There are many threads on my site where I covered the subjects of Tweeter for my Macondo and particularly the moment of integration the tweeters with Vitavox S2 compression driver. Since a few months ago I altered my idea to use the EV T-350 tweeter as the “HF contra-resonator” and decided to use a tweeter that actually embraces the S2’s idiosyncrasy I have been looking for different tweeters inspirations.

After the new unsuccessful experiments I have yesterday first sign of a “might be” right direction. Bud Purvine of O-Netics in Everett, Washington sent me to experiment his modified version of Linaeum tweeter. They are very strange linear voice coil tweeter loaded in some kind of soft walls, plastic cylindrical horns. From any rational perspective it is very strange tweeter. Furthermore Bud climes that the tweeter sound like crap but he has know-how how applying his Pattern Methodology he could make this tweeter to sound very different. The Linaeum tweeters that I got from him were already “processed” by his inventions and I spent a couple days to listen them. I put two tweeters together, making them 12R and drove it at 12.5kHz crossover point.

The sound that I heard was very familiar to me – it was the crappy sound of SET driving a low sensitively load. I have observing this effect for years (in my own and in others systems) driving different low sensitively speakers with low powers SET amplifiers. If at MF and LF it juts do not sound good or clipping then at HF something even worst happen. Any SET loaded into a low sensitively tweeter sound limp, flaccid and feeble. It sounds insipid. It feels like a live subject in a state of extreme decomposition. It is syntactic. It is extremely unpleasant and it always screams to me about own ugliness… (Now you know how 90% of HF SET out there sounds to me considering that people drive with them 90dB sensitive speakers). In fact I would like to tell you something else that no one told you. A SET driving low sensitively tweeter destroys the tweeter. Do not ask me why, I do not know why but I have seen it again and again: a tweeter that sounds perfectly fine with high power amp, after being driven for a few weeks by a low poser SET, then sound worse when it switched back to the high power amps. Do not expect from me explanations of these phenomena… To me, it is a Bermuda Triangle of SET-tweeter integration…

Anyhow, recognizing that crappy sound that I got from Linaeum and Melquiades I have to admit that I did detected some promising things: sort of softness but the different kind of softness that my phenolic diaphragms did in EV T350. Then live-saver came… I put form my storage my Lamm M1.1; it is 110W, Class A that did so well on 72dB sensitive Celestion SL600. When I connected the 110W Lamm to drive the Linaeum tweeters and properly equalized them then I was very surprise how Linaeum sounded. It was insultingly clean and elegant with no typical “tweeters chic”. The most surprising that it was a superbly match for Vitavox S2. I usually put my ear next to the exit of S2’s horn and position it that S2 and a tweeter has the same output. In the best scenarios I sound not distinct the sound of the drivers… and it was a one the best matches that I had so far. I measured the 88 db sensitive Linaeum and frankly it did not measure well. Also the HF sweep sound on this tweeter horrible – with many strange resonances and “wowy” afternoons. However it sounded in musical material unbelievably interesting!

OK, where it leads me? Defiantly am not wiling to use 100W Lamms to drive the tweeters (!), It would be possible to put a dozen of so of the Linaeum tweeters in series making the combined impedance of 60R-70R, organize them in a line array, and drive it with OTL – I still mildly contemplate it… But while I’m contemplating it one word buzzes my awareness: the ribbons. The Linaeum tweeters are not ribbons but the have the “ribbonfish” waveguides and sound almost like ribbons. I know how ribbons sound, and I never was a big fun of them, but people hardly use then above 12khz, not to mention to combine them with S2 driver (that do all necessary transient magic at MF – the thighs that ribbons lack at upper MF)…

So I wonder is any sensitively ribbons are available. There was a number of horn loaded ribbons: Decca did them, Scandinavians and Bulgarians did them but I do not see reasons to load ribbons into horn as I am looking at the above 12kHz range where horn are worthless and create more problems then solutions. So, why juts do not put bigger damn mangers and make ribbons more sensitive? So, I am looking for alnico or neodymium powered ribbons tweeters, or ribbons assembly tweeters with sensitively of around 100db. If I pile up more of ribbons then I might get up to 110dB… I do not ask how they sound – I let you know how they sound. Any contestants?

