Phase Plug?

A dustcap adds mass to the center of the membrane, a membrane that on certain speakers (PA midranges, full range speakers) is very light. Above a certain frequency (e.g. the 02 or 03 vibrational mode of the circular membrane) the outer membrane cone can get decoupled from the center cone. This largely depends on stiffness (compliance) of the membrane material at this frequency and the mass at both sides of the motion null. At higher frequencies the speaker can behave as if it has a smaller membrane (smaller area but also lower mass and different fs) surrounded by a short conical horn. I have a 10 in PA midrange where the membrane is made this way on purpose.

In the impedance measurement of the 6ND410 posted by John Janowitz, a resonance at 750 Hz disappears when the dustcap is removed. So the above might be happening with this speaker. Such a speaker can still be very usefull for PA where a midrange horn is not an option. For monitoring and hifi I would not suggest this approach, but Speaker Dave is correct that the dustcap can be a significant variable for the woofer frequency response and can reduce the woofer's effective area.

I started reading this post because I was looking for information on phase plugs at the entrance of a horn (sometimes called fins in rectangular horns). I think the high frequency dispersion of a horn can be improved with a phase plug that starts at the mouth of the compression driver. Any idea in which post there might be some information?
 
A dustcap adds mass to the center of the membrane, a membrane that on certain speakers (PA midranges, full range speakers) is very light.
Related to this is something I don't think anyone brought up so far in this thread - a dust cap can actually be attached in a number of different ways, giving quite different results in regards to resonances and high frequency performance.

On large woofers the dust cap is often glued to the front face of the cone and is a much larger diameter than the voice coil, thus is attached to the cone some distance beyond the attachment point of the voice coil. Not a good approach IMO, as you're setting up multiple resonances between the dome and the sections of cone within and beyond the dome.

I'm not really sure why a lot of woofers have gone this way - perhaps just to look more impressive and appear to have larger voice coils than they actually do. Older woofers generally seemed to have dust caps that were the same size as the voice coil diameter, albeit still glued to the cone.

Yet another approach used in some full range drivers like the Coral Flat 8 is for the (aluminium) dust cap to be directly mechanically crimped on the end of the protruding voice coil former, with the cone(s) separately glued to the former a few mm further down.

This would seem to be the best approach for extended high frequency response as there is no direct bond between the dust cap and the cone, with the glue joint between cone and former providing enough isolation at treble frequencies for the directly driven dust cap to operate efficiently. (In fact I've found gluing the whizzer cone to the edge of the dust cap instead of leaving them separately connected to the voice coil former seriously kills the treble performance)

Above a certain frequency (e.g. the 02 or 03 vibrational mode of the circular membrane) the outer membrane cone can get decoupled from the center cone. This largely depends on stiffness (compliance) of the membrane material at this frequency and the mass at both sides of the motion null. At higher frequencies the speaker can behave as if it has a smaller membrane (smaller area but also lower mass and different fs) surrounded by a short conical horn. I have a 10 in PA midrange where the membrane is made this way on purpose.
Indeed. This would seem to be one reason why a whizzer cone can improve dispersion well below the frequencies at which it's operating independently - without it at high frequencies only the middle of the normal cone is active, and it's horn loaded by the depth of the rest of the cone.

Put a whizzer cone in and now you have high frequency radiation at the centre which originates significantly futher forward, lessening the horn loading effects from the inactive part of the main cone. As well as that it acts as a phase plug placed in front of the radiation of the main cone.

I did some measurements on the Flat 8A years ago de-whizzered vs normal and found to my surprise that the presence of the whizzer cone actually reduces the on axis response from 2.5Khz to 7Khz by 3dB, whilst simultaneously increasing the off axis response out to 30 degrees by a few dB and having a significant effect even as low as 2Khz. Without the whizzer cone the driver is beaming already at 2Khz, with it in place you get at least an extra octave without beaming.

Then of course the whizzer cone itself acts as a horn for the treble radiation from the dust cap, significantly increasing it. Lots going on in a driver of that type that's not immediately obvious :)
 
I started reading this post because I wondrered by some expensive and well known drivers like the Audio Technology 18H52, ScanSpeak 18WU/8741, SB Acoustics MW16P do NOT have phase plugs while others like the SEAS Excel, Eton Symphony, Focal etc.. use phase plugs.

