TY

So it's time to build new mid-tops boxes, to fill the gaps in my current TX cabs, where the lo-mid rolls off above 300Hz. I have some questions:

- Does anyone have info about wave propogation in driver cones?

- Does anyone know of phase plug theory that talks about effective Mms reduction (as opposed to the usual stuff about path lengths and directivity)?

- Does anyone find repeatable, measurable differences between their mid-horn upper rolloffs and what is predicted by standard formulas and programs?

Not quite answers to your questions but a few things I mentioned at the meet you might want to look at....

The horn I mentioned, definitely worth a look: www.eighteensound.com/index.aspx ... ct&pid=269

Although your compression drivers will work down to 500Hz, unless you go for a very large horn they won't properly support that frequency. I think you should be aiming for a crossover point more like 800Hz. The 18Sound XT1464 will probably go a bit lower than that, as will the larger EV horns but you're trying to keep the cab smaller.

I think it's worth looking more at the D&B Q Series boxes. They use two 10" drivers and manage to run them up to 1.3kHz. I'm assuming you're going to be looking at some slightly meatier low mid drivers, the D&B boxes are B&C neo loaded. I think it might be worth looking at the PD 10MH25 as a fairly cost effective high performance driver. For a higher end driver, the 18Sound 10NMB420 looks good.

While I'm impressed at the grasp you have on the design of bandpass enclosures I personally find that they never seem to integrate into a system nicely, especially at higher frequencies.

It's a shame you're too far away to just pop round and chat/play because I love designing cabs, I just don't have the time or space to do much myself.

chaudio wrote: Not quite answers to your questions but a few things I mentioned at the meet you might want to look at....

The horn I mentioned, definitely worth a look: www.eighteensound.com/index.aspx ... ct&pid=269

Looks fairly similar to the TI2000 horns I'm using. They're CD, I think. There's a lot of vertical spread below 1K in the plots, which probably means some unloading of the driver.

chaudio wrote: Although your compression drivers will work down to 500Hz, unless you go for a very large horn they won't properly support that frequency. I think you should be aiming for a crossover point more like 800Hz. The 18Sound XT1464 will probably go a bit lower than that, as will the larger EV horns but you're trying to keep the cab smaller.

Yes, I'm very concerned about unloading the comps. But in the spirit of compromise, I think I'll be OK. Consider that the BMS mid has a 0.8mm excursion, made possible by the fact it does not have to play above 6K. Also, this being a small system, I don't expect to run it at full power. The provisional TY designs add a 1-2 inch baffle all the way around the horn. All things considered, I hope to get away with something like 650-700Hz.

chaudio wrote: I think it's worth looking more at the D&B Q Series boxes. They use two 10" drivers and manage to run them up to 1.3kHz.

I'd be interested in seeing plans - though the Q tops do sound a little thin IMO (not hugely). I guess they had to make compromises too.

chaudio wrote: I'm assuming you're going to be looking at some slightly meatier low mid drivers, the D&B boxes are B&C neo loaded. I think it might be worth looking at the PD 10MH25 as a fairly cost effective high performance driver. For a higher end driver, the 18Sound 10NMB420 looks good.

PD 10MH25: Mms 37g, EBP 360Hz
10NMB420: Mms 31.5g, EBP 200Hz
Delta 10: Mms 32g, EBP 200Hz

So the 10NMB420 has similar specs to the Delta 10s (but probably handles more power). The PD is a different proposition - I just ran a quick model of my proposed design and I think its higher mass may be problematic.

chaudio wrote: While I'm impressed at the grasp you have on the design of bandpass enclosures I personally find that they never seem to integrate into a system nicely, especially at higher frequencies.

Then you won't like what I plan to do for the midbass in the TY :shock: Yup, that's right... I'm looking at a midrange bandpass, running (hopefully) from 140Hz to 650Hz at about 106dB sensitivity. That's not quite horn levels, but the TY will be much smaller.

