Adire Extremis MT

[GrahamT]

Hi guys,
Just wondering if anyone here (Thomas, Robert, Marsh, Hank etc) is planning on designing a 2-way with the Extremis? I know Adire is going to have some designs eventually, but when and how much is anyones guess.

[JonMarsh]

Well, I have some (half a dozen) Extremis on order, but originally was considering them for a RD50 two or three way. Now I'm leaning towards 10" Dayton Reference series for that, but still haven't made a decision-I'm sure I'll evalute both.

I've also got RS-170 7" that I want to play around with for a two way, so maybe I'll try both- they're gotten in first for retro-fitting my X1 clones. But I'm considering some other designs. I'm curious to see the final Sd of the Extremis before making up my mind what to try doing with it. Knowing me, it will still be a low crossover frequency design.

I'm thinking in terms of MTM's for both the Extremis and the 7" Dayton reference series, just because even with lots of Xmax, you just don't have enough cone area with single 7" or 6", considering the desirability of keeping doppler IM reasonably low.

~Jon

[GrahamT]

Thanks Jon,
I see. I am looking to copy a simple small bookshelf design so I might have to wait for Adire. I'm not in a hurry anyway, no more money right now:W

[JonMarsh]

Well, that's something else I thought about doing, a bookshelf in about a 0.75 cu fit encolsure, like the one's I did with HiVi D 6.8's last year for my daughter, but you know, as some of my friends say, does the world really need another 6-1/2" two way, given how many have been designed to day?

But that's a pretty easy thing to do- however, probably the least expensive tweeter I'd pair with this driver is still something like one of the lower end Excel tweeters, or the Northcreek D28. Well, maybe an LPG26T is another possibility; they're only ~$40. But, as folks around here know, I don't do minimalist crossovers, because I beleive in "coddling" the drivers, and getting the most possible out of them, as accurately as possible. Which is why I don't try to do ultra cheap two ways, because good crossovers do cost.

With the Extremis, I'd probably go for a 4th order L-R at about 1500-1800 Hz. This makes the midwoofer to tweeter spacing for staying within 1/2 wavelength feasible and relatively non-critical, and stays away from the upper range of the driver, where cone modes and energy storage start to smear the sound.

~Jon

[GrahamT]

I like the idea of a 4th order LR as well. Since I am a student, I need something small and portable which is why I am looking for a bookshelf style. Dan Wiggins is looking at some tweeters still for his design, he like the Peerless 2978 for a budget design and the Hiquphon owII, SS 9500 and Morel MDT30 and 44. I would like something cheaper because I plan to build a new 2 way with the Extremis when the XBL^2 tweeters arrive, maybe a year from now I hear. That LPG 26T looks pretty good to me too, I like Ti.

[Davey]

Jon,

Did you recieve your Extremis drivers yet?

Davey.

[JonMarsh]

Adire isn't scheduled to ship them on the pre-order until mid November, so I'm just hoping they'll have arrived by the time I'm back for Thanksgiving.

They've got a lot of new info up on the web site, including some THD plots, a spectral contamination plot, and some nice views of the basket (showing extended pole piece and Faraday sleeve)

Image not available

and the VC assembly.

Image not available

Here's the complete magnet assembly- no doubt there will be minimal masking of the rear wave with this design.

Image not available

It seems like we have a surfeit of new drivers to get excited about this fall...

:B

~Jon

[GrahamT]

I cant wait to get mine. Without any plans available yet I will probably run them full range as TV speakers just for fun.

[Dennis H]

Interesting graphs. The THD looks very good although they tested at 94dB rather than the 96dB Seas uses. The impedance bobbles at 600 and 1K look suspicious though. It will be fun to see what Jon finds on his nearfield-mic torture test.

Can Jon find true love with a poly cone or will he cast it aside like all the others? Tune in next week and find out in an all new episode of As the Driver Turns.

[GrahamT]

According to Dan, the bobble at 600 is caused by a reflection of the spider, the bobble at 1k is due to the affects of the copper sleeve on the inductance.