To be honest after reading 62 posts twice I am still not sure about the merits or de-merits of phase plugs.

There are some economical (about $25 each) 6" woofers using a black anodised aluminium cone that I was looking to modify just lke Dave (Planet 10) has modified Fostex drivers with Phase plugs and EnABLing (http://www.diyaudio.com/forums/multi-way/100399-enabl-processes.html).

I figured an aluminum cone would be easier to EnABL as it can be screen-etched.
http://www.goodsoundclub.com/pdf/EnABL White Paper.pdf
 
Misnomer

I started reading this post because I wondrered by some expensive and well known drivers like the Audio Technology 18H52, ScanSpeak 18WU/8741, SB Acoustics MW16P do NOT have phase plugs while others like the SEAS Excel, Eton Symphony, Focal etc.. use phase plugs.

To be honest after reading 62 posts twice I am still not sure about the merits or de-merits of phase plugs.

There are some economical (about $25 each) 6" woofers using a black anodised aluminium cone that I was looking to modify just lke Dave (Planet 10) has modified Fostex drivers with Phase plugs and EnABLing (http://www.diyaudio.com/forums/multi-way/100399-enabl-processes.html).

I figured an aluminum cone would be easier to EnABL as it can be screen-etched.
http://www.goodsoundclub.com/pdf/EnABL White Paper.pdf


The only acoustical place a phase plug is to be found is inside a compression driver. In this setting, its merits are well known. Outside this setting, characterization of obstacles placed in front of an open frame drivers as “phase plugs” is nothing more than marketing hyperbole that will always be subject to questions of merit.

Regards,
WHG
 
The only acoustical place a phase plug is to be found is inside a compression driver. In this setting, its merits are well known. Outside this setting, characterization of obstacles placed in front of an open frame drivers as “phase plugs” is nothing more than marketing hyperbole that will always be subject to questions of merit.

Regards,
WHG

So SEAS, Eton and Focal shown below use the phase plug only becuase it looks nice or caters to some marketing myth?

http://tokospeaker.com/sites/defaul...eaker_woofer_10 inch_26cm_E0026_W26FX001b.jpg

Eton 7-200/A8 Symphony, 7" Mid/Bass with Aluminum Former | TokoSpeaker, Toko Speaker, Car Audio, Audio Mobil, Speaker, Tweeter, Midrange, Subwoofer, Capacitor, Resistor, Inductor, Audio

Speaker City sells speakers, drivers, audiophile loud: Utopia Cone 13cm w/Butyl Surround/Shielded
 
I continue to study 7" - 10" wide bandwidth midbass speakers to find drivers that can cover the 80Hz - 1500Hz vocal range with good polar response (0, 30, 60 degree SPL plots) and low intermodulation(doppler) distortion. Several of the better drivers for this task include a phase plug plus a curvelinear cone profile.

A few wave simulations from one phase plug web document are attached.


"phase plug" speaker - Google Search
 

Attachments

  • Ideal_speaker.jpg
    Ideal_speaker.jpg
    53.5 KB · Views: 540
  • Real_noPlug.jpg
    Real_noPlug.jpg
    55.9 KB · Views: 544
  • Real_Plug.jpg
    Real_Plug.jpg
    57.7 KB · Views: 538
If these simulations are real then the phase plug actually makes things worse (in my view). To bad, because the SEAS Excel Magnesium drive with it's copper coloured phase plug is a very beautiful driver (in looks).

I think this is the idea behind the simulation.

The wave simulation of the "real" speaker without the phase plug shows a dark shadow on axis of interference representing lobbing. The wave simulation of the "real" speaker with the phase plug shows less interference lobbing.

The multiple point sources on a speaker cone can cause lobbing effects. The power is radiated in a very small angle. With a phase plug one side of the cone is "separated” from the other side. There is no lobbing or wave cancellation at high frequencies. The phase plug can extend the speaker
response.
 
Issue Obstacle

When the signal frequency increases and wavelength becomes smaller than obstacle dimensions, then such an obstacle is no longer acoustically transparent; however, it is at such frequencies that the vibrating piston assumption of the model used is no longer valid. Here, as frequencies increases, the diaphragm will transition through numerous breakup modes of ever increasing complexity. These will have a much stronger influence on the radiation pattern of the driver than obstacle presence. To implement a broad radiation pattern at high frequencies requires use of a driver with an acoustically small diaphragm.

