. It has no relationship to the rest of the design, and that's the problem.

You can add organic forms to the geometric shape, but you need to do so such that it flows together as a homogenous whole.

Dig out some nature photography books and study some of the ways geometry plays with organic form. Flowers, leaves, insects...

I think it could be very slick done well.

C

[ThomasW]

After going to all the work to minimize diffraction with the faceted baffle, you're creating diffraction with the rounded element that sticks forward below the woofer. That's not a good thing.

[posix]

you see?! 4 against 1 - we win! take em off, I don't care what you call 'em but take 'em off!

[JoshK]

Its a free country, he can do whatever he wants. By I too think they looked badass without and quite strange with the ears. Thomas's point about the diffraction is a good one, it is going to be a performance drag to leave them on.

One thing I am having a hard time understanding is the box within a box concept. What is that for? Why not make the volume larger and just use the outer box, and tune them lower (or are they sealed?).

[posix]

5 against 1!

p.s. oh and it's not "free country", it's Internet and it costs money to run.

[KJP]

Well I doubt it would be THAT much of a performance drag. Next weekend I'll cut out another ear and measure it with and without both in place. If the effect is that bad I may ditch the ears.

I dont know anything about Turbo stickers on Civics. I thought the panels did follow the form. The back of the enclosures from bottem to top slope back then forward, the sides turns that into a curve instead of a sharp angle. From the side view the tapers on the baffels angle back starting from the middle, and the sides follow the same shape with a curve.

4 on 1, 5 on 1 the jury is still out. I would love to hear more opinions.

The box in a box is my attempt at a semi-constrained layer. Two layers of 30lb roofing felt is from some simple tests that Art Ludwig posted on his site. see those on this page. It seems pretty effective and I am glad I did it.

My main thing is I need to use grills, so they need to attach somehow. I like the Revels so that led me to this shape for the side. Quite reminiscent of the Revels I thought.... No? Do you guys like the look of the Revels?

Thanks for all the feedback, good or bad I want it all.

[JonMarsh]

My vote shouldn't count because of my prejudice and ties to the original design and to Avalon. :B

BUT, that said, if you want to do a Revel, do a Revel. If you want to do an Avalon look and act-a-like, you may want to stick more closely to the program. Without the ears, I think you've got a good thing going there. :T

Let's see how it measures either way.

~Jon

[dyazdani]

My main thing is I need to use grills, so they need to attach somehow. I like the Revels so that led me to this shape for the side. Quite reminiscent of the Revels I thought.... No? Do you guys like the look of the Revels?

I DO like the look of the Revel Ultimas, but their "ear" is not quite as large, i.e. it doesn't really stick out past the enclosure anywhere (just a little in one part really).

Maybe if you made it a bit smaller and left part of the top exposed. You could probably follow the lines of the back profile of the main enclosure.

I like the "original" M8ta design aka "Avalon-like" - but I am a conservative type of guy anyway...

[HMenke]

I would say that the side panels would look cool if the whole speaker had pretty much that profile...but I think they look rather incongruous on that particular faceted speaker cabinet. I would leave them off.

[Nathan P]

Here's what I think would look best on the grill holder things. I drew up an outline that follows the cues on the baffle and speaker itself. What do you think Kevin?

[posix]

is that 8 against 1 or 8 against 2 now?

[KJP]

is that 8 against 1 or 8 against 2 now?

jury is still out?

[cjd]

In the end, of course, the choice is yours.

C

[KJP]

Of course, but I would rather have some approval of DIYers whos opinions I respect very much.

[ThomasW]

From a functional standpoint remove the section of the side plates that stick out in front of the baffle.

From an aesthetic standpoint remove both the side plates completely.

Jon said the same thing in his post

[posix]

has the jury gone to bed yet?

put simply organic and geometric shapes don't mix well. you can still have your side panels if you must have them but you'd have to re-design them slightly to better follow the overall design of the speaker. make it a pointy ear if you must have ears. but seriously, your box on it's own is excellent and doesn't need anything extra added to it.