Usually with bandpass cabs, the driver's EBP is an effective upper limit on frequency range. That's because in a bandpass, the varying gain of the resonators must be counterbalanced perfectly by the loading on the driver. This only works when the driver presents an essentially resistive source impedance. Above EBP, the driver's mass gets in the way - at some frequencies it reacts with the cab to give a boost and at others it adds to give a cut. It's hard to "tweak" these ripples away.

But I've got to thinking that the phase plugs we see in mid horn throats might be doing some "magic" that appears to reduce the Mms, giving control up to maybe 1 or 2 octaves higher. This is why so many practical horns seem to out-perform the predictions of standard formulas. If I can harness this effect, I can build a bandpass with the above specs.

So I think I'm looking at driver with phase plug, set up to act as the A chamber. If it can "magically" cut Mms from 30g to say 15g, then it can work as a bandpass A chamber, (2.5l tuned to 500Hz). Then I cascade that into a C chamber (7l, 260Hz) and now I have flattish 105dB response from 650 down to 220. Finally, I port the rear (B port, 10l, 100Hz) to lift the bottom end a bit, getting to 150Hz cutoff. I can fit two of these into a 40l box which is only a little bigger than the current TX. I can reproduce similar results in Hornresp and my own software, as long as I bodge the Mms down by half.

I know using a bandpass for midrange will be distasteful to many, but I bet you're curious to know what it sounds like. As am I.

My own thoughts on band pass, is they were always ok for generating a low sound bass immediately around the cab. In a small club they were ok. (also where no long throw capacity is needed.) depending on who set them up. The long wave lengths radiate out of the ports with ease. Which is more than could be said for the high end of the bass, ( i guess this is why they could sound very low, having no higher bass to compete with).
I have always thought that the purist mid would come from a cone you are looking at, with the only acceptable interference being a phase plug. I had a mid MT502 MT602 and MT1200 that had reductions in front of the speakers and internally, turned the corners at 90 degrees. These were/are powerful cabs, but not as natural sounding as the MX1200, which has the speaker facing directly out. Trying to imagine mid frequencies in a cab where there are intrusions in the path, baring in mind, the frequencies being considered are all the fundamental voice frequencies, I don't think it would sound very clear.
Also, if above these frequencies, there is a horn for the higher mid, the two won't be compatible, as there will be inconsistency in the throw.
That's what I think anyway. I'm also happy to be proved wrong in the name of progress.

The sound quality issue is a concern, but at 650Hz crossover, surely the band above 650Hz is more critical? Anyway, compromises have to be made. Direct-driver would make my cabs bigger and still only get to about 101dB sensitivity, which isn't enough to keep up with the rest of the system.

Tony, what are your thoughts on the ability of phase plugs to raise the driver cutoff frequency? Have you used them in any context other than at the throat of a horn?

Theoretically. a phase plug should help with extending the hf in any driver. My understanding has always been that a plug properly fitted, stops one side of the speaker seeing the other side, and stops cancelation. Cancelation happens when the wave length becomes shorter than the piston diameter of the speaker. They react with each other within the cone dimension, and cancel at certain frequencies. Wave lengths longer than this are not affected and radiate in a normal way.
It should also be noted, that when the wave lengths become shorter than the cone diameter, they become directional, with no natural dispersion. When designing monitors, I have always tried to cross over into the horn before this point so that dispersion is coherent with the horn, If you don't do this, the sound from a floor monitor can come to you in two parts instead of one.
One other point, The human voice starts at around 100 hz, a soprano just about reaches 1k. The rest is harmonics.
I have only ever used a plug for mid in a horn. In the late '70's I made a large plug to put in front of 15" speaker, but it didn't make a scrap of difference, due to the wave lengths. I didn't know then what I know now.

Just a quick note...

Take a quick look at Outlines Point Source cabinets. They use 12" drivers in what they call a reflex/bandpass hybrid for their low-mid section but probably only up to about 400Hz.

I took a look at the outline site, but can't find enough detail. Maybe if I could read the patents it would say something more.

