To look at it from a neutral standpoint, the question of whether or not the HF driver used in modern Dual Concentrics is a "compression driver" or not has two prongs. First, is there a defined (or consensus agreed-upon by practitioners in the field) compression ratio that determines whether or not a driver is a "compression driver?" If not, then any drive-unit loaded with any compression by definition is a "compression driver." If so, move on to the second prong of the analysis: is the tweeter in a modern Tannoy Dual Concentric under sufficient compression to meet the defined/consensus definition of "compression driver."
I don't know the answer to either prong. I'm just throwing out the issues that are relevant and material, because that's more productive than simply asserting what you or I "see."
(Incidentally, one could apply the same analysis the to the modern Uni-Q. KEF's propaganda explicitly talks about the "compression ratio" from their Tangerine phase plug.)
That is your assertion, for which you've provided no supporting evidence. Now, it may be true. But for the reasons I mentioned above, I doubt it. The Tulip is a more complicated device than the Pepperpot, with pieces that need to be assembled. Therefore it's actually likely to be the more expensive phase plug. (Though, of course, I may be wrong. I am not a production engineer. Perhaps you have some inside information, or a greater general knowledge of the industry.)
Now, if you really mean (as opposed to what you actually wrote) that a whole obsolete Dual Concentric with the Pepperpot phase plug would be more expensive to make than a whole modern Dual Concentric with the Tulip phase plug is, then you're probably right. But the reason there has nothing to do with the phase plug. Rather, the reason is simply that AlNiCo is a lot more expensive (right now) than modern magnetic structures.
Whether or not a 50mm vc is better or worse than a 33mm vc depends on the length of the coil, other thermal management techniques in place (forced-air cooling, for instance), and the thickness of the coil. Basically, thermal mass and external cooling measures, not just coil diameter.
Driver makers have learned a few things about thermal management since the 1950s. (Also remember that, unlike AlNiCo, modern magnets don't permanently lose strength when some power's applied to the voice coil.)
But that still leaves open the question of whether any improvement in power compression is material in drivers operating above 1kHz. Keep in mind that both vc's are far larger than the typical dome tweeters, and that both diaphragms are under compression. Also, the above analysis ignores a balancing of compromises that also includes the mass corner of the driver and thus the HF response potential.
Who cares about impedance? We're in 2012! If this were 1955 and we had to worry about midrange coloration from crappy tube amps with high source impedance, OK. Sure, then impedance fluctuations matter. But today we have access to any number of competently-designed amplifiers that have low enough source impedance so as to not vary more than ± 0.25dB into even wildly varying loads.
As for phase tracking, you may or may not be correct. I don't know, though knowing what I do about studies of absolute phase audibility I am not overly concerned.
[edit]Leaving aside the issue of cost, Kindhornman's analysis of the relative merits of the two phase plug designs is spot on.[/edit]
OK. Would you prefer "speaking entirely without the benefit any relevant experience, but at least admitting said utter lack of experience"?
One needn't be rich. One merely needs to be curious. I've seen Tannoy ICT drivers available on my local Craigslist for $25 a pop. I've even bought some of them. They're just fine for background noise (in ceiling speakers in a large foyer or something) or even at home as surround speakers. High fidelity they are not. Dual Concentrics, however, if used well, can be genuinely high-fidelity reproducers.
It's a sad commentary on one's weltanschauung when one thinks my earlier comments were "elitist."
Perhaps. Everything's a compromise in some respect. Using the polepiece as the throat of a tweeter waveguide is certainly much less of a compromise than a 7" woofer combined with a tweeter on a 4" flat flange in terms of midrange polar response, though!
Whether or not you "like" something is simply out of the scope of the discussion.
You made a value judgement when you wrote, "the altec biflex and tannoy [ICT] seem like the best ideas to me. Ive never heard either. *** A CD firing thru a vented pole is a simple botch if you ask me."
