Thanks guys,
I guess that's enough to put some doubt on my foam - it must be semi-reticulated 🙂
Maybe I'll stick it in the grill, fill 'er up with propane, and light it up. That oughta do it? (that was a joke)
Dr. Geddes,
Do you care to comment on the looks of the driver exit I posted above? Based upon your experience, does it 'seem' to be a good candidate to mate to an OS waveguide (please forget it is an Altec for a moment)?
I guess that's enough to put some doubt on my foam - it must be semi-reticulated 🙂
Maybe I'll stick it in the grill, fill 'er up with propane, and light it up. That oughta do it? (that was a joke)
Dr. Geddes,
Do you care to comment on the looks of the driver exit I posted above? Based upon your experience, does it 'seem' to be a good candidate to mate to an OS waveguide (please forget it is an Altec for a moment)?
The tangerine phase plug design has never really impressed me theoretically and its poor longevity in the marketplace is kind of proof that the idea wasn't really very good. So, sorry, its not a good choice.
While we're on the topic of foam I've been meaning to ask if other materials would work here. Wool, fiberglass etc? I'm sure this would have been asked before but with more than three threads on the subject dating back to last summer it's s tough job to find a specific question.
Well lets just say that the low loss, controllable cell size and highly uniform nature of the foam is ideal. As to other materials more readily available I will have to say that it's possible, but I haven't tried.
gedlee said:
If I may, the fact that the tangerine phase plug didn't last on the market very long doesn't prove very much at all IMHO. The commercial market isn't large enough and independent enough to allow that sort of fair market outcome. (unrelated note, read "Wisdom of Crowds" by can't remember at the moment...great book).
What exactly about the tangerine phase plug doesn't jive well from a theoretical standpoint?
It doesn't solve the problem of creating a plane wave at the throat. What its stated objective was to make a phase plug that was easier to manufacture, and in that regard it was a success. But its HF performance is not very good and as such it didn't last the test-of-time. Current phase plugs are not optimum, but the tangerine is even less optimum than what is used in nearly 100% of compression drivers today.
In the early days the phase plug was cast, in parts, in alluminum and had to be machined to its final fit. In this era the plastic tangerine plug was very attractive from a cost standpoint. But today plastics have gotten so good and so cheap that acceptable phase plugs can simply be cast in plastic with no extra machining required. So the "need" for the tangerine plug is no longer an issue.
My personal favorite phase plug would probably be the "salt-shaker" design as this can do a really good job of getting the optimum wavefront at the throat. Its not that much more difficult than the tangerine and can be done in one piece.
In the early days the phase plug was cast, in parts, in alluminum and had to be machined to its final fit. In this era the plastic tangerine plug was very attractive from a cost standpoint. But today plastics have gotten so good and so cheap that acceptable phase plugs can simply be cast in plastic with no extra machining required. So the "need" for the tangerine plug is no longer an issue.
My personal favorite phase plug would probably be the "salt-shaker" design as this can do a really good job of getting the optimum wavefront at the throat. Its not that much more difficult than the tangerine and can be done in one piece.
gedlee said:
I would have to disagree with that assessment. In the AES reprint #1328 Henricksen clearly showed that the radial phasing plug actually has better high frequency performance. While the radial plug starts to roll off sooner, it does so at a much slower rate than the circumferential plug. The circumferential plug is quite good out to a certain point then drops like a rock. The radial plug has a long slow high frequency tail that extends much higher in frequency. The end result is the radial plug has a 5dB to 6dB advantage at 20KHz over the circumferential plug. My own personal experience has been it is much easier to get to 20KHz with radial plug equipped drivers. They require much less equalization to get flat frequency response which also results in a 3dB-4dB driver sensitivity advantage for the radial plug. This is because you don’t have to pad down the mid-band of the radial plug equipped drivers as much to get them flat.
I agree with JoshK's statement - There was no "true" test of time for this type of phasing plug due to the declining state of the industry at that time.
I would prefer the blue curve.
If twenty years doesn't represent the test of time, I don't know what does. The industry went through a serious peak in that time frame.
I did hear that Eminance is looking at the radial phase plug for their drivers.
If twenty years doesn't represent the test of time, I don't know what does. The industry went through a serious peak in that time frame.
I did hear that Eminance is looking at the radial phase plug for their drivers.
