Phase Shield Removal Redux

[Armand]

I hate to bring this topic up again as it's been discussed here and elsewhere, I decided to ask B&W how they get away with not using shields on their hard dome tweeters. Here is their response:

We haven't used either of these devices for some years now. Pre-Nautilus 800 Series, we used to put a ring in front of the dome and this had the effect of raising the output at very high frequencies (> 10kHz). Unfortunately, this was not a well-dispersed output increase. The on-axis response was lifted at the expense of that at wider angles (in other words, the dispersion was worse).

Removing these devices results in a better, more open sound. A similar improvement was found when we introduced a new tweeter dome support surround on the latest 800 Series Diamond. In both cases you are altering the relative phase of the radiation from different parts of the dome/surround combination.

Of course, you have to deal with the fact that the very high-frequency output is lower than one would want and the only practical solution has been to increase the efficiency of the tweeter over its whole range and attenuate more the lower frequencies through the crossover.


Best Regards,
Eric John McBride
Custom Theater Specialist
Bowers and Wilkins Service Supervisor

Now correct me if I am wrong, but I thought the shield improved dispersion and off-axis response? Also, the shield removal is "altering the relative phase of the radiation from different parts of the dome/surround combination", but he doesn't seem to address this.

Am I missing something here? I'm interested to hear what you experienced DYI'ers think. Thanks.

[dlr]

Now correct me if I am wrong, but I thought the shield improved dispersion and off-axis response? Also, the shield removal is "altering the relative phase of the radiation from different parts of the dome/surround combination", but he doesn't seem to address this.

Am I missing something here? I'm interested to hear what you experienced DYI'ers think. Thanks.

The phase shield is primarily a benefit to the on-axis, though there's something of a transition region where there is improvement to neutral to detriment as you move from on-axis to off-axis. The improvement on-axis is due to "blocking" the output from the area at the tip. On-axis this means that there's less output with time deltas. The latter is what he meant by "altering the relative phase of the radiation from different parts of the dome/surround combination" as well. It's all about relative phase.

For practical purposes, very hard domes emit the signal at nearly the same time due to speed of sound being much higher in the dome than it is in air. What this means is that the sound emitted from the tip of the dome will arrive before the rest. The sound from near the former attachment point will be delayed. The distance from surround to tip starts to become significant as the wavelength decreases with frequency rise. This causes interference, hence the droop in output. Block the tip and there's less interference.

The detriment is in the off-axis. Far enough off-axis and there is occlusion of much of the dome output (blocking) due to geometry. This is a benefit off-axis as there is much less output with significant time delays, hence less droop. Add a phase shield and you get reflected signals that cause complex interference patterns in the off-axis area.

I have a writeup using the Parts Express RS52 dome midrange that shows experimental results of various phase shield tests.

RS52 tests and measurements

Dave

[Armand]

Thanks Dave and I'm well aware of your website. In fact, I tried some different materials as the phase shield on a pair of SEAS H1212's with soft felt sounding the best based on your findings. Unfortunately, I don't have test equipment so I'm confined to listening tests only.

I was wondering if removing the shield might be of benefit with my modified Spica TC50's which, behind a very thick felt blanket, offsets the tweeter by 34 degrees such that in the listening position you are listening to the tweeter off-axis. Additionally, the reverberant lobe (on-axis) of the speaker points to the ceiling. If my attempt is to design a narrow dispersion system due to the fact that my listening environment is small (21x12x10), then my thinking is that a reduction of reflections off the ceiling might be a good thing.

What I hear without the shield is an electrostatic-like increase in detail, focus, depth and transparency but with an overall reduction in the room power response. The soundstage is less vague, but there is an unfortunate loss of image body, the FR sounds somewhat tilted up and the sweetspot narrows. Just like some B&W's I've heard! ;}

Considering the frustrating tradoffs, I might run it without the shields and adjust my acoustics in an attempt to reduce the problems noted. It's tough going back to using the shield after hearing all that extra detail and depth.

Thanks for the response.

[dlr]

I was wondering if removing the shield might be of benefit with my modified Spica TC50's which, behind a very thick felt blanket, offsets the tweeter by 34 degrees such that in the listening position you are listening to the tweeter off-axis. Additionally, the reverberant lobe (on-axis) of the speaker points to the ceiling. If my attempt is to design a narrow dispersion system due to the fact that my listening environment is small (21x12x10), then my thinking is that a reduction of reflections off the ceiling might be a good thing.

If the ceiling is low, then I suspect it might, as it will introduce a slight dip on-axis, but the change may not be much for the perceived power response from reflected energy given the frequencies affected.

What I hear without the shield is an electrostatic-like increase in detail, focus, depth and transparency but with an overall reduction in the room power response.

