CHANGSTAR: Audiophile Headphone Reviews and Early 90s Style BBS
Lobby => Speakers => Topic started by: Marvey on August 13, 2013, 03:04:31 AM
-
I've always wondered if the phase plug mod for various drivers (especially the Fostex) were bullshit or not, so I decided to do a rudimentary phase plug mod take some before and after measurements.
SEAS FAQ
What’s the difference between a dust cap and a phase plug?
Why use a phase plug?
• Reduced moving mass
• No gluing or other interaction with the cone
• Better cooling of the voice coil
• No cavity resonances
• More consistent high frequency response
Why use a dust cap?
• Avoid air leakage and “chuffing” noise
• Larger cone area
• Conceal a small voice coil with a big dust cap
• Cosmetical reasons
In the case of the FE166 driver, there wouldn't be much loss in cone area. The dust cap already has a hole in it for venting. See pic below. If we cut the dust cap off, we probably lose about one square inch of total area.
-
Here are two baseline measurements of the stock fe166en driver mounted on the BK-16. The microphone is aligned with the driver and 24" away. Graphs are 1/3 octave smoothed. First graph is directly on the driver 0 degrees. Second graph, the microphone is 30 degrees off.
We can see that the speaker has a slightly rising response when dead on. However the driver exhibits beaming behavior - we see roll-off starting at 3kHz when offset slightly from the driver.
-
So the idea behind the phase plug is that high frequencies would bounce off the phase plug and therefore we would get a more uniform polar response - less beaming. Here is a picture of the hack job I did - just for testing purposes.
-
Here is a 0 degree front on measurement of the speaker with the phase plug mod.
What I find interesting is a decrease in 2nd and 4th distortion in the bass. This is something I noticed after I took the measurements. I did take the average of three measurements however.
I strongly suspect the dust cap is the culprit for the higher distortion found in the un-modded driver. Even though the dust cap had a vent in front, I don't think the vent was enough to prevent an air cushion when low bass was playing. I figure this air cushion presented a varying load on the cone, hence the higher distortion. I noted during frequency sweeps that the driver displaced like crazy when low bass tones were being played. Still bass measurements are funny. I will need to take a few more to confirm.
I also noticed the area around 2k is a little smoother.
-
When I was big into car audio I always preferred phase plugs. Most of the speakers I end up buying myself usually have them.
Is that a DIY phase plug sawed from one of those rods at Home Depot? Not anything I've ever seen. Used to bullets.
-
Interesting, what SPL are these sweeps blaring at in absolute terms?
-
Quick answers:
- Home Depot wooden dowel. Too lazy to carve into bullet. Bullets aren't totally necessary. I've seen cylindrical phase plugs. Would be interesting to measure different types.
- 0 db = 120 SPL. So we are talking around 90db
-
So here is the 30 degree measurement of the phase plug mod.
The second graph is a comparison between the stock dust cap and the phase plug modded driver at 30 degrees. Conclusion is that this mod is not bullshit. At 30 degress of, the area at 2kHz is more even. We also see significantly less roll-off ~5db in the treble.
(http://www.changstar.com/index.php?action=dlattach;topic=925.0;attach=3191;image)
-
Any subjective differences in the original sweet spot or it just better overall?
-
It actually makes a difference. The primary reason for the experiment was for HT / movies, where I would usually sit off axis and the HF roll-off was notable. Last night I ended up resting on the couch listening to music, and I usually never do that. The new room I put the speakers is much more sensitive to off-axis response because the couch is much closer to the speakers.
The next step is to design a phase plug to tone down the rising response on-axis and maybe even further flatten out the response off-axis.
-
The next step is to design a phase plug to tone down the rising response on-axis and maybe even further flatten out the response off-axis.
I don't suppose it would be as simple as having a heavier phase plug. Mass-loading to create a HF roll-off?
This also makes me wonder about the whole modding thing that planet_10 does to the Fostex drivers.
http://www.planet10-hifi.com/fostex.html
-
Once the dust cap is removed, the phase plug isn't part of the cone anymore. The phase plug actually attaches to the pole piece. It's not part of the moving mass.
As for the EnABL stuff, it probably works. It's no different from mass loading specific parts of the cone with itty-bitty little bits of bubblegum.
-
My main effort would be trying to bring down the response at 7k. If I could simulate the potential differences in response, I might be willing to try a bit of that foolery with my own drivers.
-
You'd need a better speaker for that (this one is already pretty good though) or equalization. No real way to not drop detail extraction while cutting those frequencies otherwise.
If you are willing to sacrifice detailing, some front thin foam filter would do.
Is this for a 3-way design?