I merely pointed out that your value judgement was flawed, and without foundation. There's simply nothing the ICT does better than the Dual, except probably cost less.
That's a rather silly comment, considering the two devices under discussion are from the same brand!
Really, how can anyone reasonably see a "brand loyalty" issue when another comments that one design from company x is better than another design of from that same company x?
I disagree.
To address your latter point first, other drivers can certainly be "concentric" or "coincident." Not "Dual Concentric" unless they're Tannoys, but considering how broad "coaxial" is, the terms that are a subset of "coaxial" are useful. Just as "square" is very a very useful term to refer to some rectangles.
To your "Dual Concentric" point: Tannoy put the term in the public lexicon by coining it and then registering it as a trademark. So they get to tell us what it means. Just as, for instance, Stephen Colbert gets to tell us exactly what "truthiness" means.
Here, for reference, is Tannoy's definition of Dual Concentric: "a coincident point source, where the low frequency cone acts as a seamless, direct extension of the high frequency waveguide, yielding a constant directivity pattern ***" Source. Tannoy has another trademarked term they use for their widebanders with little inductively driven bullets: ICT. To refer to an ICT driver as a Dual Concentric is technically incorrect, and intellectually sloppy. Kind of like when Thiel calls the widebander in their CS2.4 a "coincident" driver, or for that matter when KEF and Tannoy call their respective phase plug designs "waveguides" in their marketing propaganda.
Last edited:
Tannoy stopped using AlNiCo 10 or 12 years before they stopped using pepperpots.
The reason for not using pepperpots anymore was mainly manufacturing cost. They later re-introduced them on some models in the Prestige series at a huge premium.
By what mechanism would AlNiCo lose strength when power is applied?
It can't be heat because AlNiCo is the magnet material least effected by it and the only one which retains magnetism even when red hot.
It is however very susceptible to shock. As long as you don't drop them or hit them with a hammer AlNiCos magnetic strength is more durable than any other.
Pallas,
without getting into your last post which I mostly agree with I will give this little piece of information. All of the compression drivers that I have done actual measurement of the physical phase plug use a ratio of approximately 10 to 1 of the diaphragm to initial opening area of the phase plug. I think this is fairly standard. Also what most people don't understand is the the distance and volume of air being as small as it is between the diaphragm and phasing plug that the air in this small area is controlled by refraction of the waveforms coming off the diaphragm. Otherwise it would seem that the air is simply compressed, this is not the function of this ratio.
without getting into your last post which I mostly agree with I will give this little piece of information. All of the compression drivers that I have done actual measurement of the physical phase plug use a ratio of approximately 10 to 1 of the diaphragm to initial opening area of the phase plug. I think this is fairly standard. Also what most people don't understand is the the distance and volume of air being as small as it is between the diaphragm and phasing plug that the air in this small area is controlled by refraction of the waveforms coming off the diaphragm. Otherwise it would seem that the air is simply compressed, this is not the function of this ratio.
Charles Darwin,
You are very correct about the temperature stability of AlNiCo magnets verses the newest Neodymium magnets. This is one of my greatest concerns when using the highest Mgoe neodymium magnets that are in close contact with a voicecoil. The highest force Neo material have yet to be developed with high temperature qualities. They quickly lose all magnetic energy when heated above their curie point.
You are very correct about the temperature stability of AlNiCo magnets verses the newest Neodymium magnets. This is one of my greatest concerns when using the highest Mgoe neodymium magnets that are in close contact with a voicecoil. The highest force Neo material have yet to be developed with high temperature qualities. They quickly lose all magnetic energy when heated above their curie point.
Do you think perhaps the single point source is also a little overblown? We don't typically hear real music from a single point source anyway. It would make for an excellent blind speaker test wouldn't it. Can you hear from a reasonable listening distance how close to center of the woofer the tweeter is? I would think if that was the end all that every speaker manufacture would be building concentric drivers. What do you think?