[JonMarsh]

According to Dan, the bobble at 600 is caused by a reflection of the spider, the bobble at 1k is due to the affects of the copper sleeve on the inductance.

I have a healthy bit of skepticisim, but I'm always optimistic....


Talk about contradictions!


Since the Extremis 6 is also really a 7", I may do some testing of these in my X2 enclosures, too.

The impedeance graph IS a bit unusual; I have other drivers with faraday sleeves that don't exhibit any discontinuity. Since my first thought on these was for the Saint-Saens diopole line array, usability to 600 Hz was my main concern.

It will be interesting to see how they test out near field- this is the real measure, in my experience. Of course, performance in the enclosure and so forth is important, too, but this is up to the speaker designer as much as the driver manufacturer.

The thing I really like and am curious about is the motor linearity over that first 6-7 mm of Xmax- it really looks superb. That ought to make quite a difference - just have to see what we give up, besides more greenbacks, in order to get it.

Imagine what a nice MT or MTM with a Seas Excel tweeter might do- not inexpensive, but ought to really kick butt.

Put that up on a bass bin with something reasonable - say, an RL-i 10" woofer, or two of them, in a sealed enclsoure (basically a 10" versio of the TC2+), and you would a system with ~ 90 dB sensitivity, 4 ohms, moderate size, and very high power handling and output capability.

Now, I'm supposed to be moving away from box systems, but this might be a hoot....

~Jon

[GrahamT]

I understand the skepticism, and the optimism too. Personally, if these speakers (whatever design I happen to build using the Extremis) are 10% better than my Paradigm Mini Monitors, I'll be happier than a pig in you know what.

[JonMarsh]

Well, with a good design attempt, I doubt you'll have any problem hitting that "10%" better point- I guess it just depends on which various factors you want to focus on.

One thing about midwoofers like the Extremis and the Dayton's - to get balanced system performance, until we see comparable Dayton RS or Adire tweeters with something like an XBL2 motor, you're going to have to spend a bunch more on the tweeter than you might expect. I've been "forced" into this situation with many of my current designs. I'm hoping that with a good crossover, I can get the LPG 26T to perform well in some applications, but I'm resigned to big SS tweeters or Millenium Excel's in some others.

And BTW, after inquiring a bit, it looks like early December before we'll see our Extremis midwoofers. Should be very interesting.... I'm going to have a VERY busy week on the 14th! :B

~Jon

[GrahamT]

The Paradigm's are great, but I am looking for a little more clarity and detail. It's good to know that I can beat them though. The low end will be handled by my beloved EBS Tempest and the Extremis MTs will be powered by a Sansui AU999.

The LPG tweets look good to me too, and if you have a chance to do a xover I would really appreciate it as would others I am sure. I'm pinching the pennies because I am planning to start mech engineering next fall.

[JonMarsh]

The Paradigm's are great, but I am looking for a little more clarity and detail. It's good to know that I can beat them though. The low end will be handled by my beloved EBS Tempest and the Extremis MTs will be powered by a Sansui AU999.

The LPG tweets look good to me too, and if you have a chance to do a xover I would really appreciate it as would others I am sure. I'm pinching the pennies because I am planning to start mech engineering next fall.

Can't say I blame you, Graham- school isn't cheap these days!

LPG26T should be here next week some time. They look like they'll even go a little lower than some single magnet MB Quart tweeters I've used in the past. I'll probably try something like a 1800 Hz crossover; anything higher and the center to center distances will be sub-optimal.

~Jon

[GrahamT]

Are you still thinking about using a 4th order Linkwitz-Riley for the LPG and Extremis? It sounds like a good match to me.

Also, just curious, would this driver be a good match to the Vifa XT25 because you could crossover around 3K, or is the FR of the Extremis too ragged up top?

I'm looking forward to any project you do with these drivers. I learn more about design every time, even if most of it is still going over my head.

[GrahamT]

Extremis MT ins 3/4 cu ft PE enclosure, using LPG 26 T for tweeter- wideband, tuned for low end and output in a small box. Crossover about 1.8 kHz, cauer-elliptic equivalent to 8th order L-R.