Ronny,

it does sound interesting, thanks.

As they said they were the people who supplied the Red Rose Music speakers with Ribbons? Still, while they called a statement” "Ribbons are excellent but suffer from bad efficiency" as an warranted myth they at the same rime make the Ribbons for Red Rose monitors that had sensitively of 90dB. Do you know if they sell drivers by pair or if they do custom size, sensitivity by piecemeal?  I did hear the Red Rose monitors in numerous occasions. The did have the ribbony-glyceriny sound atop It is hard to say anything defiantly as the Red Rose, similar to those Scandinavian people use the ribbons with very low crossover point in two ways systems. My crossover will be 4 or more octaves higher and therefore I will be in the very different scale. Perhaps if I do not need to have a lot of LF excursion then they could make the gaps narrower that would already yield more sensitivity… there are many people who clime “true ribbon” there are not a lot who make 103dB sensitive “true ribbon”.

Hey, a person who help to find for me 103dB sensitive ribbon will get a gift…. :-)

I emailed to Bo and he said that they can't provide individual units. There is another interesting drier out there. It is RBN180 18” toll ribbon with reported sensitivity of 110dB

http://www.alconsaudio.com/site/technology.html

I never heard it and have no knowledge about it’s level of implementation. It is slightly annoys me the fact that the RBN180 is being made by a company that targets pro audio market. The pro audio products has notoriously horrible sound… I would not even mention the claim 210W RMS power handling at 3K croosover – I wonder how better the driver might be is it was designed to handle 5-10W…..Still I look forward to try this driver…. 

Thanks Gregm.

It is interesting that I completely forgot the TAD PT-R9. I was looking at it 5-6 years back and spoke with AudioKinesis ‘Duke, who reportedly knows them well. My reservations that time were the very same as now: with 97.5 dB/W sensitively I would need 3 of them per channel and with a price of $1000 per pair it is … kind of heavy.., also I generally am not a big TAD fan, sine I tried the TAD 2001 and TAD 4001 and know what kind syntactic garbage people mend when they admire the TAD drivers.  However, nowadays I have a different demands then 5-6 years back and who knows: perhaps one single PT-R9 will be able to be driven by a SET and perhaps the PT-R9 does not sound as the rest TAD drivers. After all it is not a compression driver with beryllium cones…. Anyhow, I never heard them. Perhaps it would be good opportunity to learn how they sound. I will borrow this Sunday a pair of PT-R9 from a local guy and will try. It should be an educational….Still. it would be nice to try an array of 3-4 of them…

Yes, Paul, I know this company. It is always slightly annoyed me that it is Chinese company as Chinese manufactures very rarely, if ever, do any audio products with extreme or demanding performing characteristics. They rather do inexpensive mass-marker audio products. Also this quality control is very questionable, even worst then the Russians. Well, I understand that it all just prejudges of mine and perhaps soon or later they will make some mark in high end production. After all, the Arum Cantus might be one of them, - I do not.

The problem is that although I head, as anyone else in audio, countless among of different ribbons never tried to make up any association between the sound I heard and the type/brand of the given ribbon. In truth, up to very recently they all were the same to me and I did not like them very equally.  I’m not quite sure if my recent curiosity to ribbons is fully justified. The Linaeum driver, driven by 100W Lamm PP made me to express interest in ribbons and it is ironic because the Linaeums are not ribbons….

I'm not familiar with ribbons, never used them in my system. I will be trying the TAD PT-R9. Tomorrow and will see where it would lead me. From one perspective to use them at 12K and up would make them to not have the problem that I usually associate with ribbon’s upper MF but from other perspective there are a few things that I do not like about the HF ribbons… even before I heard then in my installation:

1) Most of the HF ribbons (the industry call with moronic word as super-tweeters) go way above 20kHz and I do not like or need it. I need a tweeter that picks from 12.5kHz and dies at 18kHz maximum, 15kHz -16kHz will do as well, but it should do the 12.5kHz-18kHz… properly, or at least affinity to my S2 driver.

2) Practically all ribbons that I have seen have this idiotic little baffle that I very much do not like. A HF transducer should be isolated in air and any baffles are deadly horrible for HF sound. HF gets reflected by this flat vertical baffles and crate HF noise. Tweeters, partially the narrow size tweeters should not have any baffles!