My opinion on this is that ultimately the coaxial or point source is theoretically superior. But and this is an engineering but, If the distance from one device to another is less than the crossover frequency then you should not be able to detect this if the crossover slope is high enough and in phase and time aligned. This is the big but though as often the crossover frequency is to high and there is a obvious problem with the distance from center to center of the two or more devices. If you look at most MTM or WMTMW designs this is almost always the case. Not only is the distance between the mid and high to great, but the distance between the two mid-range and bass drivers is such that they have comb filtering effects.
Dave S,
What you are saying is that the magnetic field surrounding a highly charged voicecoil has an effective demagnetization curve, but it has a limited effect on the ultimate Mgoe level over time. While the Neodymium material will if over heated will loose almost 100% of its magnetization. So there is a tradeoff in very high powered devices. If you never reach that temperature with the Neo it would be the preferred material.
What you are saying is that the magnetic field surrounding a highly charged voicecoil has an effective demagnetization curve, but it has a limited effect on the ultimate Mgoe level over time. While the Neodymium material will if over heated will loose almost 100% of its magnetization. So there is a tradeoff in very high powered devices. If you never reach that temperature with the Neo it would be the preferred material.
I built a pair of floorstanding coaxials using B&C 12CXT drivers, it wasn't until I built the passive crossover using the B&C schematic did I realize these are truly exceptional speakers. They are manufactured in Italy, the woofer cone is doped. Compression driver is a 1.7" Mylar, crossed over at 1.9khz. Lots of work appears to have gone into the phase plug. The FR is a bit bumpy on paper but listening to them you'd never know. I first tried them active but prefer the passive crossover using high grade components.
Greets!
True, but I don't view a high CR as a definition requirement. To my way of thinking, if there's a constriction between the driver and the radiating medium to acoustically load it at all, it's a compression driver. This could be something as simple as a driver size short tube unless the driver is XO'd high enough for it to act only as a waveguide [WG].
GM
Gm,
I was only trying to give a description of what we think of as a classic compression driver verses a load high frequency device. The horns that I have produced for both bass and mid-range usage also had a compression as you are calling it. I traditionally used a 25% loading on the drivers and I would call the transition from the driver to the minimum cross-section a compression chamber, but I wouldn't give it a designation of a compression driver in a classic sense. I will tell you that this does change the efficiency but at the same time does change the bandwidth of the device. I have been doing this for 35 years now, and it is just a way to separate a compression driver from a loaded waveguide.
Steven
I was only trying to give a description of what we think of as a classic compression driver verses a load high frequency device. The horns that I have produced for both bass and mid-range usage also had a compression as you are calling it. I traditionally used a 25% loading on the drivers and I would call the transition from the driver to the minimum cross-section a compression chamber, but I wouldn't give it a designation of a compression driver in a classic sense. I will tell you that this does change the efficiency but at the same time does change the bandwidth of the device. I have been doing this for 35 years now, and it is just a way to separate a compression driver from a loaded waveguide.
Steven
It can have an ultimate effect on the magnetic field in limiting it to zero.
I have a quite simple view on that topic: look what is used in PA systems, because if they loose a speaker in the field (e. g. a concert) they have a problem. They don't use AlNiCo, but ferrites and Neodymium. So heat is obviously not a problem (any more), but high demagnetization field is. Price is another problem, and many PA driver manufacturers introduced new high quality ferrite driven speakers, because Neodymium prices went up.
BTW, (some) ferrite magnets are stable under heat (they just loose their strength temporarily), but don't like polar winters. Neodymium magnets are quite the opposite.
Baseballbat
How interesting. I always believed that pepper pot in Tannoy DC was not a series of randomly drilled holes and that the holes were drilled precisely at exact angles to achieve perfect spherical wavefront at the throat. I believed that it was the reason why production was expensive. But hey, don't ask me, I have BA degree in Art History, not a degree in engineering, so it is quite possible that I am wrong.
They are certainly not 'randomly drilled' although precision dropped a bit during the '80s and calling them that is a wee bit disingenuous.
