If you model the chamber as throat ported, you can add a little detail - like the length of the throat port. This is easy to do and it has a significant effect.
One of the things you can't model in HR is the phase aspect of a phase plug - forcing sound from different parts of the cone to take different length paths so it all gets to the horn entry at the same time to form that plane wave front. Do you need to design the phase plug to minimize the maximum path length delta? Depends on the maximum path length distance from any point on the cone to the horn entry compared to a wavelength of the highest frequency in the horn's pass band.
One of the things you can't model in HR is the phase aspect of a phase plug - forcing sound from different parts of the cone to take different length paths so it all gets to the horn entry at the same time to form that plane wave front. Do you need to design the phase plug to minimize the maximum path length delta? Depends on the maximum path length distance from any point on the cone to the horn entry compared to a wavelength of the highest frequency in the horn's pass band.
I'm thinking he did not gap the throat, as per the Bruce Edgar article I referred to as recommended reading.
Also, I really can't say I remember the time line to when he added that mass slug
to the front of the cone of said driver.
And.. finally, can we be SURE that the driver's response actually mimics the published curves (?) Therein lies the basic question.
Sure I have the driver and the graph is accurate. I have measured the driver in some horns too. It's a pretty sweet little driver but pretty limited in bandwith like most cones. The around 2.5k dip is not going anywhere
http://www.diyaudio.com/forums/multi-way/287154-cone-midrange-horn-101-a-8.html#post4631736
http://www.diyaudio.com/forums/multi-way/287154-cone-midrange-horn-101-a-8.html#post4631736
Last edited:
This is close to the direction I was going before I started down the phase plug path.
I did measure the driver on a wide baffle somewhere here and it was close to the manufacturer chart.
On my 2nd prototype I tried gapping per Bruce Edgar and the impact wasn't very significant. More like a tweak once you have settled on a design. Something I haven't done yet.
I still have one driver with attached dust cap. But measurements have been of the driver without it unless noted.
I tried different throat openings and larger throats yield smoother response at the expense of not reaching as high. But I always had a plate thickness (15 or 18mm) between the horn start and the driver, so always an increased throat length. My next step was going in the direction of a mouth the same size as the driver (like JBL Screen Array 3-way) and eliminate the throat thickness altogether. Thinking I would have to settle with a xo not as high as 2kHz as I intended.
Then I started down this phase plug path.
That's the great thing about it, it doesn't matter if you see it or not, the physics still apply.
I didn't mean to tell you're wrong, I was just mentioning there are some more influences and advantages/disadvantages to it and a link to explain why and how.Brief Response for Now
The horn profile (area expansion) starts at [St] which is divided equally amongst the annuli openings in the phase plug that are adjacent to the driver diaphragm. This means the closer annulus is to the driver center the wider it should be. On the exit side, each annulus should be emitting an equal volume velocity. Making front cavity depth much less than 2*[Xmax] risks driver damage. Designing a phase plug, particularly for a cone/dust-cap assembly is a particularly rigorous effort. Most of the designs I see are too short to meet the necessary transformation requirements. Once we get a coherent wave front at plug exit, then a round to rectangular transformation needs to be addressed. This can be done within the confines of the phase plug as well, but requires CNC machining of a metal phase plug. At that point the BMS driver becomes a cost effective alternative, but the transition is still required. Recommend for a five inch driver that two annuli be used. WHG
OK, I'm biting the bullet. Let's hope I get to the other end with intact teeth
The horn profile (area expansion) starts at [St] which is divided equally amongst the annuli openings in the phase plug that are adjacent to the driver diaphragm. This means the closer annulus is to the driver center the wider it should be. On the exit side, each annulus should be emitting an equal volume velocity. Making front cavity depth much less than 2*[Xmax] risks driver damage. Designing a phase plug, particularly for a cone/dust-cap assembly is a particularly rigorous effort. Most of the designs I see are too short to meet the necessary transformation requirements. Once we get a coherent wave front at plug exit, then a round to rectangular transformation needs to be addressed. This can be done within the confines of the phase plug as well, but requires CNC machining of a metal phase plug. At that point the BMS driver becomes a cost effective alternative, but the transition is still required. Recommend for a five inch driver that two annuli be used. WHG
Last edited:
Force Majuere
To verify what has been simulated, requires that you measure it.
The only measurement you can make is via a microphone that gives you values that are an analog of force per unit area. To measure Power Response of a compression driver requires use of a PWT. You can also calculate power response of the system by summing a properly collected set of microphone position-measurements.
Not sure this really answers your underlying questions.
WHG
Anyone? Pleeeease?
To verify what has been simulated, requires that you measure it.
The only measurement you can make is via a microphone that gives you values that are an analog of force per unit area. To measure Power Response of a compression driver requires use of a PWT. You can also calculate power response of the system by summing a properly collected set of microphone position-measurements.
Not sure this really answers your underlying questions.
WHG
Anyone? Pleeeease?
Are you asking re' simulation or measurement?
Often I find myself look at acoustical power rather than pressure when I want to see the predicted frequency response because that is all HornResp will give me for a horn with more than one segment. Power response is the total output power integrated over all angles. For a constant directivity horn, that power is nominally evenly divided over the coverage angles, but by definition down 6 db at the edges. Because of that, for CD horns, acoustic power is far better than nothing. I imagine for non CD horns its also useful. But you really do want to know the directivity pattern of the horn. You can get that from HR via the tools/directivity menu for single segment horns, (except apparently not if you model the rear chamber as sealed rather than throat ported.)