[KJP]

I think the jury died. :rofl:

I did like the drawing that NathanP did, it more closely followed the lines esp along the back, but I have a new idea that I believe may please the imperial court.

Picture this if you will. From this line up I plan to route off about a half inch of the baffle on all sides so it keeps its shape but is now set back. I make a wooden frame to build that back up, keeping the same shape, to cover with cloth, then on the inside of the frame I install wool felt to prevent diffraction issues from the frame itself. So the end product keeps the same exact shape pictured but has a flush removable grill from the line up PLUS adds the felt diffraction control. Under the grill I will still do a leather cover. The ear would have also given me some pretty wood grain, so that leaves me wanting to reevaluate the finish on the rest of the speaker. I want wood. I now think it would look best with it all veneered with some pretty burl of some kind. Ive never veneered b4. I am scared. Would someone please tell me that veneer work isn't as hard as I think it is?

[Nathan P]

I think that will look great, I don't know about the diffraction or felt etc, but the others should have more to say about that.

[ThomasW]

Veneering is VERY tricky on a design like this with multiple facets.

The crossover is specifically designed for shape and size of the baffle.

[posix]

birdseye maple anyone?

[Dennis H]

If you're not up to veneering the whole thing, you could always do black truck bed liner on the front baffle with all the difficult angles and veneer the back part with the square angles. I think that would look quite elegant.

[posix]

I have an idea - how about you finish it off with filler, sand it all smooth, primer and give it a coat or two of paint? cream or even white would work nice, but it's up to you of course. see if you like it. if you don't you can always veneer over it (if you feel like you're still up for a challenge afterwards, that is)! :T

[KJP]

update

A little update on my dumb speakers.

Since the peer review of my original idea went so well I never took any measurements of the ears/kidneys/blobs in place to see what effect on diffraction they would have had.

Top half of baffle is covered with leather. Bottom half if just stained with a poly containing cherry colored stain. The goal is to veneer them someday with something pretty (if the stupid things end up sounding good enough to warrant the effort/expense). The back shell is truck bed liner roll on. It was not as flaw concealing as I had hoped, agian this is something I can do better down the road.

The front of the grill frame will be a layer of wool felt cut to match all the angles, glued on, and covered with standard grill cloth.

Overall I am glad I went this route and am happy with the results thus far.

I dont get as much time to work on these as I would like.

more bad cell pictures

overall



grill frame off



grill frame going on



grill frame on

[wkhanna]

I think they look GREAT the way they are!

If those are your 'dumb' speakers, I would like to see the ones with mediocre intelligence.

[KJP]

Thanks for the compliment.

The only reason I call them dumb is because of how dang long I am taking to build them. ops:

I'm quite happy with them so far.

[Nathan P]

That looks great, I love the grilles, very classy.

[clearwaterms]

and the real test is how do they sound?

[KJP]

Not done yet.

Got the felt and grills done.





Still have the base to build, then crossovers.

Looking forward to posting my sock shot. (I will borrow a real camera for that)

[clearwaterms]

Not done yet.

Got the felt and grills done.





Still have the base to build, then crossovers.

Looking forward to posting my sock shot. (I will borrow a real camera for that)

So you still have not played them yet?

[KJP]

I listened to the woofers and tweeters by themselves. The woofers sound generally bad running full range. They have a nice smooth extended low end that I am pretty excited about. I have never had stereo bass before, I think I am going to like that a lot. They are going to end up crossed low and steep, 1.1-1.2khz so I hope the problems I hear in them running wide open get all the way filtered out. The Seas 27tdfc sound ridiculously good full range. I don't know how they do it. No bass of course but very impressive.

[kano32]

I listened to the woofers and tweeters by themselves. The woofers sound generally bad running full range. They have a nice smooth extended low end that I am pretty excited about. I have never had stereo bass before, I think I am going to like that a lot. They are going to end up crossed low and steep, 1.1-1.2khz so I hope the problems I hear in them running wide open get all the way filtered out. The Seas 27tdfc sound ridiculously good full range. I don't know how they do it. No bass of course but very impressive.

