Home » Sponsored » Pi Speakers » HF Compression driver without horn flare?
Re: Pipe modes [message #47176 is a reply to message #47174] Thu, 23 June 2005 06:33 Go to previous messageGo to next message
Greggo is currently offline  Greggo
Messages: 36
Registered: May 2009
Baron
Well, if I am following this correctly I would have to guess that the pipe resonance of the bare driver has been addressed by removing phase plug and re-working or removing the throat of the TAD 4001. Also, from the web site, it seems the motivation is that the designer feels that this achieves better driver integration with regard to point source driver integration issues than if he were to use a horn. I don't have an opinion as to better or worse, but find it interesting and hope I get a chance to hear their speaker some day.

In the meantime, this forum rocks and I hope to build a pi kit at some point soon. My current hifi set up is becoming boring and I would like to hear the dynamics that all you horn fans are always talking about...

Regards,

Greg Jensen

Re: Pipe modes [message #47178 is a reply to message #47176] Thu, 23 June 2005 07:02 Go to previous messageGo to next message
Wayne Parham is currently offline  Wayne Parham
Messages: 18783
Registered: January 2001
Illuminati (33rd Degree)

A cone or dome driver doesn't have pipe resonance, because there is no pipe. A very large horn doesn't either. It's when horns are too small that they start acting like pipes.


Re: Pipe modes [message #47179 is a reply to message #47178] Thu, 23 June 2005 07:48 Go to previous messageGo to next message
Greggo is currently offline  Greggo
Messages: 36
Registered: May 2009
Baron
Yes, makes sense, I thought the TAD driver in question had some kind of throat as part of the assembly thus creating the potential for resonance if not directed into a horn. Not sure if that makes sense either, I just assumed from the earlier post that there was something about the physical housing of the driver that would create problems if run without a horn attached... I would love to try a TAD driver in a DIY project someday, but will probably start with a JBL from you when I am ready. Threads like these just open my mind up to different experiment and design possibilities outside of the box, always good to at least think outside the box now and then.

Greg Jensen

Re: Pipe modes [message #47180 is a reply to message #47179] Thu, 23 June 2005 12:54 Go to previous messageGo to next message
Wayne Parham is currently offline  Wayne Parham
Messages: 18783
Registered: January 2001
Illuminati (33rd Degree)

Compression drivers do have a throat, and they also have a phase plug. It's short, a stub really, but it is placed there because the device is designed for a horn. It would be nice if they weren't made that way, and that the phase plug were part of the horn instead. That would make more sense, because the phase plug is sort of a wave-shaping device that directs the wavefront into the horn, matching the diaphragm to it and minimizing path length differences. But the fact is that compression drivers have a phase plug in front of the diaphragm, and the exit is smaller than the diaphragm.


Re: Pipe modes [message #47183 is a reply to message #47180] Thu, 23 June 2005 19:42 Go to previous messageGo to next message
16 is currently offline  16
Messages: 4
Registered: May 2009
Esquire
All reasonable horns that might be considered for a compression driver such as the TD4001 and others of its type exibit frequency-dependent resonances depending on the size and geometry of the horn. There is a fundamantal resonance frequency, and also harmonics of the fundamental higher up in frequency. These can be detected by doing impedance vs frequency measurements of compression drivers with various horn assemblies, and they will show up as pronounced peaks in the impedance curve.
The 4001 has a 3" long conical throat that can't be avoided due to its large alnico magnet structure...that is just the way it is made. Other drivers in the TAD range do not have such a long throat section due to their very much smaller Nd magnet structures. In fact the 4002 pretty much has almost no throat to speak of.
The 3" throat inherent to the 4001 will have some resonances, albeit at fairly high frequencies. However, the point to remind oneself of is that if you listen to a comp. driver without a horn and you like the way it sounds, then your brain has already computed that any resonances it may have are either musically insignificant, or at least musically acceptable. In respect to our Model 16 design, the hornless approach had many subjective virtues associated with it that made it a compelling and appealing trade-off.
For most of the development time during the Model 16 project we were sure that the final design would include a horn, and as mentioned some fairly serious trials were undertaken in this area. But at the same time we had also put in many hours of low-level listening running the 4001 sans any kind of horn, and we had become accustomed to the way it was speaking to us. It was ultimately felt that there was something very direct and special about the hornless approach, and so we though "f__k it, let's just develop it along these lines and try and get it to work..."
There are actually many little tweaks built in to the Model 16 system that aren't obvious on paper or to the eye...the hornless thing just happens to be one of its very obvious deviations from standard practice.

Re: Pipe modes [message #47184 is a reply to message #47183] Thu, 23 June 2005 20:54 Go to previous messageGo to next message
Wayne Parham is currently offline  Wayne Parham
Messages: 18783
Registered: January 2001
Illuminati (33rd Degree)

I don't agree with your assessment of horns, particularly when used at high frequencies where wavelengths are small. Even a relatively small device is possible for use above the first or even the second pipe mode. I have reservations about the case for pipe modes of basshorns, but not tweeters.