One more idea for a mold made of metal that can be used to produce a glass-reinforced plastic waveguide:
There are basically two parts that are screwed together. Then resin and fiberglass is applied in layers by hand. After curing the two parts are simply screwed apart.
The drawing shows how screw nuts can be included so the final waveguide can be mounted with tight tolerances.
Best, Markus
An externally hosted image should be here but it was not working when we last tested it.
There are basically two parts that are screwed together. Then resin and fiberglass is applied in layers by hand. After curing the two parts are simply screwed apart.
The drawing shows how screw nuts can be included so the final waveguide can be mounted with tight tolerances.
Best, Markus
poptart said:
Take a look at the pics I posted of my Summas from the side - the face of the foam plug is spherical. You can't maintain that shape with wool or polyester stuffing.
BTW, the foam plug is spherical so that the on and off axis frequency response is consistent. If it was flat faced, there would be greater attenuation off-axis.
Markus
This doesn't solve the shrinkage problem and in fact may make it worse.
Things like this are never as simple as they appear until you try it.
We tried embedded molies in the fiberglass, but gave up as the shrinkage moved the locations and then the molies were in the wrong places and useless.
Of all the processes that I have tried the one that I am using now appears to work the best. Thats because I CAN get (and have) perfect parts this way, although they don't always come out perfect. But the fact that it can be done means that with time and further refinement I can get the process to one that is repeatable and accurate. This is why I simply don't understand all this talk about changing it. I have no desire to change it only to improve the existing one.
I learned recently that the activator settles and can do so in days. Thus it has to be agitated before each mix. And that the material cures too fast and that I need a different activator which will allow a slower cure, resulting in less heat and a better repeatabilty (less dependence on ambient temperature and humidity). Each time I make a baffle it is better in most ways than the previous one, but often with a new problem to fix.
This doesn't solve the shrinkage problem and in fact may make it worse.
Things like this are never as simple as they appear until you try it.
We tried embedded molies in the fiberglass, but gave up as the shrinkage moved the locations and then the molies were in the wrong places and useless.
Of all the processes that I have tried the one that I am using now appears to work the best. Thats because I CAN get (and have) perfect parts this way, although they don't always come out perfect. But the fact that it can be done means that with time and further refinement I can get the process to one that is repeatable and accurate. This is why I simply don't understand all this talk about changing it. I have no desire to change it only to improve the existing one.
I learned recently that the activator settles and can do so in days. Thus it has to be agitated before each mix. And that the material cures too fast and that I need a different activator which will allow a slower cure, resulting in less heat and a better repeatabilty (less dependence on ambient temperature and humidity). Each time I make a baffle it is better in most ways than the previous one, but often with a new problem to fix.
Patrick Bateman said:
You know, this is a good point. Initially I expected a concave shape to widen the directivity from a "lense" effect, but did not see that. But if the directivity were widened while the off-axis were further attenuated then the effects would cancel. My initial thoughts of the "lense" effect ignored the attenuation effect (whcih is always ignored in glass lenses) and now I can see the error of that thinking.
Initially I had wanted to use the foam as a refractive medium to further control the directivity, but it didn't work. However the sound quality was dramatically improved and I was very curiuos about this. Thats when Lidia and I proceded to look at the audibility of the HOM as a way to explain what was observed subjectivly but not explainable objectively. After several more years of study I now realize that the foam controls the HOM even if it has no effect on the directivity.
And now I understand why it didn't afect the directivity as I thought it should.
Your speakers sound so good, if I win the lottery I'm going to DONATE them to the clubs in town. Last weekend there was a local music festival that spanned four days in the city, and by the 2nd day I couldn't take it any more. The sound systems were just TOO AWFUL. All the clubs are using arrays now, which are inexcusably loud and awful sounding. We're talking about venues with 50-100 people, with a line array pointed at the audience from eight feet away. The sound is just painfully bad.
metal mandrel
This is pretty much what the shape of the bottle top section I'm working on as a proof-of-concept for all doubting Thomas's...
Stay tuned... first will be in copper, maybe with a nickel backup... then who knows
metal will solve many of the dimensional stability issues... since the technique is mature and widely used for ultrahigh precision fabrication of complex shapes
John L.
markus76 said:
This is pretty much what the shape of the bottle top section I'm working on as a proof-of-concept for all doubting Thomas's...