Tony, if I understand you you are referring to sound coming from a point on the edge of a cone, going sideways along the front of the cone and canceling with the sound coming from the opposite edge - which would start happening when the half-wave is the same as the cone diameter. Now it seems to me that this kind of cancellation would affect sideways radiation (if there were no horn flare), making the driver tend to beam. If there is a horn flare, it will force the sound forward anyway, so it seems like partial beaming isn't a problem. However, a horn gets its gain from constraining the path of the sound waves. So if there's no constraining to be done becuase the driver is already beaming, there's no gain, and you'd get the driver's free air sensitivity. A phase plug would force the sound against the walls of the horn, and permit additional horn gain.

In this whole topic, I think there are likely to be multiple ways of describing the same physical process. That means we don't all have to describe it the same way to be in agreement. All I need to figure out is whether it will work for me the way I hope it will. Maybe I should just build the boxes and be damned :lol:

I think I've figured it out.

Most phase plugs project from a ring about 1/2 to 2/3 of the diameter of the driver - this gives the shortest paths behind the plug and hence best phase coherence, and balances the cone loading nicely. But not all. The Martin W8 low-mid horn and the Funktion 1 horns all project from a ring near the outer edge. I can't comment on how F1 get so high in spite of expected inferior coherence, maybe they exploit wave propagation in the cone material. But that's *not* the mechanism I'm looking at. I notice the W8 mids run up to 800Hz (ie similar upper limit to the proposed TY mid) and at this frequency phase coherence won't be a problem, since the effective path length differences are about 1/8 wavelength.

So we can approximate the cone as though it were a single ring-shaped source at 1/2 to 2/3 the overall diameter. If we make the plug wider than that, the sound path must expand radially before it can pass the plug. This expansion rate corresponds to a horn cutoff of maybe 300-400Hz, and will add maybe 3dB of gain before the sound even escapes the plug.

This gain serves to "magnify" the driver in the usual way, accounting for an apparent Mms reduction. Of course, there are lots of ways we could add bits of horn into bandpasses to get more gain. But normally we've already coupled the driver to a cavity, and it has already developed an upper knee before we can even get the sound into a horn flare. In this case, however, the gain manifests between the driver and the first resonator (A chamber) - so that resonator "thinks" its working with a driver that has impossibly low Mms, and alignments based on that seem to work.

This explains the anomaly I mentioned a while back about the bass drivers in my S15 cabs acting as though their Mms is 70g lower than it really is. No-one can make a high power 15" driver with such a low Mms. But as the sound expands radially in the restricted front chamber (see the plans on the plans page) it gets 3dB gain from 150Hz upwards - making the driver appear lighter to the rest of the cabinet, and letting the whole 100-200Hz range come out nice and flat (without this effect, I'd have got up to about 120Hz before rolling off).

I won't call these devices "phase plugs" because actually their phase coherence is poor compared to a properly designed phase plug and so one would not want to run them past 1K or thereabouts. Perhaps the term "radial coupler" is better. They are really just a flat disk with similar diameter as the driver cone, emitting at the perimeter. They could be used in place of a phase plug if you're not going over 1K (as seen in W8). But I've already (unwittingly) used one in a bandpass to get 50% higher upper end, and soon will use one in the TY.

Actually I don't think the above is right. Having done some more modelling and background reading, I now believe the phenomenon I was looking for is cone flexture.

I can get typical 12 inch horns to model up to an octave higher cutoff by adding a simple cone flexture model, and the adjusted results are closer to reality I believe. If anyone reading this wants to try the adjustment on their horn model, let me know! In the TY models (TY is a bandpass, not a horn) I can get better control in the tricky 300-650Hz range with this adjustment.

I'm looking at a design that goes 250-750 at about 108dB, and has a shelf at about 105dB down to 120Hz (good for overlapping with subs) using the same Delta 10 drivers (2 in each box). I'll probably go for a wider dispersion horn for the HF and a trapezoidal cabinet too. Now I just have to figure out how to fit everything in (it's a very "busy" design).