Comparison Bohlender & Graebener Neo3W and Neo3PDR in dipole mode

[keyser]

Today I did some measurements on the B&G Neo3W and Neo3PDR. Biggest differences:

- Neo3W has approximately 3 dB higher sensitivity
- Neo3PDR has better dispersion in the top octave
- Neo3W has a smoother transition from dipole to beaming

EDIT: too bad, the pdf is too big to upload. Try this link: [removed link]
I hope it works.

EDIT2: apparently not too well. I'll have to try something else.

EDIT3: I've got a zip uploaded, but without the pictures.

EDIT4:




http://img843.imageshack.us/img843/6...ementsetup.jpg
http://img651.imageshack.us/img651/5...thtweeters.jpg

[twest820]

Quite the bloom on the PDR. The sonograms are derived from the 15 degree data?

[keyser]

The sonograms are derived from the same data as the frequency response curves: 0, 15, 30, 45 and 60 degrees. I omitted the data at greater angles, as the exact setting of the angles becomes very critical if you want to be able to accurately compare the drivers.

BTW, I've been using the Neo3W up till now, but I think I'll give the PDR's a second chance.

[cuibono]

Nice work, this has needed doing for a while.

If you're interested, try adding foam edging front and back to the PDR, about 15mm thick total. It helps smooth things around 7kHz, but preserves the top octave performance.

[erjee]

Nice job!

You can try adding some foam/felt/etc on the outside of the Neo3W to create a narrow source for high frequencies, similar to the Neo3PDR. You might be able to reduce the beaming in the top octave while avoding the bloom at 7.5 kHz.

Good luck!

[Rudolf]

Keyser, this is a very informative comparison!

From your 3rd picture in the pdf I see a difference of 5 dB at quite some places. Since the PDR has foam added behind half of the openings, they appear to be almost blocked (total blocking would result in 6 dB difference imho). I would have hoped for a less radical solution by BG.

In consequence the active driver width (source) has been halved, while the baffle width has stayed the same. The differencies between the polars very nicely document the difference between a wide source and a smaller source on the same baffle:

-the baffle width dictates the position of the dipole notch (6 kHz in this case)
-the smaller source will start to beam later than the wide one - regardless of baffle size.
- this opens the "blooming" region between the dipole notch and the onset of beaming. Polars of the ORION for instance show the "excessive" (compared to the Neo3) width of that region clearly.

Originally Posted by erjee

You can try adding some foam/felt/etc on the outside of the Neo3W to create a narrow source for high frequencies, similar to the Neo3PDR. You might be able to reduce the beaming in the top octave while avoding the bloom at 7.5 kHz.

As desribed above: By widening the baffle you draw the dipole notch to lower frequencies. You get the same width of the "blooming" region as with the PDR, with everything starting at lower frequencies only.
The most elegant (I would say "best") solution would be to have the widest source in the narrowest baffle.

Rudolf

[Saurav]

Now I too own a pair of PDRs as well as regular Neo3's. I'll try and run equivalent comparisons like this some day. Subjectively, I prefer the sound of the PDRs, even though the measurements show a deeper notch. My theory is that the wider upper treble dispersion somehow works better in my room.

[Saurav]

Originally Posted by twest820

Quite the bloom on the PDR. The sonograms are derived from the 15 degree data?

I think they're referenced to 0, so the 'bloom' you see at 7kHz is really the on-axis dip affecting the data normalization. I think. Keyser, if you have the data saved, could you plot the same charts, un-normalized? I find it useful to look at it both ways.

[twest820]

Regardless of how the data's displayed the finding's still the same; more bloom on the PDRW than the W.

Originally Posted by Rudolf

The most elegant (I would say "best") solution would be to have the widest source in the narrowest baffle.

For nude and near-nude drivers such as in this study, yes. Horns represent the opposite end of the design continuum and appear to be viable too---it's mainly midway between the extremes where performance gets dicey. Certainly easier to mill down the flange on the Neo3 than to build a horn (or waveguide), though the desirability of doing so depends on how low the Neo3 needs to cross.

[Saurav]

Originally Posted by twest820

Regardless of how the data's displayed the finding's still the same; more bloom on the PDRW than the W.

I would disagree You could have two drivers with identical off-axis curves, and one of them has a deeper notch on-axis. The normalized curves would show this same difference. Would you describe that as one driver blooming more than the other? I wouldn't. I guess technically you would be correct, but that's not what I think about when I think 'bloom'.