Didn't see your post on HTT till now. You are going to be very busy with all these projects. And 8th order! Sounds pretty cool. I've never heard speakers with 8th order crossovers before. I'll have to do some research on that and look closer at the 8th order on the M8 project. I cant wait to start building this design.

[JonMarsh]

The Extremis is too ragged about 2 kHz, IMO, to want to use it much there. Just my inate conservatism. I suspect some mild breakup modes; nearfield investigation and energy storage measurement should tell. MarkK has agreed to test an Extremis using his Praxis setup as soon as I can send him one.

Re the crossover, I've been doing this sort of off the beaten path cauer-elliptic crossover which has one more reactive component in each filter section than conventional 4th order network, but I tune to an acoustical transfer function for the first 48 dB or so of attenuation to between 6th and 8th order LR(actually, it can be tuned steeper, and to other Q profiles, but then I don't like what that does to the group delay). The incremental cost adder isn't very significant compared with a standard LR-4. (IMO).

The M8 8" two way with HiVi M8a woofer and Vifa XT25, as written up in a three part AudioXpress article last year uses the same concept, if you want to see more details- or search the site here; there's an example in the M8ta thread, on the floor standing version of the M8 two way with new drivers.


I think these are going to work quite well in this "small" project, but what I'm really cautiously hopeful about is the potential in my big dipole system, the Saint Saen, where I hope to use 8 or 10 per side with RD50 ribbon, and probably some Fountek JP2 about 4-5 kHz. With this system, I won't have to use the Extremis any higher than 500 or 600 Hz.

~Jon

[GrahamT]

Re the crossover, I've been doing this sort of off the beaten path cauer-elliptic crossover which has one more reactive component in each filter section than conventional 4th order network, but I tune to an acoustical transfer function for the first 48 dB or so of attenuation to between 6th and 8th order LR(actually, it can be tuned steeper, and to other Q profiles, but then I don't like what that does to the group delay). The incremental cost adder isn't very significant compared with a standard LR-4. (IMO).

I like I like! :T The most important part of a design IMHO is the crossover network. It was a really good idea for the M8 and it should work well here too. I agree that it shouldn't cost that much more than a normal 4th order LR.

I have been looking at the response of the LPG and it is hard to find info. It does appear to have an amazingly flat FR both on and off axis. It has a pretty high power handling too. I am surprised more people haven't used it. Have you heard it before Jon? How does it compare to metal domes used in commercial speakers like Paradigm, B&W etc or the popular Seas 27mm Al domes?

If I didn't have so much homework and if it wasn't freezing outside, I would be out there right now cutting up some MDF and butcher block. :lol:

[JonMarsh]

Aerial Acoustics and Totem use it in several models. The response curve is pretty good on and off axis, and it has a genuine metal aluminum faceplate, whereas most tweeters at this price point use plastic.

I've heard some folks think it was only "so-so", but then getting the most from a tweeter is dependent on the crossover, and if you pair a "revealing" tweeter with so-so electroncis, the result will be definitely "so-so". I've heard the same things from folks about the big Focal tweeters (TC120dx2), which from 1800 Hz on up are quite good, and very efficient.

So, we'll see. The real "budget" performer would be trying to get the Dayton RS180 to play nicely with the LPG26T. If that can be pulled off, should be cool, except folks will probably gripe because of the cost of the crossover components! There's always something.... :rofl:

The PE cabinets shipped last Friday... just wish Adire would, now!

[GrahamT]

The PE cabinets shipped last Friday... just wish Adire would, now!

Me too. I have emailed them and they say "soon". Which PE cabs did you order, the 20X8 MTM or the 17X10 MT cab. I assumed the 17 by 10 but I just want to make sure before I cut my baffles.

I am going to learn a lot from this project, cant wait. Hopefully I will be able to fit the crossover in the cab. That xover for the M8 MTM was huge!

[JonMarsh]

I ordered the 17" X 10" x 12.5" 0.75 cu ft cabinets for MT system, (Extremis plus LPG26T), and I orderd the 1 cu ft MTM cabinets (22" X 9" X 13.5) for the MTM's. At this time I'm just planning on doing an MT Extremis system; considering the driver Xmax, it should perform well.