Regards,

WHG
 
With a phase plug one side of the cone is "separated” from the other side. There is no lobbing or wave cancellation at high frequencies. The phase plug can extend the speaker
response.

How so? It looks worse near the phase plug, but in "far-field" it seems quite a bit more uniform.

Hence a phase plug really only makes sense in the midrange. Above what frequency does it make sense? 300hz?

When the signal frequency increases and wavelength becomes smaller than obstacle dimensions, then such an obstacle is no longer acoustically transparent; however, it is at such frequencies that the vibrating piston assumption of the model used is no longer valid. Here, as frequencies increases, the diaphragm will transition through numerous breakup modes of ever increasing complexity. These will have a much stronger influence on the radiation pattern of the driver than obstacle presence. To implement a broad radiation pattern at high frequencies requires use of a driver with an acoustically small diaphragm.

So if one has a 3 way with a woofer at say 8-10" up to 300hz and a tweeter at say 3khz with a small 3.5-4" midrange covering "the middle decade" 300-3k does a phase plug on the midrange help?
 
Hence a phase plug really only makes sense in the midrange. Above what frequency does it make sense? 300hz?
I took some measurements on an 8" full range driver with 30mm voice coil recently, dustcap vs no dustcap vs various size and shape phase plugs, none of them had any effect below about 1.4Khz, so on a smaller driver I would expect the minimum frequency where it made a difference would be even higher.

You definitely won't be seeing any changes at 300Hz, other than a slight overall change in sensitivity due to change in cone mass and cone area. (Exchanging a dust cap for a stationary phase plug reduces cone mass and cone Sd area)

Generally phase plugs have two effects vs having a dust cap:

1) They can eliminate the cavity resonance behind the dust cap, which in my particular driver was about 1.6Khz. (If the cavity is eliminated, so is the cavity resonance)

2) They can eliminate the dust cap resonance. (what's no longer there can't resonate...) On this particular driver the dust cap resonance was around 9-12Khz.

On the other hand you now have to worry about diffraction from the phase plug - if the shape is wrong it will introduce large variations in response due to diffraction. On this driver I tried various shapes including deliberately squared off ends and the result is large fluctuations in response of many dB from about 4-8Khz.

The smoothest response through this region was a traditional "bullet" shaped phase plug, where the end smoothly curves to a point. However the smoothest response I could obtain with a phase plug from 4-8Khz was actually no smoother than the original dust cap.

Contrary to popular belief I didn't find any improvement in high frequency dispersion with the phase plug, in fact dispersion in the treble was worse, due to elimination of the radiating (aluminium) dust cap. (with a paper dust cap there probably would have been little/no change to treble dispersion)

Phase plugs in midrange/full range drivers are one possible design approach but they don't help with dispersion, at most they can help eliminate cavity and dust cap resonances which can also be dealt with in other ways, (dust cap material damping, vented and damped pole pieces etc) and must be carefully designed so as not to introduce response irregularities due to diffraction.

Although it seems to be a popular tweak in full range circles, almost taken as a given to be "better", retro-fitting a phase plug in a driver designed with a dust cap is by no means guaranteed to give a better result, and can very easily give a much worse overall result. At the very least a lot of careful measurement and iteration of the shape of the phase plug would be necessary to achieve an acceptably flat response on an arbitrary driver, and chances are that high end treble output will drop significantly.

(Overall output will drop due to a reduction in Sd, but treble response will drop proportionately more due to the dust cap being a larger percentage of the effective cone area at treble frequencies)

If the driver comes with a phase plug you can reasonably assume the manufacturer designed the phase plug correctly to match the rest of the driver. (Such as its cone profile)

One other comment about phase plugs - sometimes you see drivers where the "phase plug" is actually attached to the cone like a bullet shaped dust cap, rather than being stationary and attached to the pole piece.

