Geddes on Waveguides

MJL21193 said:
Thanks Earl, I appreciate these concise answers.
I will have to explore this for myself, to see if it makes a reasonable improvement in SQ.
I would like to try it with a regular dome tweeter though, but in your conversation with rcw, you made it seem as though this would not be a good approach.

See, I'm not always snide. 🙂


NO, that is not what I said to RCW. I objected to his assumption that a dome in a conical horn was the "optimum" approach. That I don't agree with. But a dome in a conical horn is far far better than a dome in a baffle.

I know that its not your fault, but it is so very difficult when people misunderstand what is said and then I take heat for it. RCW's discussion and mine was on a fairly small point, nothing that would completely invalidate what he suggests. I just believe that there are better ways to do things. But as I told him, in one of my designs I actually did use a dome tweeter on a conical horn just as he suggests - because it is dirt cheap, not because it is the optimum.
 
gedlee said:

...
The use of asymmetric waveguides is of extreme interest to me and has been for years and I would dearly love to pursue the topic both with actual designs and constructions as well as technical discussions. But the construction part has alluded me for years owing to the exhorbitant cost increase involved in their fabrication. But discussion is important because it helps to hone the design objectives and so perhaps this should be continued in the hope that someday I will have the resources to pursue the actual designs.
Have you actually tried it and measurements indicate that they need foams?
 
So I could play with a dome tweeter version to see how I like the results, then if I did, I'd go to a compression driver.

For me, the construction is not the hard part, so I can easily spare a few hours to mock up a viable example.

Here's a quick section view. Is this approximately correct?
 

Attachments

  • 1a1.gif
    1a1.gif
    4.2 KB · Views: 537
Most dome tweeters I've seen (not cheap ones) have a bit of a flare on the faceplate around the dome. See the revised pic below - the black is the faceplate.
Construct the horn to match the contour of this flare.
 

Attachments

  • 1a1.gif
    1a1.gif
    4.8 KB · Views: 524
gedlee said:


A rational comment - how refreshing!!

No actually the long axis is horizontal. Thats where the shorter distance comes from.


I realize that the shorter distance would come about with the long axis horizontal. What I was thinking is that the large axis would provide a more narrow polar response than the narrow axis, and therefore you'd want the long axis vertical to narrow the vertical response.

If the long axis has a wider response, is it due more to the angle of the waveguide?
 
A point about dome tweeters that Earl made is that various makes of them have domes of different curvatures.

Metal domes are generally too flat for 80-90 degree waveguides, and ideally a transition region to "bend" the wavefront onto a different curvature should be used.

With the D25AG I put a 50mm. hole saw through the face plate and provided the waveguide mounting plate with a spigot that fits into the hole and clears the center of the surround by a half millimeter.

The seas 29mf has a rated .5mm. peak diaphragm excursion and I modeled it on winisd.
With a biquad at 1500Hz. and a fourth order L-R filter at the same frequency it can produce peaks of 113db. with a 90Watt input and the excursion has a maximum of 0.27mm.

In a wave guide you can expect at least 5db. more output bellow the mass corner, meaning in the average room a pair can give short peaks of 120db. at only half the rated linear excursion, that should be about enough for anybody.
rcw
 
rcw said:
A point about dome tweeters that Earl made is that various makes of them have domes of different curvatures.

Metal domes are generally too flat for 80-90 degree waveguides, and ideally a transition region to "bend" the wavefront onto a different curvature should be used.

With the D25AG I put a 50mm. hole saw through the face plate and provided the waveguide mounting plate with a spigot that fits into the hole and clears the center of the surround by a half millimeter.

The seas 29mf has a rated .5mm. peak diaphragm excursion and I modeled it on winisd.
With a biquad at 1500Hz. and a fourth order L-R filter at the same frequency it can produce peaks of 113db. with a 90Watt input and the excursion has a maximum of 0.27mm.

In a wave guide you can expect at least 5db. more output bellow the mass corner, meaning in the average room a pair can give short peaks of 120db. at only half the rated linear excursion, that should be about enough for anybody.
rcw


Hi RCW,

This is not a criticism of your idea, but just some food for thought. You can pump 90 watts into a dome tweeter for a short duration, or you can furnish an 108dB sensitive compression driver 16 watts all day long. Both equal 120dB (tweeter per your analysis). However, the compression driver will have multiples less distortion and will not go up in a poof of smoke. Why chase down this path? If its just in the spirit of DIY, then that's cool. It just doesn't seem a logical path to me. Once again, no offense intended, just sharing ideas here.

