[Nyal Mellor]

So I'm going to build an OB! Would like some assistance with driver selection to meet my design specification:

- two channel system
- active amplification
- digital crossover and time alignment
- (heresy) using subwoofers crossed over at 80-100Hz
- dipolar above the subwoofer (i.e. radiates equally to front and back)

I currently have a Linkwitz Orion, and this will be a long term project to try and improve on it. After living with it for quite some years (5 infact) and comparing it to a number of very high end (100k+) systems, I feel that there is some improvement potential to be made in:

- the low bass where the Orion is displacement limited
- the midrange / treble, where the Orion has a certain 'grain'
- imaging, where the Orion is slightly 'diffuse' and does not possess as much of a 3D character as I have heard from say a standmounter

By using high slope crossovers I hope to work around some of the challenges that limited Linkwitz in terms of choice of XO frequency, driver displacement issues with the tweeter and out of band resonances from the midrange.

I am not wedded to dynamic drivers and potentially there might be a nice configuration that could be arrived with via ribbon or electrostatic means.

[ronpod]

Nyal,

If you haven't already seen the version 3 update to the Orion
http://www.linkwitzlab.com/orion-rev3.htm
you might take a look. You may get some of what you are looking for for a $20 swap of electrical parts...

At any rate, best of luck in your journey.

[D OB G]

Hi Nyal,

As far as a ready to go design is concerned, JohnK (John Kreskovsky) of Music and Design has a new design coming out before Christmas, I think.
It is called the Nao Note (you may well have heard of it - however it is three channel above the woofers).
Makes use of logically selected recently released drivers on a logical baffle.

With the Orion, do you use the Thor subwoofer (or maybe subwoofer(s))?

I am currently progressively updating my subwoofers to four sealed boxes, with a total of eight 10" XXLS Perless drivers, so no displacement limitation, but the point is that they can be used up to 80-100 Hz.  The current Nao crosses to the "sub" at 120Hz.  (I don't think one, or even two Thors might be enough , but then I haven't tried them!- depends on the size of your room of course).

Rudolf, and others, e.g. JohnK, make the point that for each driver the baffle should be an absolute maximum (and preferably less) of 2.2 times the fuctional diameter of the driver for decent polar response (the Orion seems to meet this criterion).

If you are using FIR filters (I am), an even polar response is absolutely crucial.  It may be that this is difficult (impossible??) to achieve with a two-way on an OB.  (Geddes believes that the Orion doesn't measure with an even polar response pattern- maybe due to the way the tweeters are implemented).  (Geddes claims though that the 200-500Hz range of the Orions may be the best he's heard, and it's hard to get his praise!).

Toole believes (I think) that "spaciousness" and pinpoint imaging are polar opposites on a spectrum.  This might be what you call the "diffuse" character.  i.e. if you sit more or less in ther far-field, then the 3D that OBs are capable of is a spacious 3D.  From my experience, the sort of imaging a boxed monopole provides is just always going to be different (but then I haven't heard a million speakers!!).  (Mind you, StigErik on DIYAudio, who sits in the near field- less than two meters I believe!!!, seems to get the best of both worlds).

It's interesting that you can hear the mid's break-up resonances (I take it), as I've always been sceptical about the potential success of his notch filter.

As ronpod says, the improvements with Orion 3 may well eliminate the "grain" you describe.

I would emphasise again that if you are using FIR filters, polar response is EVERYTHING.  And getting correct time alignment at the front means time alignment at the rear is made worse (in practice, I personally don't think that is a problem- for various reasons).

With your long experience with the Orions, my comments may be totally superfluous, but here they are anyway for what they're worth. 

[Nyal Mellor]

Nyal,

If you haven't already seen the version 3 update to the Orion
http://www.linkwitzlab.com/orion-rev3.htm
you might take a look. You may get some of what you are looking for for a $20 swap of electrical parts...

At any rate, best of luck in your journey.

Hey Ron, I have the parts sitting on my desk, just need to find some time to get them in! Not sure the tweaking is going to sort the bas out...

[Nyal Mellor]

My simple first idea was basically to take the same drivers and baffle as on the Orion and see what improvement could be made by crossing over to subs and using finding a way to use higher slope filters on the drivers whilst maintaining polar response. But I'm quite up for something else.

[PDR]

You might want to see if someone in the Bay area has a set of these to audition.
I have many different speakers.....from arrays to 60" ribbons on quads of 12' OB subs.

The pair I own are the little brothers of these....they are absolutely stunning,
and seem to fit some of your criteria....


http://www.audiocircle.com/index.php?topic=72715.0

[Rudolf]

Nyal,
I would see two areas where you could really improve the Orion concept:

1. distributed subwoofers (NOT as bases for the dipoles): http://mehlau.net/audio/multisub_geddes/

2. dipole tweeters without baffle like done by StigErik and others, preferable planars like the BG Neo3 or Beyma/Mundorf AMTs. http://www.diyaudio.com/forums/multi-way/142015-my-open-baffle-dipole-beyma-tpl-150-a-82.html#post2132463
SL surely has achieved the best possible with two conventional tweeters, but they still work completely outside the range which should truly be called "dipole".