This throws away (IMHO) the two possible advantages of a phase plug - because it's attached to the cone it will be radiating, and basically is a bullet shaped dust cap which will have a dust cap resonance. Likewise because its attached to the cone there must be a cavity behind it to allow it to move, so there will be a cavity resonance, unless the pole piece is vented and damped. (Albeit at a different frequency to a normal cavity resonance due to the larger cavity)

It still won't be doing anything to improve the dispersion at high frequencies, so I'm not sure that I see the point of a "phase plug" that is attached to the cone rather than the pole piece. I would venture to say it may be more about looks than anything else... ;) That's not to say that such a driver is necessarily bad, just that the "phase plug" is not doing anything special.
 
Last edited:
I'm glad this thread got bumped up as it has been a very interesting read. I still need to go back through it again and see if I can make heads or tails of it, but I must say, it's been interesting non-the-less.

The only argument I can add to the conversation (albeit less scientific), would be that, a large number of well respected wide-band drivers have them. However; A quick glance through the full range drivers at PE, reveals many designs with phase plugs and many without. Many of them with plugs appear to have nice SPL charts, some, not so much. At the same time, there are quite a few drivers without phase plugs in the same category, some look good, some not so good as well.

My gut tells me, that like any other driver characteristic, nearly all things come and go with a handful of trade-offs that need to be properly balanced to get a respectable performance. Like under-hung vs over-hung coils, or poly cones vs aluminum cones, they all work, just differently, and they can all work well, if implemented and utilized properly. The phase plug, I have to imagine, falls into a similar category.

Eric
 
Last edited:
how?

The two main issues that it addresses (cavity resonance around 1-2Khz and dust cap resonance around 10Khz) can both be addressed in other ways IMHO, without the drawbacks of a "phase plug".

What are the other ways to solve the issues?
I'm especially interested how the cavity resonance is solved ... without cutting something, and what do you do if you have a wizzer. The dust cap resonance I presume is solved like any other resonance, with felt pads.
 
Last edited:

AndrewJ

Member
2001-09-05 6:37 pm
With regard to phase plugs on bass/mid drivers, I was always very sceptical regarding claimed improvements in off axis performance. So I did the experiment!
I measured a typical polypropylene cone/dustcap driver, both on axis and at 15 deg off axis increments out to 60 deg.
I then plotted the change in response at each angle compared to on axis. This is then a measure of the directivity response of the driver.
Next I removed the dustcap and re-measured the on axis response. Of course, this was messed up because without the dustcap, we have a cavity resonance formed by the tubular well formed by the voice coid former and top of the pole piece.
Next I fashioned a fixed "phase plug" to fill this cavity, the top end of the plug sitting level with the start of the cone.
I then remeasured all the on and off axis responses. The on axis response was now much closer to that of the driver with the dustcap, reduced in level almost 1dB due to the lower effective area having removed the dustcap, but the basic response was quite simliar to the one with dustcap.
However, the change between off axis and on axis was identical, within closer than 0.5dB, to the dustcap version out to beyond 7 or 8kHz.
Extending the phase plug even further out made no noticable changes to the directivity response, and only marginal changes to the on-axis response.
The conclusion is that a so-called phase plug has no real effect on driver directivity, especially so for a midrange driver that will be crossed over to a tweeter.
 

AndrewJ

Member
2001-09-05 6:37 pm
Interesting link about the definitive Technology dual surround woofer.
We used this technique extensively on all of the KEF coupled cavity woofers, starting with the 104/2. We need it to isolate the front and rear of the cone because the coupled cavity loading technique was very sensitive to these leaks.
The center roll surround was needed because we used a rod bolted between the drivers for force cancelling of the woofer vibrations, so we could not use a conventional dustcap.
This was back in the 1980's, so I would think that the technique is considered prior art and is not therefore patentable, unless they have some other tricks up their sleeve and a good patent attorney :)
 
No!

Hence a phase plug really only makes sense in the midrange. Above what frequency does it make sense? 300hz?



So if one has a 3 way with a woofer at say 8-10" up to 300hz and a tweeter at say 3khz with a small 3.5-4" midrange covering "the middle decade" 300-3k does a phase plug on the midrange help?

Wave length of a 3 kHz signal (about 5”) is larger than the diameter of the specified driver and even much larger than a center obstruction (co) considered here; so, it (co) remains acoustically transparent and irrelevant.
Regards,
WHG