Producing pepperpots required a skilled machinist and time while the tulips use pressed components clipped together by unskilled labourers.
I suppose the reintroduced pepperpots are cnc'd these days.
Charles Darwin,
Perhaps the drilling wasn't exactly random as I called it, though the picture I have seen was not precision, as a machinist I would have no problem saying that. Maybe the newer CNC version would be much better. But the problem with drilling holes that small is that most drills in those sizes have a tendency to turn, not always follow a given path, they twist to put it nicely. But that isn't even my major gripe, a straight drill bit does not in any way expand, it is as I just said a straight non expanding hole. You can add the hole diameters together in any manner and the total area is staying constant. This is not an angle drill they are using or a tapered end mill or any type of taper that I can see. So you have produced a transition that is collecting the area together into perhaps a coherent wavefront but you have caused the first section of the expansion rate to be zero in one of the most important areas of the device. I just fundamentally disagree with this basic premise, I would never consider this a proper design for that simple reason. Radial grooves or even the early tangerine Altec phasing plugs all attempt to follow an exponential expansion rate, though many were actually conic, it is fundamental to a good compression driver. Anything less than that will be an inferior reproducer, no matter who's brand name is attached. No matter what dispersion angle the tulip phasing plug takes it follow that basic premise. And to say that because you could use castings or precision machined parts to produce these parts does not make them inferior in any way to a drilled Pepper-pot design. I would say just the opposite, it was cheaper to not have to produce proper tooling and to drill a series of holes in a steel plug. The precision of the Tulip design can be very high with modern die casting methods and the centering simplified with proper design.
Perhaps the drilling wasn't exactly random as I called it, though the picture I have seen was not precision, as a machinist I would have no problem saying that. Maybe the newer CNC version would be much better. But the problem with drilling holes that small is that most drills in those sizes have a tendency to turn, not always follow a given path, they twist to put it nicely. But that isn't even my major gripe, a straight drill bit does not in any way expand, it is as I just said a straight non expanding hole. You can add the hole diameters together in any manner and the total area is staying constant. This is not an angle drill they are using or a tapered end mill or any type of taper that I can see. So you have produced a transition that is collecting the area together into perhaps a coherent wavefront but you have caused the first section of the expansion rate to be zero in one of the most important areas of the device. I just fundamentally disagree with this basic premise, I would never consider this a proper design for that simple reason. Radial grooves or even the early tangerine Altec phasing plugs all attempt to follow an exponential expansion rate, though many were actually conic, it is fundamental to a good compression driver. Anything less than that will be an inferior reproducer, no matter who's brand name is attached. No matter what dispersion angle the tulip phasing plug takes it follow that basic premise. And to say that because you could use castings or precision machined parts to produce these parts does not make them inferior in any way to a drilled Pepper-pot design. I would say just the opposite, it was cheaper to not have to produce proper tooling and to drill a series of holes in a steel plug. The precision of the Tulip design can be very high with modern die casting methods and the centering simplified with proper design.
Last edited:
Charles,
Without proper data I really couldn't make that claim about cost either way. Not knowing the production volumes and tooling cost that would be a difficult question to answer. Perhaps you have much more information in that regard, but I just couldn't answer that question. My basic point really is that baring cost as a factor the Tulip design which could be changed for different dispersion requirements is the superior design. That was my real point, not price or any other thing. I would just consider the Pepper-pot design as an inferior attempt at this one aspect of the device. Overall I can't make that call.
Without proper data I really couldn't make that claim about cost either way. Not knowing the production volumes and tooling cost that would be a difficult question to answer. Perhaps you have much more information in that regard, but I just couldn't answer that question. My basic point really is that baring cost as a factor the Tulip design which could be changed for different dispersion requirements is the superior design. That was my real point, not price or any other thing. I would just consider the Pepper-pot design as an inferior attempt at this one aspect of the device. Overall I can't make that call.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.