Why would the phase plug create turbulence? Why would its length make this difference?
Hornresp will assume with a zero Vtc that you have an ideal horn attached. The phase plug should aim to be nothing more than the beginning of the ideal horn collected from a less than ideal source shape.
As mentioned the power is proportional to the pressures so I'm not sure you have the right question, or that I understand your point.
In my mind, the air moving from the center of the diaphragm (right behind the center of the phase plug) around the phase plug and into the throat must be moving non-linearly given the shape of the obstruction (phase plug). The longer the plase plug, the smooother it tails towards the end of the throat. And the shorter it is, the more abrupt the transition from the wider part of the phase plug to the end.
With Hornresp I'm attempting to model how the horn will once built, and hence I iterate the design until I get to the best balance I can, expecting the horn will measure accordingly.
HR cannot do sound pressure level for multi segmented horns. That's a limitation of Hornresp.
I understand at low frequencies acoustical power is more appropriate because those frequencies are less directional. But what about 300 to 2000Hz? In the last simulations I showed the SPL chart with phase plug had an ugly spike caused by the phase plug. On the acoustical pressure chart the spike was less severe. Should I reduce the magnitude of that spike? The raise was so severe the response at 400Hz was out of whack with the response at 1.5kHz, so by raising efficiency at higher frequencies that phase plug might be screwing up the middle of the passband. But if I look at the acoustical pressure chart then the impact is less severe. So I need to know what to look for if I expect to use HR as a design optimization tool.
But you are modelling a single segment horn, at least so far. I'm sure that your final design will have more segments, if only some form of mouth termination. But so long as you still have just a single segment, look at the SPL and directivity, there is more info there. HR will be able to show you the improvements to the power response that a 2nd segment or mouth termination will give but not to any directivity improvements.
re' smoothing the power response - always a good idea. The experts can correct me if I'm wrong but corrections to the power response apply at all angles and thus can be done by EQ. Smoothing the power response will also smooth the pressure response but the shapes of the curves may be different. You can see that comparing power vs pressure graphs from a sim of your single segment horn.
But looking just at the power response hides any directivity deficiencies that a horn has and conical horns do have them. This may not be an issue if you have a single seat that you care about and are going to listen on axis but it will be if your goal is a wide sweet spot.
re' smoothing the power response - always a good idea. The experts can correct me if I'm wrong but corrections to the power response apply at all angles and thus can be done by EQ. Smoothing the power response will also smooth the pressure response but the shapes of the curves may be different. You can see that comparing power vs pressure graphs from a sim of your single segment horn.
But looking just at the power response hides any directivity deficiencies that a horn has and conical horns do have them. This may not be an issue if you have a single seat that you care about and are going to listen on axis but it will be if your goal is a wide sweet spot.
If by non-linearly you mean not in a straight line, it must, because the source isn't otherwise right for the horn.
N.B.
1) Only at low frequencies does phase velocity become infinite and the air actually moves as a unit.
2) Under those conditions, the closer to the diaphragm center the smaller the volume of air that is so moving.
3) For the frequency domain, compression ratios and passage widths you are most likely to use, 1) and 2) should not be at issue.
4) For Phase Plug Passages: Perfect!
WHG
1) Only at low frequencies does phase velocity become infinite and the air actually moves as a unit.
2) Under those conditions, the closer to the diaphragm center the smaller the volume of air that is so moving.
3) For the frequency domain, compression ratios and passage widths you are most likely to use, 1) and 2) should not be at issue.
4) For Phase Plug Passages: Perfect!
WHG
More
Visco-Thermal processes taking place within the compression chamber and narrow phase plug passages will severely attenuate some response spikes shown in HR simulations. HR will not address all of the design issues that you will encounter as you attempt to arrive at optimally performing horn/driver system.
For a cone driver, there will be two principal sets of standing wave modes in the compression chamber. One set across the dust cap and another set across the remaining cone and compliance surface. Besides these, the diaphragm will have its own breakup modes as well. Due to band-width limits, only the first two of each should be of interest. Note suppression of each mode will require a dedicated annulus.
Here is treatise on phase plug design you may find helpful.
https://eprints.soton.ac.uk/348798/1/Jack%20Oclee-Brown%20PhD%20Thesis.pdf
A review of Bob Smith's work follows in the attachments.
Regards,
WHG
Visco-Thermal processes taking place within the compression chamber and narrow phase plug passages will severely attenuate some response spikes shown in HR simulations. HR will not address all of the design issues that you will encounter as you attempt to arrive at optimally performing horn/driver system.
For a cone driver, there will be two principal sets of standing wave modes in the compression chamber. One set across the dust cap and another set across the remaining cone and compliance surface. Besides these, the diaphragm will have its own breakup modes as well. Due to band-width limits, only the first two of each should be of interest. Note suppression of each mode will require a dedicated annulus.
Here is treatise on phase plug design you may find helpful.
https://eprints.soton.ac.uk/348798/1/Jack%20Oclee-Brown%20PhD%20Thesis.pdf
A review of Bob Smith's work follows in the attachments.
Regards,
WHG
Last edited:
Midrange experimentation is paused. I have too many variables and too little time. In fact most development work was paused due to lack of time. And I'm missing an amp to be able to do 4-way active now. So I will work on the midbass section and tweeter, buy the amp, then further explore midrange.
Cone-driven horn evolved some. Then looked into 2" compression-driver horn for mids, then also direct radiators in single, MTM, line array. But all this was theoretical rather than experimental.
I think this year I will be able to move forward more.