[Nyal Mellor]

Rudolf

I couldn't agree more!

Bass - yes it will be monopole multi-sub with lots of EQ. I have found that home theater type bass - lots of power and EQ'd flat frequency response is much more satisfying than dipole bass. And more importantly it solves the issues 1) compromise between best location for bass and best location for soundstaging 2) speaker boundary interference related suckouts due to phase cancellation from forward and backward wave from speakers 3) in the Orion in particular, lack of surface area and need for insane amounts of EQ. It generally can't be a good thing to add 32dB of EQ (or whatever it is) to sort out the dipole cancellation at bass frequencies.

Tweeter - was looking at the BG Neo 3 PDR...The AMTs that I've looked at in the past had TERRIBLE off axis response! Have they improved?

[Nyal Mellor]

NO baffle! now there is an interesting thought!

We need a new forum - the NO baffle forum

[Rudolf]

Tweeter - was looking at the BG Neo 3 PDR...The AMTs that I've looked at in the past had TERRIBLE off axis response! Have they improved?

Look at those new AMTs with small neodymium magnets and removable back covers, like the Beyma APT 150 http://www.beyma.de/fileadmin/seiten/download/pdf/Beyma_2010/TPL-150_en.pdf or the Mundorf 2310 or 2510 http://www.hs-devices.com/english/amt.htm. They should be used without the face plate to keep the profile as narrow as possible. That way the radiation pattern will keep the dipole figure 8 to much higher frequencies than the Orion+ can.

[Nyal Mellor]

I will look at those - the only AMTs I measured were like 20 years old and they were BAD off axis.

Currently thinking BG Neo3 and 2 Seas W18 Excel drivers in a two and a half way with high slope XO between Neo 3 and Seas around 1.8k. It'll be tight and may be a complete flop

I can understand why people end up with 3 ways, doing a 2 way OB with good polar response is challenging. But then again most of them don't have steep ass active XOs. Hmm. More thinking required.

[Nyal Mellor]

Quote from Beyma: "In these conditions, the horizontal dispersion of the TPL-150 is completely
awesome."  I didn't know 'awesome' was a term typically used in audio engineering circles

[Danny Richie]

I went down this same road not too long ago and did something almost just like that.

Currently thinking BG Neo3 and 2 Seas W18 Excel drivers in a two and a half way with high slope XO between Neo 3 and Seas around 1.8k. It'll be tight and may be a complete flop

I still have the test baffles, measured responses, etc.

[maxro]

- digital crossover and time alignment

Won't electronically correcting for misalignment of acoustic sources create double the original time alignment error as viewed from the back side?

[Nyal Mellor]

Maxro, no it should fix the problem, basically we are delaying the tweeter to align it with the acoustical center of the woofer. It will be aligned on the woofer from either direction.

[Nyal Mellor]

I went down this same road not too long ago and did something almost just like that.
 

I still have the test baffles, measured responses, etc.

Please send to me!! Thanks!! Both the Beyma and the BG Neo look good, infact the BG Neo has better CSD but the Beyma seems like it would work no problem in a straight two way, whereas the BG Neo looks a little iffy.

[maxro]

Maxro, no it should fix the problem, basically we are delaying the tweeter to align it with the acoustical center of the woofer. It will be aligned on the woofer from either direction.

Methinks this needs a rethink.
OK, so the tweeter is physically forward of the woofer and needs to be delayed, yes? Now, turn the baffle around. The tweeter is now behind the woofer and it is the woofer which needs delay. It would take quite the circuit to solve that pickle. Of course physical alignment solves the problem without the paradox.

[Danny Richie]

Maxro is right about the driver alignment.

Please send to me!! Thanks!! Both the Beyma and the BG Neo look good, infact the BG Neo has better CSD but the Beyma seems like it would work no problem in a straight two way, whereas the BG Neo looks a little iffy.

Nothing iffy about the BG if you know how to handle it.

I think you guys would get a kick out of hearing what I tried, what worked, and what didn't. Since the whole thing never became a product and was just research that led me somewhere else, then I would imagine no one will mind me posting it here.

I'll upload some pics and elaborate. Give me a little time.

[Saurav]

Add me to the list of people interested in your design study.

[Saurav]

Maxro, no it should fix the problem, basically we are delaying the tweeter to align it with the acoustical center of the woofer. It will be aligned on the woofer from either direction.

As maxro said, this is not correct. An electronic delay doesn't move the location of the tweeter, it makes the tweeter's sound come out later, regardless of which side you're measuring. I've had to do this with my speakers, and it helps the on-axis response, but makes things worse to the sides and the back.

Check John K's site (www.musicanddesign.com), he has an article where he discusses this.