Rgs, JLH
 
rcw said:
In a wave guide you can expect at least 5db. more output bellow the mass corner, meaning in the average room a pair can give short peaks of 120db. at only half the rated linear excursion, that should be about enough for anybody.

... reminds me of that '640K ought to be enough for anybody' quote.

Above everything here, we are talking about the qualities and properties of the waveguide. Maximum sound pressure is not one that seems to be lacking in any way (as used on the Summa speakers), and 'about enough' is certainly not something that I have in mind.

It is certainly interesting to see what does and does not work in a waveguide though.
 
hi Earl

i don't know it this is offtopic, but something i would like to know your opinion. I had a 300hz tractrix horn with BMS 4592nd compression driver in the past, which was covering all the way up to 20khz. The treble was not sound right to me . Thinking about it, i can immagine, that the reason was HOM. Now i used a Radian 8" coax driver for a project, with a 1" compression driver covering from 1,2khz up, for a small room, about 4m x 4m . It has a 90 degree opening like your waveguide, but is not OS. Well, the treble is simply perfect . I did't here anything, that might bother me because of HOM. Since it is a point source, the sound is so coherent, no phase shift, no detection of the source, the sound simply flows in the room. If it would be 2 channels, with a woofer and a wave guide above, certainly the source would be detectable, and the sound not so coherent as with a coax. What makes you prefere the Summa design to a design with a coaxial driver ? I mean , the advantage of a point source is not insignificant, while i didn't bother with HOM, and directivity might be the same as your OS.

Angelo
 
pooge said:


If the long axis has a wider response, is it due more to the angle of the waveguide?

Yes, the angle of the walls determines coverage angle if 1) the waveguide is large enough 2) the frequency is high enough such that there is more than a wavelength across the mouth. Below this frequency, the polar response gets very complicated and hard to predict.
 
rcw said:
In a wave guide you can expect at least 5db. more output bellow the mass corner, meaning in the average room a pair can give short peaks of 120db. at only half the rated linear excursion, that should be about enough for anybody.
rcw


We don't agree on this point as I would caution that while you may not burn up the dome, thermal compression will be quite evident at these levels. The compression driver will still be idling.
 
angeloitacare said:
hi Earl

i don't know it this is offtopic, but something i would like to know your opinion. I had a 300hz tractrix horn with BMS 4592nd compression driver in the past, which was covering all the way up to 20khz. The treble was not sound right to me . Thinking about it, i can immagine, that the reason was HOM. Now i used a Radian 8" coax driver for a project, with a 1" compression driver covering from 1,2khz up, for a small room, about 4m x 4m . It has a 90 degree opening like your waveguide, but is not OS. Well, the treble is simply perfect . I did't here anything, that might bother me because of HOM. Since it is a point source, the sound is so coherent, no phase shift, no detection of the source, the sound simply flows in the room. If it would be 2 channels, with a woofer and a wave guide above, certainly the source would be detectable, and the sound not so coherent as with a coax. What makes you prefere the Summa design to a design with a coaxial driver ? I mean , the advantage of a point source is not insignificant, while i didn't bother with HOM, and directivity might be the same as your OS.

Angelo

These comments are all subjective based on some causual listening and they are being compared to a situation that you have not heard. I have heard both and I don't find your comments correct. "the treble is simply perfect " - these kinds of superflous statements are always disturbing to me.

I would rather use a "optimum" waveguide, even if it can't be coaxial, than the highly compromised waveguide required for coaxial mounting. In my experince the tradeoffs are not good ones - the treble is simply "not" perfect.
 
These comments are all subjective based on some causual listening and they are being compared to a situation that you have not heard.

hello Earl

my comments come not as a scientist, as you are, but based on what counts, which are the normal subjective listening experiences in a home environment . I know the setbacks of more than one source, it can be perceived clearly when sound comes from more than one source, integration is never as perfect, as from one point source. The treble is perfect for me, when i am satisfied and pleased with most kind of music. Of course, this is a personal subjective opinion, which others might not agree with.

I would rather use a "optimum" waveguide, even if it can't be coaxial, than the highly compromised waveguide required for coaxial mounting. In my experince the tradeoffs are not good ones - the treble is simply "not" perfect.

the quest remains which trade-off is worse, two different sources, or a highly compromised waveguide, which does not aloud a " perfect " performance .

Angelo