S16 - Constant Directivity Dipoles

[gainphile]

This is an investigation whether the high frequency response of typical narrow dipole radiators can be improved by employing a waveguide (like Physics CS2)

Gainphile: S16 - Constant Directivity Dipoles



I was pleasantly surprised to find out that cheap waveguides like Dayton 10" and Econowave measured impressively. They had opened the door for me to investigate as I thought only Dr. Geddes could make those plots. Hence thanks to Zilch and folks at AudioKarma.

Graph:
Dayton 10" Waveguide plot
Dayton 10" Waveguide sonogram
Econowave plot

The dispersion pattern of those waveguide surely will be different and so is the phase, but they should be better than typical back-to-back domes.

First, I measured the dipoles + forward firing Waveguides. These are indoor takes and there are plenty of artefacts.



We can observe the dipole directivity transition which goes for -6db@60deg to -6db@45 deg for the waveguide. The transition is very smooth. There is no blooming as normally experienced by back to back dome tweeters.

Graph:
dome tweeters sonogram (O.P.)

I can almost see the graph as gedlee speakers without the transition to omni in lower frequency, or Orions without the irregular tweeter response. A "best of both world?"
(I saw Orion's plot before in Gedlee plot, but could not open it anymore due to app error, anyone has it?)

The sound of this system at this stage is very smooth. One needs to get used to the waveguide sound as they are nothing like dome tweeters and intial impression would be rolled off highs. They do not sound bright even when equalised as flat (1m). Yet there is a big hole in te reflected spectrum which makes the system sounded unbalaced. This could not be rectified by changing the tweeter level.

The obvious solution is a rear firing waveguide, however this is challenging in terms of installation and aesthethics, so I use dome tweeters instead on the back, firing out of phase. I assumed the rear radiation would be diffused enough when listened from the front.



The tweeters become omni directional as it approaches the xo frequency and influence front radiation. The level is quite low at -10 to -12 db but no excuses here they should not be there. A possible solution might be to use very directional ring radiators or one of those Seas DXT tweeters.

The sound however is correct and "full". Like those dipoles with back to back domes, but nothing of the usual signature which is impossible to equalise and caused much questionable qualities.

From the data and subjective impression I see this as a possible approach for great dipole speakers and further improvements may be made with correct type of rear firing tweeters.

[Saurav]

I tried a WG + dipole midrange/woofer and was never quite satisfied. I toyed with the idea of a rear-firing tweeter, but then made the leap to the Neo3 / dipole tweeter. Looks like you're getting very good results with the rear tweeter. The second sonogram - is that the rear radiation? Or the front radiation with the rear tweeter active?

[mige0]

Nice plots.
I'd be interested in what you are coming up with regarding "illuminating" front and back equally in the upper range

XO somewhere at 1.5kHz ?

Ever thought of cutting two Dayton-Horns down to roughly fit a NEO ?

Michael

[Patrick Bateman]

While I ended up buying Gedlee Summas, I *did* evaluate a set of dipoles with waveguides. This was at CES back in '07 or '08 IIRC. I have some posts about them in this forum. Can't recall the name ATM.

They were very VERY good, but not quite as good as the Summas.

I also enjoyed a set of German studio speakers from Klein and Hummel.

One of these days I'd like to build something like a bipole Summa. I've tried running the Summas back to back (in mono) and the improvement in the "room sound" is noticeable.

Of course, using four Summas would require a lot of $$$ and a big room, but that's the law of diminishing returns for ya.

[dabbler]

Originally Posted by Saurav

I tried a WG + dipole midrange/woofer and was never quite satisfied. I toyed with the idea of a rear-firing tweeter, but then made the leap to the Neo3 / dipole tweeter. Looks like you're getting very good results with the rear tweeter. The second sonogram - is that the rear radiation? Or the front radiation with the rear tweeter active?

Hi Saurav,
What was it about the WG that you werent happy with? I have found the compression driver / WG combo to be similar but more satisfying then the Neo 3. The Neo 3 is clean and transparent but lacks dynamics in my experience.

[gainphile]

Originally Posted by Saurav

The second sonogram - is that the rear radiation? Or the front radiation with the rear tweeter active?

That is the front radiation with rear tweeters active. As seen, it influences the dispersion pattern somewhat.

[gainphile]

Originally Posted by mige0

Nice plots.
I'd be interested in what you are coming up with regarding "illuminating" front and back equally in the upper range

I haven't planned anything yet and will see what options will be available. It's quite a construction challenge. But it also could be as simple as putting smaller MCM waveguide on top of the dome tweeters.

[Saurav]

Originally Posted by dabbler

Hi Saurav,
What was it about the WG that you werent happy with?

A disconnect in the soundstage, where the treble was always forward from everything else. Granted, this was mostly before I got the DCX, I didn't spend a lot of time with that driver combination and the DCX.

The Neo 3 is clean and transparent but lacks dynamics in my experience.

Yes, I definitely miss the dynamics. Maybe some day I'll try a WG + rear dome / bullet tweeter combination. I'm still not convinced about how important it is to have "true dipole" response up into the treble, vs. just have a reasonable power response match. I definitely haven't tried both ways for myself.

[gainphile]

Originally Posted by Saurav

A disconnect in the soundstage, where the treble was always forward from everything else.

This is what I observed when I had no rear tweeters.

It's due to transition between dipole radiation to forward directional.

Once the rear tweeter is added, the soundstage moves back behind the loudspeakers and "full" as I described it. The speakers "dissapear".

[twest820]

Nice build. Be interesting to see how it settles in over time. The 8W4P two way I ran for a while had essentially the opposite directivity issue on the rear wave. That was my first full dipole so it took a while for the directivity glitch became audible. After about three months it was starting to be annoying, but then I'm pretty fussy about directivity.

I assume it's a three way. Are you using the fourth channel on the MiniDSPs to time delay the rear firing tweeter to time alignment with the rest of the back wave?