Another interesting opportunity was expressed by O-Netics’ Bud Purvine who propose a crazy but quite reasonable hypnotizes why my SET can’t driver low sensitively tweeters. It will take some time to validate his proposal….

Paul S wrote:

With the Wright amps the G2Si did not seem to live up to their efficiency specs using my little cap and coil crossover.  I can't offer good descriptions of their sound with the ML2s until I get my TT going, but there is obviously a MUCH better match now.  MUCH MUCH better, in fact.  I am very encouraged that the hardness has come up for the correct "air", which used to be somewhat hiss-y but is now properly, forcefully ragged when appropriate, and it is melded in with the sound, not just added on top.  Also, before I blame the Arum Cantus for anything I think I will try other caps in that crossover; right now I think I have a baby Auricap for that one.  All in all, I think the Arum Cantus still has plenty of "potential" for ML2, so IMO it is that "good", anyway, meaning at least it is not obviously limited with ML2.

I do not know what kind sasses you have with your Arum Cantus when you switched from Wright amps to older Lamm ML2 but I assure you that it has nothing to do with Arum Cantus. I have seen many speakers that the “vintage ML2” converted from a pile of crap into something very acceptable. The older Lamm were able to do it. I have no animosity to Wright amps, but I herd them many time and they are really kindergarten amps in comparing to older ML2. In fact the older Lamms were did not have so much worsening of sound when they drove low sensitivity tweeter as my Milq has. Lamm runs global feedback in ML2 but Melquiades has no feedback and therefore much more sensitive to load. Anyhow, if you careful chose a tube for input stage and slightly upload the output stage (use 4R tap against the high impedance load) then the vintage ML2 will give as good HF as it possible to get from a full range SET. It will not be able to play aggressively of rough if music called upon but all together it might be very much a reference how a HF should be structured by a full-range amplifier.

I tried the driver and generally I do like it’s sound a lot. It more sophisticated then Linaeum, it is more noble tonally and alike the Linaeum is seamlessly melts with Vitavox S2. In fact I am developing recently a theorem according to which a compression tweeters is… might be very much a fundamentally faulty way for HF reproduction.  Do not make a conclusion that I’m idiot too fast – I have a lot of reasons to think this way, not only because my practical conclusions but rather the conceptual reasons (it will be another thread)

The TAD PT-R9, despite it's superb sound, has 3 problems:

1) It has a lot of ambiance noise, that makes it “always there” instated to be there only when it called upon. The Linaeum practically free form this problem, however they are treated by Bud Purvine  with his EnABL Patterns: 

EnABL Patterns (pdf)

http://www.positive-feedback.com/Issue21/standingwaves.htm.

I did not experimented with virgin Linaeum tweeters but Bud claims that the original Linaeum are way different (as far as noise concern) then his treated Linaeum. So, presumably if the EnABL Patterns were applied to TAD PT-R9 driver then it might loose its “ambient noise” as well? Or perhaps that large stupid buffle should be removed at all?  It was a reference to my post above:

http://www.goodsoundclub.com/TreeItem.aspx?postID=2989#2989

2) TAD PT-R9 has some very very very vey minor “common denominator sweetness” that is identically across all recordings. It is auditable what I drive tem with Milq and it is not auditable when I drive them with 100W amp… Sounds familiar?

3) The TAD PT-R9 driver specification said that it has 97.5dB at 1W/1M sensitivity. Well, when I say that those marketing idiots should be shot in their stupid heads then do you think I was kidding? OK, tell me: what made those cretins to advertise that the drivers have 97.5dB?  Below are 2K and up sweeps driven by Super Milq HF channel’s at the same volumes. The first one is 109dB sensitive Vitavox S2 driver. The Second is reportedly 97.5dB sensitive TAD PT-R9 that looks more like 91dB sensitive… Sucks!!! The last sweep is the TAD PT-R9 driven by a single full-range Melquiades at full 15W power.

I think no further comments are necessary….