IF you plan to do MTM's with the Extremis, are and making your own enclosures, I'd go with the same panel layout (22" X 9"), but make them deeper to increase the internal volume to 1.5 cu ft if you want to run them ported wide range. For two Extremis drivers in one box, you should be using a 3" flared port. I wanted something the same size as the speakers I built for my daughter last year, which were pretty kick butt for a small set- the HiVi D6.8 with 3" voice coil and neodymium magnet tuned to 32 Hz in that box had some nice bottom end on the James Taylor cut "Gaiia" from the Hourlass SACD (folks at the No. CA DIY meet thought a sub was turned on!). The Extremis should be even better...




2" flared port exits the rear.

The key to making the crossover fit in a smaller box is component selection; I'm going to look at that pretty carefully; probably Jantzen inductors, since they're narrower and taller; also, I use Solen caps as required in parallel with values to minimize the count and size as much as possible. I've found by splitting up the crossover into three boards (HF zobel, HF network, LF network), the individual boards are more reasonably sized.

~Jon

[GrahamT]

At this time I'm just planning on doing an MT Extremis system; considering the driver Xmax, it should perform well.

I agree. Plenty of displacement and your MT design is the one I plan to copy. I'll cut the baffles to 17 by 10 inches.

I plan to use all Solen components because I am in Canada and also use seperate boards for easier xover placement.

[ThomasW]

According to Dan Wiggins the build date for the drivers is Dec 7th.

Shipping will be shortly after that.

[GrahamT]

Thanks Thomas. I know they are a small company and delays are common. I just wish I had those drivers sooner. Looks I will be busy over the Christmas break.

[GrahamT]

The CSS website says that the Extremises will be delivered during the week starting Dec 12th.

[GrahamT]

Well I'm still waiting to get my Extremises, but I noticed that there are some new FR and impedence graphs on the Adire website. They dont look as bad as before. I wonder if they are smoothed or if this is just the FR of the final drivers and not the protos?

http://www.adireaudio.com/TextPages/Extremis6PageFrameText.htm

[JonMarsh]

That frequency response plot is at 10 dB/division! Which isn't exactly kosher... without knowing the test baffle and measurement details, it's hard to comment fairly- but note between 500 Hz and 1.5 kHz about an 8 dB difference in response.

Maybe they should have gone to Peerless for an HDS cone, as in the 850439? Meow...


The impedance curve looks more "real", though it's amazing how it stays so flat then rises rapidly above 5 kHz or so. Doesn't show the same copper cap behavior as an SS cone driver. Well, we'll just have to see....

~Jon

[GrahamT]

On the CSS website:
"Extremis 6 drivers have been delayed but are still expected next week." :cry:

I think I will get some Baltic Birch ply or something to start making the cabinets over the break.

[JonMarsh]

If you have a Rockler store in your area, they carry 12" and 2-1/2' wide pieces of BB ply. I bought some on that trip for the Mini-Maxx and Saint-Saens projects.

Got one of their sweatshirts, too, on my last trip to the Pleasant Hill store here in the Bay area.

It seems a little weird that we get these updates on delivery from CSS, but Adire has been relatively closed mouth. Well, I can understand they've probably got a pretty full plate, getting production going in their new Chilean facility.

~Jon

[GrahamT]

None of those stores in Toronto but I can get cut BB ply at a few different places and I have a table saw for more accurate cuts. I think I will make the box a little bit deeper than the PE one but use more radial braces.

Should I add some depth and if so about how much to account for the huge crossover (using Solen caps and inductors) or just make the adjustment for the decreased Vb by changing the port length? Or would it be better to build an external crossover?

I was looking at and thinking about the new FR plot. (Assuming they are accurate) Could you flatten out that hump centered at about 700 Hz by starting the BSC a little lower?

Thanks for the help Jon

[JonMarsh]

Well, it's something to take into account. I tend to not trust stuff I haven't measured myself from many companies, unless I've correlated my measurements with theirs successfully. It's a weird deal. For example, I've always had good correlation with Focal's published data and what I measure; not so with Accuton.

