Adire Extremis MT

[Al Garay]

By the way, Al Wooley (www.rawacoustics.ca) will have new kits using the Extremis 6.8 and OEM Arum Catus G2 ribbon for $459 in USD.
I visited his house yesterday to pick up my new DEQX and got a chance to listen to many of the kits coming out with the Extremis and the 4.5" XBL driver. Most are using a soft dome tweeter made by Usher, looks identical to the NorthCreek D28. Al will be using the G2 ribbon with his kits. Overall, these have a very smooth, with good depth even when driven by inexpensive Marantz 5300 receiver and a cheap Pioneer dvd player. The G2 ribbon had more of the expected top-end sparkle, more forward.

[Davey]

Oh my gosh Al.....a new DEQX.....you've gone over to the dark side.

Davey.

[Al Garay]

I put all my eggs into my 2-channel system. I rationalized it thinking that I will not have to call you asking questions about your measurement system (already has it built in) with software wizards and auto-eq (have not tried it yet). No more questions about 6-channel preamp with remote control for more than 2 inputs so I can control the volume in my music source and movies. Built-in DAC and most importantly, it has a black faceplate.

[denverdoc]

Al,

Welcome to the DEQX club. The only downside to this most remarkable piece of gear I can see is that soon you will be thinking about 3 of these to run a ne plus ultra HT set-up!

Do make sure to join the users group as you will likely find that altho well written, the manuals can only take you so far. By the way I have tried 3 very different speakers ranging from 200/pair JBL's to a DIY using over 3K in drivers and in each case, the improvement ranged from blatant to jaw dropping. I will never ever use again any speaker employing a pasive XO unless its for the bathroom or computer!
John

[Al Garay]

Thanks. Let's continue the DEQX discussion in the AC thread which now seems to belong to another John. Back to Extremis maximus and the quest for true measurements and good music.

[denverdoc]

Well send my regards as I no longer participate at AC for a variety of reasons. Certainly not trying to hijack this thread as the other John seems wont to so--and not an alias, post was simply a potentially helpful sidebar as to what has been helpful for yours truly, and a reminder to all that the days of XO torture may soon be behind us :T
John

[GrahamT]

Here is the latest measurement. Ungated, 0.4m far field, MLS.

[JonMarsh]

My suspicion is that the roll off in the very top end may be due to the RS mic. How does the high end balance sound to you, Graham?

Remember, here's the revised Modula MTM with almost the same crossover on the 27TDFC (1400Hz instead of 1200Hz)- check out the Modula MTM thread.



The top end on the 27TDFC certainly isn't dropping off.

Is this last measurement you posted after tweaking the LPAD since your post about resistor values, or with the same values you feel it's a little bright for?

[GrahamT]

The top end is not dropping off, for sure it's my mic.

I think that maybe the tweeter is just one or two dBs louder. They sound great, very detailed and clear, but after extended listening I think the tweeter level is just a little too high.

I'm going to play around with some Lpad calculators like the one you linked me to and try that in speaker workshop. I'll let you know what I find.

The rest of the crossover is right on. The BSC is just right and the transistion is very smooth. In a big room like at the DIY meet, people thought the tweeter mid balance was just right.

[JonMarsh]

Depends on how much absorbtion or early reflections are going on- smaller room, nearer side walls, power response will seem a little brighter regardless of anechoic balance. Tune until it sounds right; probably just lower till the 3 k level is the same as 500 Hz, for example.

[Kevin Haskins]

By the way, Al Wooley (www.rawacoustics.ca) will have new kits using the Extremis 6.8 and OEM Arum Catus G2 ribbon for $459 in USD.
I visited his house yesterday to pick up my new DEQX and got a chance to listen to many of the kits coming out with the Extremis and the 4.5" XBL driver. Most are using a soft dome tweeter made by Usher, looks identical to the NorthCreek D28. Al will be using the G2 ribbon with his kits. Overall, these have a very smooth, with good depth even when driven by inexpensive Marantz 5300 receiver and a cheap Pioneer dvd player. The G2 ribbon had more of the expected top-end sparkle, more forward.

