*** Waveguides ***

[ralphs99]

I've added a new page to my web site about Waveguides!

http://www.aeronet.com.au/waveguide.htm

It documents some of the studies I've been doing recently into the characteristics of waveguides and their use with direct radiators.

I hope the information is useful to those people doing research into waveguide operation! :T

Cheers, Ralph

[TacoD]

Wow great website, I bougth the software from Earl Geddes but I have not tried a waveguide yet.

[Patrick Bateman]

Based on your graphs, it looks like the compression driver on a waveguide "drops" less off-axis than a conventional tweeter. By "drop", I mean that the compression driver has more energy off-axis than a dome, ON THE SAME WAVEGUIDE. Any idea why?

[Tommythecat]

I think his ideas of the difference between the dome tweeter and the ring radiator - and a phase plug - deals with the issue. Especially if using a ring radiator with a phase plug, like the Vifa XT series tweeters, would seem to achieve a better power response (and also help out the XT's rampant low-end distortion ie below 2k).

So i think the compression driver is immediatly better suited to a horn/waveguide since it has better phase control when used in a WG(in comparison-straight out of the box). But for "normal" listening levels, the eq'ed direct radiators can be just as good.

[EdL]

A compression driver provides something closer to an "ideal" plane wavefront. A dome, with the wave being generated by a section of a sphere, produces a wave that is different from the start. This difference is one of, if not the only reason why metal dome tweeters have phase shields in the wavepath. Cloth domes are more compliant. Their flexibility is one reason they are less likely to be found with phase shields.
Ring radiators flex between their rings. It is a design feature. Though not technically correct to call a filter, you could consider it to be a mechanical filter. The compliance between the rings means that some parts of the radiating surface emit frequencies that other parts do not. Since the radiating surface more closely resembles a plane, the ring radiator does provide a better wavefront than a dome to couple with a waveguide.

Now, if low freqency extension is one of the aims, a metal dome becomes a necesity. The higher excursions required by lower frequencies expose the diaphragm to compression by the resitance of the air being driven. A fabric dome is not up to the task. It will distort on its way to those lower frequencies. Only a metal dome is stiff enough not to distort...or use a compression driver.

[ralphs99]

Thanks for the comments, everyone!

-Patrick,

I think your question has already been well answered by Tommy & Ed.

A compression driver appears to load the waveguide in a better fashion. The advantage is reduced for large included angles however. So direct radiator loading, particularly with a ring radiator, works almost as well on practical waveguides.

Overall I favour direct radiator loading. While the coupling into the waveguide throat needs extra attention, direct radiators have less colouration and I think can sound superior to a compression driver/waveguide combination at normal listening levels.

Cheers, Ralph

[Patrick Bateman]

Thanks for the comments, everyone!

-Patrick,

I think your question has already been well answered by Tommy & Ed.

A compression driver appears to load the waveguide in a better fashion. The advantage is reduced for large included angles however. So direct radiator loading, particularly with a ring radiator, works almost as well on practical waveguides.

Overall I favour direct radiator loading. While the coupling into the waveguide throat needs extra attention, direct radiators have less colouration and I think can sound superior to a compression driver/waveguide combination at normal listening levels.

Cheers, Ralph

Sorry to hear that. I have listened to everything from the new Vifa XTs, to B&C DE25s, to JBL 2470 compression drivers.

IMHO, compression drivers on a good waveguide can sound as sweet as a direct radiator, with VASTLY improved dynamics and low-level detail.

Of course they're expensive, very expensive, but they're worth every penny. And they're fairly affordable on e-bay.

[ralphs99]

Hi Patrick,

I think compression driver based systems can sound very good indeed!

I've spent the last 25 years listening to mostly 2 way systems with compression driver/horn HF. I've lived on a day-to-day basis with JBL 4435's, Tannoy Gold's, EV Sentry III's & Altec Model 15's amongst others, along with several custom systems using JBL2420's, 2426's and most recently 2407's. I like many of the attributes of compression driver/horn setups.

But I have to say the 2470 is one of my least favourite compression drivers. The phenolic diaphram can sound really nasty with several major break-up modes inside the HF passband. If a compression driver is restricted to operating between 800Hz and 2kHz, then it seems to me that there's little point to the exercise.

My current research is primarily directed to finding something smoother than the best available compression drivers while still maintaining the positive qualities of compression driver/horns combinations.

Cheers, Ralph

[EdL]

With recent info provided in the H1212 thread, I'm giving thought to keeping the phase shield in the throat of a waveguide. I do want to use a smaller throat than the hexagrid allows. This will require removal of the grid, cutting it to a smaller diameter, providing a slight groove in the wall of the guide and re-insertion of the gird....without changing the distance of the shield from the dome....applied geometry at it's best...