Things sound different when the crossover is employed and the drivers are placed in the enclosure. I hope you don't hurt your tweeters feeding them a full range signal. Low frequencies + tweeter = ouch!

[KJP]

making progress

I did some measurements last weekend and have been working on a crossover design. I would like to present this work for some review before I start ordering part$$$.

Here is the unaltered response of woofer and tweeter.



The networks



And a combination of highpass and lowpass predicted response/phases with goals and summed responses with normal and reversed tweeter polarity.



Am I crazy or does this looks really good?

TIA for any input.

Keivn

[JonMarsh]

That looks a lot like what you should be shooting for, but do you have an adequate amount of baffle step comp? What does the raw transfer function look like of the filter?

This is the LP transfer function on the "original". Pretty much a full 6 dB, as these are intended to be put out away from the wall.



Here's the crossover- note that the first inductor is rather a bit larger than the second on the woofer, that's because it has a different corner frequency and is responsible for BSC as well as one of the poles in an LC pole pair.



Just suggesting you give this a look, otherwise you seem to be well on track.

~Jon

[KJP]

Thanks for the insight. Speakerworkshop doesn't generate a networks transfer function so what I did was make a new driver, associate the real woofers impedance curve and a flat frequency to it and then simulated the networks effect on that. What I saw was only about 3db BSC. Does this method sound correct?

These are going to sit with the woofers 3 feet from side walls and 4 feet from the back wall so I do think I will need more BSC than 3db. So I reworked everything to this end.

new network



new networks transfer function showing about 5.5 db BSC



and the new combined response



If this looks ok I am ready to submit an order for parts.

Thanks again for the input.

[JonMarsh]

How about fine tuning the Q at the LF corner frequency? Notice your transfer function has a some peaking (higher Q) at 1 kHz, with a bit of a dip before that- I've found that to correlate with some resonance in the midrange in female vocals- note that the TF on the original M8ta doesn't have any peaking there. Of course, you may be experiencing some diffraction effects that are different from my baffle, which seems to cause you to need that response. I'd suggest looking at the 15 degree off axis curves on the RS225 and tweeter also, sometimes I average the two and use that for network development, though not in this case. It should be possible to get a nice phase control through the crossover region (verified as you do with the out of phase test), without any peaking in the LF crossover (higher Q) near the crossover point.

[Jed]

How about fine tuning the Q at the LF corner frequency? Notice your transfer function has a some peaking (higher Q) at 1 kHz, with a bit of a dip before that- I've found that to correlate with some resonance in the midrange in female vocals- note that the TF on the original M8ta doesn't have any peaking there. Of course, you may be experiencing some diffraction effects that are different from my baffle, which seems to cause you to need that response. I'd suggest looking at the 15 degree off axis curves on the RS225 and tweeter also, sometimes I average the two and use that for network development, though not in this case. It should be possible to get a nice phase control through the crossover region (verified as you do with the out of phase test), without any peaking in the LF crossover (higher Q) near the crossover point.

Translation- smooth out the peak at 1k :B

[JonMarsh]

Translation- smooth out the peak at 1k :B

Yeah, that's one of my shortcomings, using 250 words to suggest something that can be condensed to six or less- :W

~Jon

[KJP]

I can't get it to go away and still maintain good tracking of the target.

How much deviation from my lp target is ok to avoid this peak in transfer function? If it comes to choosing evils, which would be more important, target tracking or filter Q?

FWIW the difference between the saddle and 1khz peak is about .7db. Optimizing for a prettier TF yields way more than .7db amplitude deviation.

[Jed]

Yeah, that's one of my shortcomings, using 250 words to suggest something that can be condensed to six or less- :W

~Jon

Actually, I learn more when one's mind ponders all the variables. Your suggestion of looking at the off-axis response is something I'll consider when I get further along on my 3-way projects.

[Jed]

Keep playing with LCR trap circuit values/add one- you should be able to get rid of that peak and also maintain good tracking- and as Jon mentioned, maybe it is a baffle artifact that will disappear 15 degrees off axis, so therefore- not to worry.

Jed