Al.... I'd be interested in your thoughts of the Extremis plus the Usher. That is pretty close to the final design and I'm trying to get as much feedback as possible.

[GrahamT]

After tweaking the Lpad a little and doing some more listening I've decided this version sounds the best to me in my room. They sound just lovely. I'm done tweaking for now.

The low end response is incredible aswell, I dont even bother turning my Tempest on anymore.

Here is the final version of the crossover that Jon posted before.



I might make a new thread with the important details of this project because a couple of people have told me they are interested in building it aswell.

[JonMarsh]

Thanks for the update, Graham. Starting a new thread with all the details from your experience and a summary of the design would be a great idea! Hope you have the time to do that!



Jon

[Al Klappenberge]

Jon,

I think we are doing similar stuff:

Take a look at my "extreme-slope" crossover networks for Klipsch speakers:

http://www.alkeng.com/klipsch.html

Al K.

[JonMarsh]

Hello Al, and welcome to HT Guide!

I was just over at your site looking at your Klipsch crossover designs and whitepaper. It looks like you have a real love love affair with the big Klipsch models and have done some interesting things to upgrade them. Have you been working with these speakers for a long time?

My father was an Episcopal minister, and for several years in the 60's we had a church over in Hope Arkansas.

I think it's curious that there haven't been more folks using this filter type in audio, but then I first tried it out in a class D amplifier output filter design in the late 80s, and only in the 90's started applying it to speaker crossovers.

Interesting reading!

~Jon

[Al Klappenberge]

Jon,

I have had a set of Belle Klipsch since about 1975 or so. Having been L-C filter product line engineer at K&L Microwave (a filter company here in town) I started looking at the crossover in them and realized they were simply an audio "triplexer". That's something I had been designing at microwave frequency for years. Since I have (and sell) the software to synthesize filters like this (zero - placed or elliptic), the future got sort of obvious! It wasn't until I actually built an extreme-slope set of networks I realized what a step forward they are. There's two reasons they are not used very often. The first is their expense and the second is people's fear of "phase" destortion. Of course designing them separates the men for the boys a bit too!

Al K.

[JonMarsh]

Jon,

The first is their expense and the second is people's fear of "phase" destortion. Of course designing them separates the men for the boys a bit too!

Al K.

Well, commercial speaker crossovers at times strike me as being akin to trying to build (or rather, market) what is described as a "V-8" but only seems to have enough parts for a V-6 or inline 4 at best... so from that viewpoint, I understand the expense comment, though these implementations of cauer-elliptic networks are really only slightly more complex than a typical 4th order L-R, which is common in "high end" speakers. But the flexiblity in achieving one's goals, compared with adding an impossible number of components in a conventional ladder network topology, gives all sorts of advantages, as I see you understand also.

and then there's your second statement, too....

~Jon

[Al Klappenberge]

Jon,

Absolutely! I have heard about first order networks with longs strings of all-pass equalizers that they THINK will compensate for time dealy errors. I have done a few experiments that have convinced me that my ears (at least) are totally def to phase. With the extreme-slope networks, only one driver at a time is making the sound. No combing (almost) and no multiple copies of the same sound getting to your ears. The difference is quite obvious. So far, everybody who has installed one has commented on how the stereo image has opened up. Even if they do cost more, they are worth it!

Al K.

[GrahamT]

Okay, I said I was done tweaking but I just cant help myself and leave well enough alone. Maybe I'm neurotic, or maybe it's the solder fumes? This is a lot of fun. I guess I'm addicted now.


I'll start a new thread like the Modula when I finally make up my mind.:lol:

[GrahamT]

I'm almost done. The only thing is the tweeter padding and that is a lot of personal preference. It's pretty close right now.

[KevinP]

Greetings Mission Possible DIYers,

Great forum here. I’ve been doing so much reading here the last few weeks that I feel like I know some of you personally.