What enclosure width are you planning, Graham? Since I'm taking the lazy way out, with the PE enclosures, and have designed and tested a baffle layout for 7" midwoofer and tweeter using BDS, if you can stay to a front panel width of 10", that would be cool. I'd allow the equivalent of 120 - 160 cu inch extra for the crossover - that's not really much. I'll be trying to opmize the crossover design for size- don't want it too big myself, either. Try to keep the net volume at 28 liters or the port length will become a problem.

~Jon

[GrahamT]

Thanks as always Jon. I am planning to use a 10" baffle and if you could send me a pic of the layout or the location of the drivers on the baffle I would appreciate it. I'll calculate the volume to be about 28 L so this should turn out to be a pretty big box for a 7" driver.

Those PE enclosures sure look nice for the price but I enjoy making the enclosures myself, even in my shop which is about -4* celcius.

[JonMarsh]

Ouch! That's not a shop, that's a refrigerator! You Canadians are made of sterner stuff then a lot of us folks from the US of A.

I'll prepare you a panel drawing; concept works out pretty easy (I'll explain in the next post on this).

BTW, do you want to do an "Extreme" version of the Extremis? I was playing with Unibox this afternoon, with the Extremis model, and if you aren't using a sub but want to optimize the very deep bass output capbility, then consider a fairly deep box, 34 liters, tuned to 25 Hz; it would have an output capbility from 23 Hz to 40 Hz at 1 meter, one box, anechoic, of > 100 dB, with a max power input of 50 watts. Of course, in a room with boundaries, two playing, etc, you get more output. That's not a small box for a 7" driver, but you could minimize the "impact" by making it deep- which you'll need to do for the port, anyway.

Let me know if you want to go that way, and I'll draw up some simple AutoCAD plans and PDF it for you.

~Jon

[GrahamT]

Let me know if you want to go that way, and I'll draw up some simple AutoCAD plans and PDF it for you.

~Jon

Yes please! You are a gentleman Jon. I have an EBS Tempest but I dont want to use it for 2 ch so the extreme version sounds good to me as long as I can still use my 2" flared ports (I can make the center tubes any length).

[GrahamT]

I have thought about this design and I dont think it will work with my 2" ports will it? The air speed will be very high at those frequencies. Do you recommend the 21 L ,28 L, or 34 L design for use with 2" flared ports?

Thanks for the help Jon.

[JonMarsh]

For a driver with this much swept area capability (more like an 8" woofer), you do need more port area to fully utilize it's capabilities. A 2.5" port or even 3" port is preferred.

With a 2" port, you'd get some compression at the highest output levels - you'd be OK up to output levels around 95 dB/cab at 23-35 Hz (worst case point), or about 97-98 dB at 30 Hz and above.

The smaller box would reduce the LF output and of course the port speed/output, by rolling off the frequency response a bit more. At the same time, it relies a bit more on port output through 50 Hz.

Personally, if I was building a box from scratch, even if I only had the 2" ports, I'd go with the larger design tuned a little lower. The port length is about 10" long, for 2". For 2.5", it's about 14", which is why getting the extra volume through depth is important if you go that route (besides keeping the front panel the same would mean no changes to crossover for BSC). If you got some cheap ply or PB, you could build rough boxes for either size, and compare.

To summarize, 34 liter net internal volume, minimum 2" port for 40 watts peak at 23 Hz (just under the "redline"), and about 94-95 dB in the mid 20's.

~Jon

[taz13]

-4'c

Man you easterners (southerners) and your heat waves. :rofl: :rofl: My shop temp is currently - 35' C with no wind inside and 3 radient heaters (110v) I am comfortable but tools can be chilly so 1 heater is just for keeping the tools warm.
Canadian tire and Lee Valley both sell workshop 110v radient heaters that work great, so you are warm but don't have to heat up the shop air. I may work for a utility but I hate paying for power. :M
I like this speaker as I am planning on using it with the CSS w125s or w125st for the mids and still looking for a tweeter, but may try the t140 from CSS.

