Geddes on Waveguides

Jump to Latest
#2,983
#2,984
ZilchLab said:
...

with a recommended crossover frequency of 1.9 kHz, I don't believe 1.4 kHz is "well within its operating range" of 4540....
Click to expand...


I won't worry much about the lower limit of such a driver in home use.

114dB/1W/1m sensitivity on CD horn
max SPL = 132dB @ rated power 60W

How can one approach that in home use? I bet it'd be OK in several hundred Hz range as long as the loading is adequate and the input power is low. It's low at home use anyway. I guess 1W is already the ultimate limit in ordinary domestic environment.

In the chart, I see it pretty good down to about 800Hz, where the 2nd HD is even lower than that of 1.5kHz and 3-5kHz.

And the distortions at 1kHz are lower than 2kHz.

Interesting.
 
#2,985
ZilchLab said:
You're not seeing the radial re-entry slotted compression chamber on the back side of the BMS diaphragm, and with a recommended crossover frequency of 1.9 kHz, I don't believe 1.4 kHz is "well within its operating range" of 4540:

http://bmspro.com/4540ND.bms_4540nd_compression_driver.0.html

4552nd will play there, though:

http://bmspro.com/4552ND.bms_4552nd_compression_driver.0.html

These are not so inexpensive as they once were, BTW....
Click to expand...

But the 4540ND is uniquely suited for playing the top octave cleanly. I am not aware of another compression driver that can play as high. The 4552, like most compression drivers, runs out of steam half an octave earlier.

Yes, it opens up the whole debate about ultrasonics, but I can hear the difference like night and day. Every time I switch from the 4540 to a conventional compression driver, it sounds like the entire top octave is AWOL. It is NOT subtle.

The output beyond 20khz is smooth and measurable, I've done a ton of work with the 4540.
 
#2,990
catapult said:


Earl, are you sure it was German-made BMS? I think this came up before and it was another brand with a similar name that you had used.
Click to expand...


I haven't looked at their products in years. It was a BMS, it was a coaxial unit. On my waveguides the crossover between the diaphragms created a polar mess at the crossover point which was right in the hearing bands most sensitive region. Thats what I didn't like. I know that their ring diaphragm and unique phase plug can have a very good performance.

But as I have said before, to me compression drivers are a commodity. I will most likely use the B&C until I can have a custom one made for me in China. I just don't see any compelling reason to "check-out" all of the individual choices out there.

I look at one thing first - price. If its not competitive with what I get the B&Cs for then I go no further. If it is, like some Eminence stuff, then I look deeper. In the case of Eminence they were not able to get me the hole mounting patteren that I needed and I would have had to change all of my fixtures and mounting. It wasn't worth it.

Only the Chinese made drivers that I sampled several years ago ever really got my attention. Extremely close copy of the B&C, extremely cheap and I could have told them all the things that they were doing wrong. Someday I will buy enough drivers to get them from China, made right, even with some improvements.
 
#2,991
A 2005 AES convention paper by Gunnes points out that a major failing of compression driver phase plugs is that even if the dimensions are optimum only half of the incident power actually makes its way through the phase plug to the throat, the rest of it is reflected .

If we look at a ring radiator we can see that the reason for the slots and holes in the average phase plug is removed because the phase difference across the radiating surface is small, because the radial distance across the radiating surface is small.

The point about this is that then the phase plug can consist of a central body and an outer annulus that does not cover any of the radiating surface and yet still provides suitable compression.
Such a design should reduce the energy storage mechanism outlined by Gunnes.
Rcw.
 
#2,992
rcw said:
A 2005 AES convention paper by Gunnes points out that a major failing of compression driver phase plugs is that even if the dimensions are optimum only half of the incident power actually makes its way through the phase plug to the throat, the rest of it is reflected .

If we look at a ring radiator we can see that the reason for the slots and holes in the average phase plug is removed because the phase difference across the radiating surface is small, because the radial distance across the radiating surface is small.

The point about this is that then the phase plug can consist of a central body and an outer annulus that does not cover any of the radiating surface and yet still provides suitable compression.
Such a design should reduce the energy storage mechanism outlined by Gunnes.
Rcw.
Click to expand...

When I worked at B&C we concluded that the main advantage of the ring diaphragm was reduced mass. That's because the center of the diaphragm does not contribute very much radiating area, but it needs to be stiff at that point in order to minimize diaphragm resonances. The ring reduces the radiating area far less than one can reduce the mass and still maintain adequite stiffness. We did not at that time see any advantage from the phase plug and I still don't.

Do you have a reference on the Gunnes paper? Preprint number?

The phase plug reflects energy by virtue of the compression ratio but this is completely independent of how that ratio is achieved. The ring radiator allows for a lower compression ratio without substantial phase cancellation, thats true, and the energy reflection is reduced by virtue of the lower compression ratio, but a lower compression ratio also reduces efficiency so the net result is not much change, which is the case with the ring radiators - a little more output and a small increase in the high end, but its not dramatic.
 
#2,993
gedlee said:



...
Only the Chinese made drivers that I sampled several years ago ever really got my attention. Extremely close copy of the B&C, extremely cheap and I could have told them all the things that they were doing wrong. Someday I will buy enough drivers to get them from China, made right, even with some improvements.
Click to expand...
For orders around 500~1000, normally they should be able to do some work and modifications. However, when driver costs get to significantly low compared to the whole system, it should not be an issue. There could be different ways to work with companies.
 
#2,996
The Gunns paper is a convention paper and the copy I have has no pre print number and was presented at the 119th. Convention.

I was thinking of a low compression device that has nothing in front of its radiating surface, therefore no definable cavity as such.

The thought I had was that if you do this the trade off of phase plug to diaphragm spacing verses upper frequency roll off is relaxed and you can use a larger diaphragm displacement.
In this the phase plug section of the device is a converging tube, virtually a back to front annular conical horn.
It occurred to me that when mated to a suitable wave guide a conjugate match might be able to be obtained between the convergent and subsequent divergent sections, allowing you to get a purely resistive diaphragm load .
rcw
 
#2,997
At a trade show, I spoke to a very well respected transducer engineer at JBL about his thoughts on the BMS compression drivers. He described them as "over engineered" and left it at that.

I wasn't fluent enough in this sort of thing to take that conversation much further.

I understand that, at one point, JBL licensed this technology and used it in a couple of products (like the 2408H / 2408J compression driver) but it appears to have been ultimately abandoned or at least is out of favor, for whatever reason.

I anecdotal heard something from Jack at Assistance Audio about BMS manufacturing the diaphragms for JBL at one point, as JBL was having trouble with them but that is just anecdotal.
 
#2,999
soongsc said:

I don't quite understand. If drivers are a commodity, and if they can be produced by commonly available methods, they should be pretty low compared to specially designed enclosures.
Click to expand...

In China they are, but those drivers are not available in the market here. That was precisely my point, that a very low cost driver from China can be as good as a much more expensive unit sold in the market here. Pretty much shows that they are a commodity.


rcw said:

It occurred to me that when mated to a suitable wave guide a conjugate match might be able to be obtained between the convergent and subsequent divergent sections, allowing you to get a purely resistive diaphragm load .
rcw
Click to expand...

What is the facination with "purely resistive diaphragm load"? What difference does it make? The load on a piston source isn't purely resistive. When it is reactive its just a simple mass load - whats the big deal? At any rate what you suggest wouldn't do that anyways. There is always going to be a reactive component whenever the load is changing such as at the lower end of the passband. All you can do is move the passband of the waveguide well below that of the source, which isn't practical given that it won't really change anything.
 
You must log in or register to reply here.
Top Bottom