Paul S wrote:

OK, Romy, that EnABL stuff got my attention! I can't recall all of the bad ideas borne of good research, but this EnABL thing sounds a little fishy, and certainly it sounds too good to be true, as existing in a simple and really useful form.  Can it be just so? Of course we've known for some time that we could alter the sound of various drivers by "doping" them, putting light foam stuff on them, etc., basically by messing with their mass, but also altering their skin effects and radiating patterns with this and that. But how can this EnABL thing/stuff work on anything that's supposed to go high?  I mean, how, for example, would it work on a ribbon or a 'stat panel, which driving surfaces certainly exhibit boundry effects, too?  Surely the theoretical benefits have to be offset by the mere presence of something more on the radiating surface? Obviously, I don't get it. BTW, I am surprised that TAD, which siupposedly caters to "pros", would give bad info on efficiency, but I will be more surprised if you get a stock unit to work for the top of the top end on your otherwise "perfect" hi-fi speakers.  Few "pro" speakers are drop-ins for hi-fi, in my experience.  I understand that TAD drivers are used in an expensive version of a JBL Hartsfield redux, but then, I remember the Hartsfield as fairly colored to begin with.

Yes, I also was VERY surprised that TAD had way off the mark sensitivity. From a different perspective PT-R9 is not really a really a professional driver but rather a Japanese toy for the freaks (perhaps like me). They made it as a kinky marketing statement and I did not see they sold them a lot. I think for the first 5 years they had some kind of reticular price of $2800 per driver. I think Michelle Framer go 4 of those drives and he suggested that when he placed them under the legs of his TT’s stand then the “soundstage because wider”. I presume that the fact that the quoted sensitively is way off (8db!!!!) then it might be an evidence that the driver never was seriously used. It is no surprise the Pioneer drove the form for the driver down very aggressively since then…. Still, despite my bitching about the PT-R9 I have to admit that it sounds very good.

BTW, I am very seriously contemplating of getting that Arum Cantus G1 of yours, despite of my prejudices. Is your G2Si driver have the quoted sensitively? Do you know if the Arum Cantus use the neodymium manger of the ceramic one?

About the EnABL patterns, I relay do not know. Do you remember when Einstein hammered a hors shoe atop of his door and reporters attacked him blaming him that how such a materialistic and rational person as Einstein was could believe in a such bogusness. Einstein replied that he hors shoe bring luck to everyone, even to the people who do not believe in it. I think I feel that same with the EnABL patterns. I do not really understand it and do not look forward to understand it….. However the EnABL treated Linaeums sound “different” (be advised the I did not hear the untreated Linaeums).

The EnABL thing sounds a little fishy? A vacuum tube or a transistor idea is also sound fishy to me, so many other things and particularly in audio. I do not know yet if the EnABL works. The result with Linaeums was very pleasant and I am sure that if I get your Arum Cantus then I will try to EnABLed one of the driver and to see how it behaves. I think that the baffled ribbon tweeter should be the most benefited by the EnABL methods.

Chris, I did ask myself the same question: perhaps my given PTR9 is demagnetized. I called the last night to the owner of the tweeters and he said that he uses them for 7 years and they driver should be fine. He also expressed a lot of doubts if the drivers are 88dB and he said that it more likely like 93dB. Actually when he told me that I did not believe him because I thought that it would not be possible. I tried the last light only one driver, later on I will try another – it is highly unlikely that they both are damaged and I think that 91dB is what they in fact are.

Unfortunately Nelson Pass’ comment are irrelevant:

Nelson Pass wrote:

Figure 6 shows the same curve, but isolates the ribbon tweeter, which is driven by a current source and actively crossed over at 12 dB/octave at 10 KHz.  You will notice that it doesn’t actually go away until down around 6 KHz, a typical example of why you don’t want to take textbook crossover values too seriously with real world drivers.

So, since he uses an active crossover then he has dedicated power amps to drives the tweeters. It is unknown what gain those amps have and it is most likely that the amp has 8dB more then the 99dB sensitive Lowther DX4s. Not to mention that in his graph the PTR-9 hit in the pick over 100dB… obviously it was amplified….

cv wrote:

I think at the frequencies involved you would NOT get the summing of sensitivity you are after as the radiation patterns would not overlap too much.