Rick

Wide Mouth Mason Stew

[GrahamT]

To summarize, 34 liter net internal volume, minimum 2" port for 40 watts peak at 23 Hz (just under the "redline"), and about 94-95 dB in the mid 20's.

Sold! I will build the 34 L design and keep the front baffle the same dimensions as yours (10" by 17" I believe). 2" flared port with 10" center tube. I will be using a real 50w per channel amp and my nominal listening level is 85 dBC (jazz, blues, classical). I picked up some ply and will try this box out when I get the baffle layout and PDF. I really appreciate all the help with this project Jon.

Man you easterners (southerners) and your heat waves.

Rick,
I know we have it pretty easy compared to you guys. I have some heaters but unless I am out there for a long time I dont bother turning them on. I know what you are saying about the tools though.
CSS has a design posted for the WR125s and the Audax TM025f1 at the top of the website if that interests you.

[taz13]

Already ordered the parts for the Xover and hoping they will make a nice bookshelf speaker.
As for heat that is the beauty of radient, you stay nice and comfortable and the air is cold but the mass around you is what heats up. And only use 1.4 Kws @ $0.053 per kilowatt hour, that to me is affordable(about 7 cents an hour/ heater). :T I hate paying utility bills. :M

Sold! I will build the 34 L design and keep the front baffle the same dimensions as yours (10" by 17" I believe). 2" flared port with 10" center tube. I will be using a real 50w per channel amp and my nominal listening level is 85 dBC (jazz, blues, classical). I picked up some ply and will try this box out when I get the baffle layout and PDF. I really appreciate all the help with this project Jon.

Will be watching construction progess and waiting for listening review.

[JonMarsh]

Yeah, if only Adire will ever ship!


OK, Graham, I'll post the front panel layout later today. Sounds like we're synchronized.

~Jon

[GrahamT]

Yeah, if only Adire will ever ship!
OK, Graham, I'll post the front panel layout later today. Sounds like we're synchronized.

~Jon

Sounds good to me Jon, I am your humble apprentice willing to learn. I'll be picking up some more supplies later today.
I just applied for engineering at UofT and Waterloo here in southern ON. I wont hear anything for a while but it is exciting to start a new direction in life.
Time to celebrate by making some saw dust! :T

[GrahamT]

Hi Jon,
I have been working on the box design this morning. For a net volume of 34L the box will need to be 10" wide by 17" high by 20" deep when made from 3/4" board. I have subtracted bracing (3 radial braces), a 2" double flared port with 10" center tube, the drivers, and the estimated 120-160 in^3 for the crossover. I tried the box in winISD (yeah I know but unibox wont work on my computer for some reason) and the extension looks great. Looks like I will have to make some new speaker stands aswell to hold these monsters! I might draw the box in AutoCAD if I have time.

[GrahamT]

Jon,
Does this look like a good box plan? I am still learning how to use AutoCAD.

One thing I noticed in unibox (finally got it to work) is the affect of port resonance. Is it a big problem?

Image not available

Here is a bigger version of the drawing:

http://clubpolk.polkaudio.com/forum/attachment.php?s=&postid=246670

[JonMarsh]

This looks like a good plan. A surprising number of newer speaekrs are coming out with greater depth than height or width. From the viewpoint of having adequate port length and absorbing the midrange portion of the driver rear wave, this is good. The enclosure can be lightly stuffed with Dacron polyester, which helps kill the mid rear wave, but doesn't lower the box Q very much. I also use Whispermat on the rear walls, sides, and top/bottom.

The port resonance is down quite a bit in level, and rear firing it minimzes the audible impact, as opposed to say, having a front firing port.

Since I'm running behind a bit, here's the info for driver positioning:

Woofer is centered in the baffle, 7" from bottom edge.

The tweeter should be between 4" and 4.5" from the top, offset from the center by 1", mirror imaged between the two cabinets.

Double check against driver diameters, you may need to nudge the tweeter up just a little to have at least 1/4-1/2" clearance from woofer.

This combination of offset locations distributes the boundary loading and diffraction effects to get the smoothest response on this baffle; here's the BDS predition of the baffle loading and diffraction effects.




~Jon