Since you have 3" of throat mounted on the device, it seems to me that the best thing for you to do would be to complete the horn by adding a flare. A 3" section has a quarter-wave mode at 1kHz and half-wave at 2kHz, so I think the best thing you could do would be to complete the flare by adding a horn as the device was intended. I would prefer to either have a direct radiator or a full sized horn than a truncated pipe.


Re: Pipe modes [message #47186 is a reply to message #47184] Thu, 23 June 2005 23:14 Go to previous messageGo to next message
16 is currently offline  16
Messages: 4
Registered: May 2009
Esquire
Hi Wayne, thanks for your post. If we take an example horn of length in the range of 250mm to 300mm, depending on the chosen geometry and mouth size there will be resonances loosely centered in the 400Hz-500Hz range, the 800Hz to 1kHz range, and the 1.2kHz to 1.5kHz range. I think the point you are making is that if we are using the device such that we are filtering it pretty high, then the resonances will be less of a problem as the higher harmonics are of less magnitude than the fundamental mode. On paper this is correct, although it is fair to say that the second mode (which is still of considerable magnitude) is still in the picture.
I also agree that intuitively one can't help but feel that a 3" conical throat truncated abruptly is something to be avoided like the plague. Well, the thing is, is that I kept telling myself this over and over again, and at the same time was spending many hours listening to the 4001 raw and thinking "this sounds bloody good, but I am looking forward to hearing it when it is properly coupled to a suitable horn". The plain fact is that when I started playing with horns I was never able to achieve certain subjective characteristics that were apparent with a hornless implementation. Just like the comparison of wide baffles giving more loading and support to a woofer, and narrow baffles losing some of that support in trade for a more "open and spacious" presentation, so it is that using a device like the 4001 without horn can open up a world of alternative sonic possibilities. The other interesting thing is that during our subjective evaluations we had all of TADs drivers on hand simultaneously. This meant that we had drivers such as the 4002 and 4003 which have vastly shorter throat sections than the 4001. As mentioned earlier, the 4002 has virtually no throat to speak of so you actually get to hear the thing speaking to you right off the phase-plug. Once again, I thought many times that either you should have a fully developed and extended horn, or no horn at all like the 4002. The in-between situation of a 4001 with no horn was intuitively the worst of both worlds. Well all I can say is that I just listened and listened and could not in the end ignore the sheer musical energy the 4001 sans horn was capable of.
Don't they say something about the bumble-bee not being aware that the scientists say it can't fly...so it just does?

Re: Pipe modes [message #47187 is a reply to message #47186] Fri, 24 June 2005 06:33 Go to previous messageGo to next message
Wayne Parham is currently offline  Wayne Parham
Messages: 18783
Registered: January 2001
Illuminati (33rd Degree)

Pipe modes are caused by standing waves which setup where there are abrupt transitions in the pipe. The discontinuity caused from a straight pipe or one with only a slight flare suddenly terminating in an open end is much greater than a pipe that flares widely and gradually. A compression tweeter horn has 50 or 100 times more mouth area than throat area, and that prevents response peaks in the passband.

One thing that might be cool in your application would be to radius the tweeter exit opening on the baffle. A radiused baffle would help smooth the transition and might reduce pipe modes. You would essentially be adding a little bit to the throat flare, and by rounding it to match the baffle, it would help transition the wave from the throat to the half-space condition of the baffle.


Re: Pipe modes [message #47191 is a reply to message #47187] Fri, 24 June 2005 20:14 Go to previous message
16 is currently offline  16
Messages: 4
Registered: May 2009
Esquire
Yes pipe modes are caused by abrupt terminations, and it is also true that many commonly used horn flare geometries at some point or other manifest these sudden transitions. The fact that any horn, by virtue of the fact that it IS a horn, loads a driver more severely then any baffle (and we all know that even baffles have diffraction effects once the wave reaches the point of zero support) means that at some point you have got to start making the transition to a baffle...i.e. increasing the flare angle towards the end of the horn (unless as you mentioned earlier you decide to filter the driver such that it is operating way above the cut-off frequency of the horn in question). How successfully this transition is accomplished is I guess in simple terms the reason why people experiment with curves like circular mouth tractrix flares and the like in the first place...they are looking for that ideal smooth acoustic impedance transformation.
I'm pretty happy with what I've been able to accomplish in subjective terms with the Model 16 concept. As far as adding an extra little radius section to the front of the 4001...I did try it actually and I didn't like it as much...and on another occasion I sent a 4001 to the CNC shop to radius the end of its throat a little (which required filling behind its plastic throat with resin as well as accepting the fact that I would be cutting slightly into the iron magnetic return path...don't really wanna go there again...but if anyone else wants to try it and let me know the results I look forward to it...)
Best regards,
Hassan.


Previous Topic: 2206
Next Topic: Pro 3 Pi
Goto Forum:
  


Current Time: Sun Nov 17 15:54:17 CST 2024

Sponsoring Organizations

DIY Audio Projects
DIY Audio Projects
OddWatt Audio
OddWatt Audio
Pi Speakers
Pi Speakers
Prosound Shootout
Prosound Shootout
Miller Audio
Miller Audio
Tubes For Amps
TubesForAmps.com

Lone Star Audiofest