Well, with 2 drivers is will be questionable as connected in series they would make 16R and will sink the amp’ output minimizing the summation (unless another transformer ratio was used). With 4 drivers it would be possible to maintain the same impedance and regardless the frequencies involved the summation will take place. However, I agree that the amplitude of the summation will not be as predicable as at lower frequencies. The acoustic pressure will rise but at the same time, since the driver are organized into it is a line-array they will act as a small low-pass filter rolling off the HF. On the other hands, the line-arrayed 4 driver will shape the valve up front and more focus… effectively creating more pressure at the listening position and narrowing the latitudinal spread of HF. Of course to predict if exactly would be kind of difficult. The line-arrays might be complex thing…. :-)

Screw the Canadian ribbons with 92dB sensitively. Chris, now I am very pissed on you. I am keep wondering in dark and after all I am learning that you already use the 8 Aurum Cantus G1s drivers. Are you hiding form me or what? Anyhow tell me more about them and please send the picture of your custom baffle. Are they really 102db? – after the TADs I’m getting into sensitively paranoia. The 8 Aurum Cantus with removed baffles of exactly what I am considering now if I find a good sores to buy them. With the current leading price of $340 it is unlikely but I hope to find a better deal. Since the Aurum Cantus G1 should have very narrow horizontal dissipation I was considering putting them into something like a horizontal ark array….

Allo

Well, my baffles are for a vertical array. A friend of mine built a "clone" of the Nola Grand reference and got a very good deal on the G1 ribbons... will send you a PM.

Don't be pissed about me keeping you in the dark regarding the ribbons; your comments elsewhere on the Nolas and ribbons in general had me loathe to make any recommendations. I think we may have very different tastes in terms of treble reproduction; I like my ionofanes, for example, but more to the point, I have not not spent much time yet trying to integrate a tweeter with the S2s, which is what this all about. And the ESG array is for another project which is shelved for the near future...

I wonder if for 10kHz and up, you want to use narrower ribbons than that on the G1 for best results.

The guy who runs RAAL has made a custom, multi-ribbon dipole arc of adjustable curvature looking exactly like what you want (save perhaps for diplar operation). There was a picture of it here:

http://www.diyaudio.com/forums/showthread.php?s=28fd849ad2d3e2f2e65669cd3232c54e&threadid=62391&perpage=10&pagenumber=168

but I think it's gone. See post 1673 for example. I have been playing with FEMM lately looking at something similar but without the curvature.

Anyway, I will try and get some measurements of my G1s for you to see what the sensitivity is like relative to some references.

cv wrote:

Don't be pissed about me keeping you in the dark regarding the ribbons; your comments elsewhere on the Nolas and ribbons in general had me loathe to make any recommendations. I think we may have very different tastes in terms of treble reproduction; I like my ionofanes, for example…

cv wrote:

… but more to the point, I have not not spent much time yet trying to integrate a tweeter with the S2s, which is what this all about.

cv wrote:

The guy who runs RAAL has made a custom, multi-ribbon dipole arc of adjustable curvature looking exactly like what you want (save perhaps for diplar operation). There was a picture of it here:

cv wrote:

Anyway, I will try and get some measurements of my G1s for you to see what the sensitivity is like relative to some references.

Romy the Cat wrote:

I can’t see the pictures. What I was thinking to connect the ribbon’s array in series and drive them with one custom better Bad’s transformer. Or perhaps if it were I would say .3R ribbons then 5-6 drivers would make a couple Ohms and this impedance would be possible to drive by a dedicated HF channel with a special OPT, eliminating that, perhaps compromised, China-made transformer in the each tweeter… What do you think?  It might be an interesting project….

Gentlemen,

I suppose it is time to defend my sterling reputation here. The EnABL process does in deed work across the full frequency band. Works on all driver types and permits the very qualities that Romy seems to appreciate in the treated Linneum tweeters.