I have built a few speaker kits for friends and family, so I sortof know my way around this stuff. This project has me thinking that now its time for MY speakers.

There are a few modifications I am thinking of that I would like to hear some opinions on.

First I would like to make this into a tower. I think a tower would be way more volume than the Extremis needs so how should I consume that space? I’m thinking make it a transmission line, maybe tapered. 1/4 wave of a 25-30ish Hz tuning is a 10ish foot line. I'm thinking 10ish feet folded could be towerish size. Is this just stupid?

I really like the look of the M8ta enclosures so that’s what I want to shoot for looks wise. Maybe without the backward slope, or maybe with, minor detail there I think

Looking at the crossover here:

https://www.htguide.com/forum/showpost.php?p=145035&postcount=237

It appears to me that BSC is happening at the crossover frequency and just padding the tweeter down for it? Is that correct? If so I think I would like to shoot for that type of approach. Am I being stupid here and creating a mess for myself?

I have Speakerworkshop working good enough to get repeatable measurements of things and I'm not afraid of taking lots of time to get this done. I ordered the woofers and tweeters yesterday so the trigger has been pulled.

Am I stupid? Be honest.

tia,

Kevin

[GrahamT]

Welcome Kevin.

If you make either design (a tower or TL) you will need to adjust the BSC. I dont know how that works in a crossover with elliptical filters though.

You could make a ported tower with the Extremis with a down firing port a la the M8ta. You could shoot for a volume of about 35 to 40 L tuned from 25 to 30 Hz.

BTW a couple of people are working on a TL over at DIYaudio.

Am I stupid? Be honest.

Not stupid. But maybe just crazy like the rest of us. :

I'll make a suggestion though, try building this design with a test box like I did. Then decide what you want to build later. I think if you decide to build a tower, TL etc you may only need to change the first inductor, but I'm speculating.

I'm still tweaking the Lpad. I think I have it done but I'll post a couple of versions in a new thread with a parts list when I am finished.

[KevinP]

Hey Graham,

Thanks for your input and speculations. And thanks for the pointer to the diyaudio Extremis TL efforts, I'll be watching that progress.

I really think I'm settled on a TL, even a big 40l ported doesnt seem to become towerish dimensions like a 9-10 foot folded line could. I think....

[Kevin Haskins]

Personally I'd build the MMT Extremis Dan designed. It sounds fantastic and would be much easier to build. It is tower size and has great dynamic range...better than what you will get with a single driver. Two drivers helps with the effiicency and the low frequency output is incredible. Great midrange also.... contrary to what you might think based on some other reports.

[KevinP]

Hi Kevin,

I'm not familiar with the design you speak of, and two extremis per side is what I really wanted to do. MTM style probably. However I am electronics challenged with an AV reciever that limits my speaker load to the 8 ohm range. At CES it was hinted at that a dual 8ohm VC Extremis was on its way, I tried to contact Adire about it for an ETA and recieved no response. Plus this design seems to take into account certain things that I have always felt to be important and I've been trying to come up with a nice project for myself for some time now so... but I cant make anything easy on myself, thats just not in my nature. ;D

[Kevin Haskins]

Go for it then! The process is a good part of the fun.

The MMT is going to use two 8 Ohm drivers. I doubt there is going to be a 16 ohm version anytime soon. Most amps are 4 ohm stable and the dual 8 ohm version is an easy load hovering around 5-6 ohms across most of it's frequency range. ETA for the MMT is soon.... very soon.

[GrahamT]

Here is what I am happy with right now. I have tried about a dozen different Lpads and like this one the best. It also measures pretty flat. I changed R7 to 8 Ohms and R3 to 3 Ohms. I came up with these using Speaker Workshop. Call me crazy, but it sounds damn good.

Here is 1m ungated MLS. The relevant frequencies are in between the markers. The response doesn't really drop off on the high end, they are probably pretty flat to 30 kHz.

[DWiggins]

Hi all, got a ping to check this thread out... Lots has happened!

