Dipole idea...

[augerpro]

So Javier's dipole thread inspired me to buy a bunch of drivers I didn't need to build a speaker I wasn't even considering! Dang you Javier! Anyway I've wanted to dipole for awhile and even with limited shop access I figure I can whip something up fairly fast.

My idea is a 2.5 way MTMMM or MMTMM using Zaph ZA14's and Aurum Cantus AST2560 which can be run dipole (and is well behaved). These will be crossed to small subs I already have between 100-300hz. From a discussion I had with John Kreskovsky a number of years ago there is a big advantage to crossing a dipole mid with monopole woofer using LR filters as it creates a cardiod response at crossover, which is perfect when crossing 100-300hz because that is right where the back wall tends to muck things up. Prototyping will be done with a minidsp as always, but it will probably stay active once I'm happy with it.

Looking for feedback on which layout to use in regards to room interaction. I like the symmetric MMTMM intuitively, but I do wonder just how well those upper mids will blend with the sub (and just how cardiod the pattern could then be). MTMMM is better in this regard but has a drooping main lobe in the midrange.

Also curious about how to terminate the edge of the baffle?

Any other dipole advice appreciated. I do have John K''s dipole sim and rereading all his stuff now.

I'm actually really looking forward to this, I had used the AST2560 and ZA14 in a driver shootout a year and half ago and really liked them both, maybe the best of the drivers I used in that shootout, but neither fit the requirements of the resulting designs. I already had the AST2560 and John Krutke had a number of QC samples sitting around so these will actually be fairly cheap to build.

[5th element]

Actually if you think about it and visualise the radiation patterns it's rather easy to understand.

At low frequencies the bass module would be acting as an omni directional radiator throwing sound everywhere. The dipole module on top would be throwing sound forwards and backwards with the nulls at either side. The sound being thrown forwards by the dipole and bass module would interfere constructively and create a large bubble of forward going radiation. Around the sides output would fall and then around the back you'd have cancellation with the main panel.

[fjhuerta]

I never had any real issues other than the bass boost you'll need to integrate the OB to the VB. Anything from 160 to 120 Hz works very well. The ZA14 is pretty much equivalent to the RS-150's I'm using.

If I could build them again I'd definitely do an MMTMM in a 2.5 way configuration! That, and a sub, would be a definite improvement over my design, I think.

[augerpro]

Why just 120-160hz? Any other design and construction tips you have, I'm all ears!

[fjhuerta]

Why just 120-160hz? Any other design and construction tips you have, I'm all ears!

Over 160Hz made bass sound muddier, and male voices started coming out of the bass module.

Under 120Hz made the little woofers work too hard, and I actually bottomed them out.

Then again, if you are doing an MMTMM, I'm sure you can lower the frequency response to... 100 Hz? 80 Hz? And cross them over to a sub. Just thinking about it makes me wish I had thought about it before building my own speakers!

[augerpro]

Makes sense. I'll have to look at the numbers in John K's sim to see if I'll have xmax issues so low. When I played with the ZA14's last year I remember that they sounded clean right until the point they bottomed out with a pop. Need to make sure I'm nowhere near that!

Any thoughts on MTMMM vs MMTMM?

[augerpro]

So as I understand it, a dipole with a flat on-axis response will radiate less power as you go down in frequency versus a monopole. In fact it should be flat sound power response versus the typical sloping response of a monopole in-room. Does this imply that the overall on-axis response should be tilted down? I know that the preferred tilt found by Harman varied according to directivity so perhaps I shouldn't expect that in-room dipoles should also have this downward slope.

Whether or not it needs this tilt, the ability to do this does seem useful. How would one implement this in a minidsp? I assume you have to directly calculate the biquad filter but I don't see this ability in the worksheet at their site.

[sfdoddsy]

Linkwitz recommends a treble roll-off for his speakers. You can achieve it with a Minidsp by using a shelving filter on the input.

[Juhazi]

There are many simple ways to tilt overall response in minidsp. Both input and each "way's" responese can have several filters. Biquad programming is possible but not necessary.

When we discuss of dipoles, it s imporant to know the real measured directitivy "profile" of the speaker. There is no such thing like fullrange dipole! Alas placement in the room is even more important than with monopoles. It is not wise to give general rules.

This is how my speakers measure in my room now. On-axis at 1,5m is straight when speaker is in middle of the room. Two mic positions, REW sweep 1000ms window. Bass is monopole up to 180Hz LR2.

[augerpro]

There are many simple ways to tilt overall response in minidsp.

Such as? I don't think shelf filters are really a solution, unless you can use super low Q values, but minidsp Q values are limited to .5 I believe.

[Juhazi]

I have Advanced plugin and minidsp 4x10HD. Easy and fun!
More here http://www.minidsp.com/applications and here http://www.hifizine.com/2010/12/prot...e-minidsp-2x4/

Here I show 2 active settings, both high shelf -2.5dB Q=0.5 Easy, just don't boost output over 0dB level!
Notice that 3 other settings are left passive
-
ps. This is just an example, I don't use this!

[augerpro]

I'm really looking more for a straight line EQ. But I suppose a shelf filter with PEQs at the corners to flatten things would work...just seems so in-elegant