The process controls the boundary layer, absolutely. Since the boundary layer of any emitter completely describes what can and cannot be emitted into the adjacent air, or water, to control it's spurious ringing, without deconstructing the coherent transient information in both micro and macro signals, is rather important. Since the boundary layer is amenable to lateral control mechanisms, being a carrier of lateral compression waves as they express into the adjacent medium over a period of time, a low mass diffraction grating at both ends of the membrane will perform a multitude of important functions.
The simulation of an infinite length between terminus edges. In this the energy density in the boundary layer, through the flat wave diffraction grating, is far greater than on either side of the grating. This forces energy that has passed through and is now at the edge of the membrane to exit the edge. This remaining energy will exit in phase and time coherence, with the energy already expanding into the adjacent air. Theoretically there are tiny diffraction artifacts from the grating blocks, back into the boundary layer, but they are extremely small in amplitude and have no chance to form a standing wave.
The elimination of null emitter bands within the boundary layer allows phase and time coherent emission into the surrounding air as the wave front crosses the emitter surface.
The amount of mass required to effect this control in a 6.5" mid/woofer is equivalent to writing your name, in script on an 8.5" X 11" piece of paper with a ball point ink pen. This relationship holds true regardless of the size of the emitter being processed. So, this is not a bulk mass damping system. It is instead a mass damping, of sorts, of the boundary layer, in the lateral plane only. Since the boundary layer is a very low mass high energy area, a small amount of mass, relative to the total emitter diaphragm mass, is all that is required for complete control of reflected energy that would otherwise form standing waves.

What has always fascinated me about this process is that it works on woofers, well below their natural piston point, so it even controls the emitted energy with wave lengths far greater than the dimensions between rows of the applied pattern.

What it removes is exactly what Romy is complaining of, the uncorrelatable sound that is left over as transient ringing, from the passage of a wave across an emitter surface. It also brings some other benefits but this is the most important one and a full range system controlled in this fashion does not emit ANY spurious information. Nor does it exhibit ANY colorations from the drivers. Nor does it have any "location" perceivable in a stereo spread. And last but not least, there are no limits to the resolution available.

I am aware that this sounds fishy, many others have made the same claims and then failed to perform. I welcome any audiophile, who cares to journey to the west coast, to come by and listen to a full range system treated in this fashion.

You can even poke and prod around, looking for the hidden alternate reality generators.

This is real and it does work.

cv wrote:

Ok, I managed to get my measurement setup running. Behringer ECM8000/PC soundcard kind of thing. I measured the following tweeters: Focal TLR TAD PT-R9 Aurum Cantus G1 Fane 5022 I don't have an absolute level calibrator and I need to take time to try and make the measurements accurately, ensuring the distances are correct in each case. But having said that, if we assume each of these drivers is an 8 ohm load (I left the signal level and gains untouched), then I measured, in relative terms, sensitivities commensurate with the makers figures. In other words, the TLRs (rated at about 95db/W and coming with individual calibrated response plots) were a db or two less sensitive than the PTR9s, which in turn were a bit less sensitive than the G1s. Early indications are that the 102db/W is off for the G1. They don't sound like 102db either. But I think the PTR9 spec would seem to be honest, at least with my units. The Fane has such a crappy humped response it's impossible to draw any conclusions there! Anyway,  this is very rough - will try and find time to take some more careful measurements later in the week, hopefully tomorrow.

Well, Chris, I need to check out my other PT-R9 driver. Perhaps the one that I tried was discharged … I do not think so but I will do it within a day or so.

BTW, currently my completion for more interesting high sensitively ribbon tweeter looks like coming to the end and it is possibly that you will be the winner. I spoke today for a quite long time with the proposed by you RAAL’s Alex and I like the guy quite a lot. He has very valuable from my point of view combination of knowledge and desire to use his expertise in applied manner to benefit the objectives of my project. I have very-very pleasant feeling about him and I very much like how he thinks. We did not decided yet but perhaps we can cook a custom 107dB ribbon driver. If we move forward then you will be declared a winner and that BMW Z4 Roadster will be shipped to you.

Thanks,
Romy the Cat

PS added in a few hours: Chriss, I juts got another PT-R9 from the box. I did not measure it but just A/B the subjective output of the both tweeters and I confirmed that they are identical. I was pretty convinced that it would be the case. The only explanation that I might see is in the efficiency of my room at HF. I have, as you remember, quite a lot of going on in my room that reduces the room response at HF (all that hanging thing, the book and record shelves, the cats t-shorts and so on…) Perhaps it is that eats the affectivity HF in my room. 9dB, Hm, quite a lot! I never consider my room over-dumped but it defiantly never was “live”. I really am not wiling to change anything with the sound of my room. Probably it is better to continue beaching about the tweeters manufactures….