Any comments?

[Patrick Bateman]

Hi Patrick,

I think compression driver based systems can sound very good indeed!

But I have to say the 2470 is one of my least favourite compression drivers. The phenolic diaphram can sound really nasty with several major break-up modes inside the HF passband.

That's what I've heard too. For $180, replacing the stock diaphragms with aluminum Radian diaphragms is a easy upgrade. Which is just what I did!

[JoshK]

With recent info provided in the H1212 thread, I'm giving thought to keeping the phase shield in the throat of a waveguide. I do want to use a smaller throat than the hexagrid allows. This will require removal of the grid, cutting it to a smaller diameter, providing a slight groove in the wall of the guide and re-insertion of the gird....without changing the distance of the shield from the dome....applied geometry at it's best...

Any comments?

Hey Ed! Haven't spoken to you in a bit. Anyway, I like your idea, pretty clever. :T I had thought about using a toothpick or something equivalent (long pin, thin rod, etc) attached to the waveguide's side (or drill a small hole to insert toothpick through) and attach the shield to the toothpick.

[noah katz]

"My current research is primarily directed to finding something smoother than the best available compression drivers while still maintaining the positive qualities of compression driver/horns combinations."

Have you tried B&C DE250 or BMS 4550/4552ND?

[ralphs99]

With recent info provided in the H1212 thread, I'm giving thought to keeping the phase shield in the throat of a waveguide. I do want to use a smaller throat than the hexagrid allows. This will require removal of the grid, cutting it to a smaller diameter, providing a slight groove in the wall of the guide and re-insertion of the gird....without changing the distance of the shield from the dome....applied geometry at it's best...

Any comments?

Hi Ed,

Yes, keeping the phase shield in place will be the way to go. It will help eliminate the on axis dip which you will otherwise experience. Good luck with the modification. In my tests, where I added a phase shield, I used a piece of wire with a plastic washer glued to middle of it. Crude but effective!

Cheers, Ralph

[ralphs99]

Hi Patrick,

I think you made a smart move. I've never knowingly listened to a radium diaphragm, but I believe they are very good. Definitely way better than the phenolic!

Cheers, Ralph

[ralphs99]

Hi Noah,

It's an interesting question. I have it on pretty good authority that BMS supply parts of the 4552ND to JBL who then sell it as a 2407.

They certainly look very similar inside & out!

If the rumours are true, then yes, I have tried the BMS 4552ND in the guise of the JBL 2407.
I like it as a tweeter, ie used above 3kHz. But I think the old fashioned 1" exit compression drivers like the 2420 sound smoother and more natural even though they miss out on the high frequency extension.

Cheers, Ralph

[EdL]

Thanks, Josh & Ralph

I think I see a similarity between the phase shield Seas incorporates and the foam plug Geddes advocates. Both affect (improve) on-axis response.
The response from a waveguide provides a more significant null. Geddes' foam plug is said to lessen the higher order modes of refraction occurring within the guide.

Am I trying to draw a connection between different things?

[ralphs99]

Hi Ed,

I think you're on the right track. I also believe the effects to be related.

Although Geddes uses compression drivers to load his waveguides, I think there are still some phase issues in the throat region. These aren't very noticeable with a rapid expansion and large included angle, but are probably exacerbated as the directivity of the wave guide increases, to the point where some measures to counter the phase issues become necessary. ie the foam inserts he uses.

I wonder if the foam inserts would be even more effective is the density could be varied, being most dense at the centre and progressively less dense towards the periphery.

In any case, for a regular dome tweeter, the idea of a foam insert isn't effective enough. I've tried it with various acoustic foams and felts, and the phase problems were still in evidence. Without a driver that uses a normal phase plug, a phase shield or similar acoustic device is needed for waveguide loading. (even without a waveguide in some cases)

Thus I think (and I've observed) the best compromise is to load a waveguide (with a reasonably large included angle) with a ring radiator.


Cheers, Ralph

[noah katz]

"I think I see a similarity between the phase shield Seas incorporates and the foam plug Geddes advocates...Am I trying to draw a connection between different things?"

I think so. From Geddes' description of what the foam does, it's simply absorptive, which absorbs more of the HOM because of their longer path length.

OTOH phase plugs alter the path length of the desired signal.

[EdL]

During a drive this morning I thought about the mounting of the phase shield within a guide. Previously I thought of resizing and using the existing hexagrid. Here is an alternate: Salvage the phase shield. Make a 3-wire (tripod) armature to support a small metal (brass?) tube. Insert a shaped wooden plug (golf tee?) into the tube to position the phase shield within the throat. This allows for adjustment of location.

Comments?