Stay tuned... first will be in copper, maybe with a nickel backup... then who knows
metal will solve many of the dimensional stability issues... since the technique is mature and widely used for ultrahigh precision fabrication of complex shapes
John L.
markus76 said:
A more "commonly" use method is to simply glue the parts together
But its not very hard to make a 2-part mold consisting of the waveguide mold and on top the mold fore mounting plate part
Before laminating the 2 molds are screwed together with a little clay in between
Another advantage from this is the maintained precision of throat
Picture shows the 2-part mold
Attachments
Your construction is even simpler and better. But still there's the problem with the fiberglass shrinking...
So I'm really curious how John's electroformed part comes out.
Best, Markus
So I'm really curious how John's electroformed part comes out.
Best, Markus
But what exactly does an electroformed part do to solve anything? Its just a skin - there is a lot more to a waveguide than this, like walls thick enough and damped enough to be non-sound transmitters, and mounting the driver and the waveguide itself. If the electro-formed part only becomes a model on which to make other parts then this is already easily done to very high tollerances with machining.
I'm lost as to why an alternate process is desirable - to save on cost? How is that going to be accomplished? I can't imagine anything costing less than the cast parts that I make now - they are a fraction of the cost of the fiberglass ones. I have no problem with bantering about different ideas, but I think that it would be wise to have some sort of a goal against which to judge these ideas.
I'm lost as to why an alternate process is desirable - to save on cost? How is that going to be accomplished? I can't imagine anything costing less than the cast parts that I make now - they are a fraction of the cost of the fiberglass ones. I have no problem with bantering about different ideas, but I think that it would be wise to have some sort of a goal against which to judge these ideas.
gedlee said:
yep..you're lost.
isn't necessarily a skin.. but it can be. The process is alot simpler than casting, especially after the first CORRECTLY MADE piece is confirmed. Trust me I've built assemblies much more complex with constrained price points than you can imagine, but you seem to be a non-believer in an ivory tower, so i'll just proceed w/o your endorsement.
How much of your time does it take to cast each waveguide? 15', 1/2 hour? I understand if you don't want to divulge this, seeing as you seem to believe your method is IP. However, very little machining is involved to duplicate an existing waveguide or mandrel. That's one of the advantages, you can make multiples of either the mandrel or the finished part. As for being a skin, that's only one possibility. You're also not limited to symmetrical shapes (think of an elliptical guide) spend lots of time on a hand-built mandrel, then easily duplicate it in metal at RT.
I really could care less whether you think this is a good idea. I'm looking beyond your OS guides to other shapes, etc. asked for by others. Me thinks thou doest protest too much
John L.
Is the goal with electroforming simply to make the outer face of the waveguide, thus allowing some material to fill in behind the metal to provide the desired damping? That seems like a reasonable approach to me in that it reduces the requirements for precision molds and maybe alleviates some of the issues with curing by allowing different materials to be used for adding the additional mass.
I certainly have no frame of reference for whether or not it would be a more cost effective approach, but it does seem like it would be an easier way to construct them.
Here's how I'm envisioning the general approaches and someone correct me if I'm off base. I'm going to use a jello mold that is for appearances and not consumption for my example.
Current situation. The jello pan is very precise and strong and costs a fair amount. High quality jello must be used to ensure the jello sets well and retains the proper shape. Jello is removed from the pan prior to sale.
Electroforming. The jello pan is very precise but also somewhat thin. The pan is sold with the jello, so really any crappy jello can be used because all we care about is the shape of the jello and the thin pan has that part covered, the jello is just there for mass.
I know that's a simpleton's view but that's roughly my level of understanding of these processes.
I certainly have no frame of reference for whether or not it would be a more cost effective approach, but it does seem like it would be an easier way to construct them.
Here's how I'm envisioning the general approaches and someone correct me if I'm off base. I'm going to use a jello mold that is for appearances and not consumption for my example.
Current situation. The jello pan is very precise and strong and costs a fair amount. High quality jello must be used to ensure the jello sets well and retains the proper shape. Jello is removed from the pan prior to sale.
Electroforming. The jello pan is very precise but also somewhat thin. The pan is sold with the jello, so really any crappy jello can be used because all we care about is the shape of the jello and the thin pan has that part covered, the jello is just there for mass.
I know that's a simpleton's view but that's roughly my level of understanding of these processes.