Graham, glad you like the design and drivers! Looks like John did a great job on the XO; not too complex on the woofer.

Have you listened to the woofer raw, by itself? It's something we push hard here - a quick listen to a naked driver will quickly turn up any audible problems, and gives you a good idea of how a driver will sound. You have the perfect test case - disconnect the external crossover!

Also, here's a quick shot of the FR (spliced at 300 Hz) for the MMT.

Image not available

FR is +/- 1.5 dB from ~25 Hz to 20 kHz. The "depression" below 300 Hz was left there on purpose, since those tend to be "bloom" frequencies for most 8-10 foot ceiling rooms. Especially when the speakers are placed within a foot or two of a boundary (as is usually the case). In a really dead room or pulled out, it doesn't sound hollow, but in a live room it does tame the bass wooliness a bit.

Impedance is pretty clean as well; it has a minimum of 5 Ohms at 3.5 kHz, and averages ~7 Ohms from 20 Hz to 20 kHz.

Image not available

I'd consider this speaker a nominal 8 Ohm load; I'd have no qualms about running this with any 8 Ohm only amp. The "price" paid for the low voltage sensitivity (NOT efficiency) is the high impedance of the driver, which means that a pair of them will yield a nominal 8 ohm load without much work at all. Phase is also well controlled, staying +/- 25 degrees from ~80 Hz and up.

There's quite a few new designs that will be rolling with the Extremis and OEM variants in the next 3-6 months, from some of the very big names in the DIY/boutique community. Rather than the couple of clients showing Extremis-based speakers this last CES (3 months ago), there should be a couple dozen next year.

Dan Wiggins
Adire Audio

[ThomasW]

Hi all, got a ping to check this thread out... Lots has happened!

Caught redhanded..... :E

And all this time we though this was occuring well under the radar ...:wink:

[Kevin Haskins]

I took some measurements this morning of the KIT61 with the Usher. These are all 1/6th octave smoothed and taken from 1M. Measurement gear is Liberty Instruments Praxis & autopod with M-Audio Transit & calibrated mic (good out just past 20K).

On axis 1.0M

Image not available

Measured 1.0M 15deg off-axis above the plane.

Image not available

Measured off-axis 30 deg to the side.

Image not available

And finally 45 deg off-axis to the side.

Image not available

The low frequency data below 500HZ isn't representative.

[GrahamT]

Graham, glad you like the design and drivers! Looks like John did a great job on the XO; not too complex on the woofer.

Have you listened to the woofer raw, by itself? It's something we push hard here - a quick listen to a naked driver will quickly turn up any audible problems, and gives you a good idea of how a driver will sound. You have the perfect test case - disconnect the external crossover!

Dan Wiggins
Adire Audio

Thanks Dan,

I did listen to the drivers raw like you told me over at HTT. I must admit I cant really hear any grit on the top end. I'm not a golden ear though, maybe I dont know what to listen for. I'll give it a more thorough test, it's easy enough to bypass the crossovers.

[capslock]

Mark, can you confirm that your 7 in. driver tests were also done with a mic distance of 0.5 m? What kind of gating did you use?

By the way, I have some interesting distortion measurements of Dayton RS vs. Excel, PN me with an email address that will hold large files if interested.

[Mark K]

Mark, can you confirm that your 7 in. driver tests were also done with a mic distance of 0.5 m? What kind of gating did you use?

By the way, I have some interesting distortion measurements of Dayton RS vs. Excel, PN me with an email address that will hold large files if interested.

My tests were done at 0.25m and unbaffled for nonlinear and 2 inches from the plane of the driver for the linear and FR, IIRC.

[Shawn Solar]

Man I just read the entire thread...I gotta say I'm getting impatient. My kit61s have been sitting here for two days and there not done yet lol. By monday I'll have kit61s, kit41s and gr research av2s to listen to...can't wait. I had the kit41 hooked up today. I gotta say not bad at all! Doing some research on some speaker kits those Daytons seem very enticing as well..lol it never ends