That would, I believe, be due to the lift in the upper frequencies of the direct response on your listening axis, the off-axis of the tweeter.

The soundstage is less vague, but there is an unfortunate loss of image body, the FR sounds somewhat tilted up and the sweetspot narrows. Just like some B&W's I've heard! ;}

The direct sound increase in the off-axis would be in line with this change in perception, since it's a wide bandwidth effect.

Tilt up the response for more detail and something else has to give way. That's one of the good things about DIY, you can tailor it, sometimes in very small ways that have a surprisingly large perceived effect. You could be adding as much as a full db increase in the upper tweeter range on your listening axis. Though the on-axis will droop, the off-axis increase possibly increases the tweeter power response in its affected range.

Dave

[Paul W]

Used to live in Medford NJ...so "hello" from an ex neighbor.

You seem serious about this so I suggest you invest in inexpensive measurement gear so you can be more certain of the impact and compare results of subtle changes.

[Brian Kingsbury]

It might be fun to try putting some pieces of foam a la' the Raal 140-15D to reduce directivity and increase dispersion and off-axis response. They sure are fugly but I'm guessing they do what they're supposed to or they wouldn't be on those REALLY expensive and highly regarded tweeters.

[Armand]

Great analysis Dave. Thanks and you're probably right.

The differences with shield on and off is quite frustrating. I've come to the conclusion that the tonal balance without the shield is too peaky and I don't have the means to even attempt to compensate in the crossover so I decided to change the voice coils yesterday and get back to square one. I think the originals were somewhat compromised from experimenting with the shields on and off. Being curious can be expensive sometimes.

So it's back to a slightly vague soundstage which is a shame because the Spica's are imaging champs. Without the shield the focus took on an electrostatic like quality but like electrostatics, you were forced to put your head in a vise. But the excellent tonal balance is back and that's more important to me. Thanks again for the education, Dave.

[Armand]

Used to live in Medford NJ...so "hello" from an ex neighbor.

You seem serious about this so I suggest you invest in inexpensive measurement gear so you can be more certain of the impact and compare results of subtle changes.

Hi Paul. I can see from your site that attempting something similar in a 150 year old home in Medford might be quite difficult! I hope you listen with earplugs. ;}

Honestly, I wish I could afford the equipment as I'd jump on it in an eyeblink. Geez, I don't even have a basement to work in because of the high water table here in Marlton.

Are you still on the East coast?

[Armand]

It might be fun to try putting some pieces of foam a la' the Raal 140-15D to reduce directivity and increase dispersion and off-axis response. They sure are fugly but I'm guessing they do what they're supposed to or they wouldn't be on those REALLY expensive and highly regarded tweeters.

I find it hard to believe that you can control dispersion using foam unless the intent is to limit the vertical dispersion. For that kind of money, I expect a ribbon wrapped around Sophia Vergara........

[Paul W]

No longer on the east coast. I've retired to the foothills in northern Georgia, just a few miles from the southern end of the Appalachian trail. Family here, the weather is milder, large lake nearby...and property taxes are less than 1/5 what I was paying in NJ .

[Armand]

Wow, talk about culture shock.

Our best friends just got back from visiting her relatives in GA. To pay for room and board, he was assigned the job of taking care of the hogs, LOL. I had to ask him if he had to buy overalls a straw hat and took up the washboard. He did always wanted to play an instrument.

Actually I've met her family at weddings here and they were terrific people. I'm somewhat jealous of their rustic lifestyle as they seemed so laid back. Sounds like a healthy way to live to me.

[Armand]

Dave: I noticed that SEAS does not provide a pressure release (hole), drilled into the plastic phase shield. I see that you performed this on a midrange driver on your website to avoid any Helholtz resonance. I also noticed a hole drilled in the shield on a pair of Energy speakers in my AV system.

Do you know why SEAS does not do this? Is it advisable to poke a hole in the shield?

[dlr]

Dave: I noticed that SEAS does not provide a pressure release (hole), drilled into the plastic phase shield. I see that you performed this on a midrange driver on your website to avoid any Helholtz resonance. I also noticed a hole drilled in the shield on a pair of Energy speakers in my AV system.

Do you know why SEAS does not do this? Is it advisable to poke a hole in the shield?

It doesn't always help. I try everything when testing, but in some cases it's detrimental, primarily for the on-axis response. It's too variable for me to suggest it without testing.

Dave

[penngray]

No longer on the east coast. I've retired to the foothills in northern Georgia, just a few miles from the southern end of the Appalachian trail. Family here, the weather is milder, large lake nearby...and property taxes are less than 1/5 what I was paying in NJ .

The good fortune of low cost living in Georgia, Very nice!!

Come down to Florida where there is no state income tax but the property tax is sky high (about 2% of the house value and the house values were high before the crash), Yikes!!