op.9,

to put your comment I perspective it is necessary to know what surface you had on your Lecleach horns before. If you search the site (use world textured or structured) then you would know that I am a vocal opponent of smooth surfaces on horns. When I see those manufactures sell their glossy and shiny horns I truly smile. Somewhere in this site I pitched enter theory that sound shell be scratched off the surface of the horn instead of sliding and slipping from the walls of the horns. Pretty much an abrasive sand paper is one of the greatest surfaces… In the time when John Hasquin was still in business he taught me to finish the horn surface with textured paint, it is what I do and moderating the dryiness of the pains it is possible to moderate the granularity of painting. Alternately you might use hairy wood of rough plaster surface as Jessie might do with his plaster horns.

Now about the Mr. Gedde’s idea of open cell foam.  It would work but I do not like the solution as foam instead of demanding diffusion does also absorption. I did use foam for experiment in past and I have seen HOW it works. Well, properly if to use very fine and very thin (a few mm) sheet of foam and to cover the horn very accurately then it will created the necessary boundary effect of surface randomanization however I would propose a slightly different angle.

Would it be possible that overly-absorbing characteristics of the Geddes foam redeem the upper range problem of your BMS 4592ND driver? In my past I did corrected the problem with some on my compression drivers by foaming the horns, then I learned that it is preferable to get a better driver. Sure, I have an agenda and I sweepingly do not like the BMS drivers. The improvements that you repost do suggest that the BMS did not work out well before in your horn. So, it might be an opportunity to look at the problem from different perspective.

Sure, it is a matter of personal preference but to my taste anything shiny of glossy should not be used in horn or in audio… unless you are French of course :-)  As far as I concern everything should be absolutely black, completely not glittery, have no reflection of any link and provide as less as possible visual impact. In my past even experimented with large drapes of black nylon covering the speakers. But it is me…

The surface of my fiberglass horns is quite rough. The mould was sprayed from a distance with some water based paint - so it all landed in fine globules. Maybe this is too fine a rough texture though. I like the idea of thick artists oil paint…

One more variable: The meeting of the driver to the horn throat seems very critical. I think I previously had a leak on one side. . . .and the centering of the driver seems to be very critical too. If I place a small piece of absorbent foam right after the driver, these problems disappear, but that is no fix, as then the sound loses dynamic quality. How do you all mate your drivers to the horn? Just a rubber gasket? I now understand why those screw threads were so popular.

I need to do more research, but I'm pretty sure that the foam filling has a different effect from that of the rough horn texture. For example I've now been quite shocked at the sound of live applause. There is now 'wetness' to the texture, and the sound seems to ‘hang’ all around the room.

As to those horrible BMS's, its all I've got at the moment, I’d love to have some other 2'’s to compare one day... But why care if their problems can be successfully controlled by this foam?!

Question, can you tell me what difference open foam has on a 'good' driver in a rough horn?

op.9 wrote:

As to those horrible BMS's… why care if their problems can be successfully controlled by this foam?!

op.9 wrote:

Anyhow, what are the 2" comp drivers that ARE any good and are available comercially? I didnt go with any JBLs because I had a very bad time with the 2427 1" version, could never get any music out of it.

I would be surprised if you get any reasonable sound out of 2427 driver. I do not had it but I true a number of the drivers from the same family. My observation is that those drivers with small cones are very-very poor, unless you use them somewhere above 3kHz, but still the do very bad sound. The 1” Altecs were better, the older 802 or 808 with the 802 cones (not with the 808 cones). Still the Altecs were slow and had very small/ stiff-suspended cone to sound reasonably. The large Altecs were more fun and the 2” JBL were way more sophisticated, particularly with put aluminum cones. You might try the 2440/375 but do not bite into the newest titanium diaphragms. With all cones and pros of 2440 they will have VERY different sound then even a dozen BMS drivers. BMS juts do not go where Sound lives and no foam will help them. The entire reputation of the BMS driver is based upon the hype that a half dozens of idiots- manufacturers blow up abound the BMS driver but in reality it is all very-very third-rate drivers and any single product built around the BMS drivers is a testimony of nothingness.  You might stop by in your local pro audio shop and lease for a weekend a speaker with 2440/375. It will cost you ~$60 over a weekend. At home you would need to clean and to align the driver and give it time to play. It will be very different then JBL one-inchers. There are other options but it is would be an easy way to learn how much you do not get with BMS....

The best driver I have heard is the TAD 4003.

It does not use the secondary resonance of the suspension, as described by Duke LeJeune above for the TAD 4001-4002, to extend the response towards 20 kHz, but has a natural roll off from app. 15 kHz.

Interestingly the sound of the 4003 is much better than the 4001-4002, the density of information and ability to articulate, also in the treble, is on a different level.

The sound of the TAD 4001-4002 might impress for a few notes until you realise that everything they do in the high frequencies have the same monotonous character and is not able to express the "smell" of the instruments very well.

If one looks at the AES paper for the 4001 it can be seen that their response over 8-10 kHz is due to deliberately placed resonances.

The JBL 375 was also made with the same trick, it is generally more coloured than the TAD 4001, but in a pleasing way.

be wrote:

The best driver I have heard is the TAD 4003. It does not use the secondary resonance of the suspension, as described by Duke LeJeune above for the TAD 4001-4002, to extend the response towards 20 kHz, but has a natural roll off from app. 15 kHz. Interestingly the sound of the 4003 is much better than the 4001-4002, the density of information and ability to articulate, also in the treble, is on a different level. The sound of the TAD 4001-4002 might impress for a few notes until you realise that everything they do in the high frequencies have the same monotonous character and is not able to express the "smell" of the instruments very well. If one looks at the AES paper for the 4001 it can be seen that their response over 8-10 kHz is due to deliberately placed resonances. The JBL 375 was also made with the same trick, it is generally more coloured than the TAD 4001, but in a pleasing way. All these drivers where designed by Bart Locanthi, the 4003 was, if I am not mistaken, his last completed driver before he died, apparently he realised his mistake with the others.

horny wrote:

be wrote: It does not use the secondary resonance of the suspension, as described by Duke LeJeune above for the TAD 4001-4002, to extend the response towards 20 kHz, but has a natural roll off from app. 15 kHz.

be wrote:

The best driver I have heard is the TAD 4003. It does not use the secondary resonance of the suspension, as described by Duke LeJeune above for the TAD 4001-4002, to extend the response towards 20 kHz, but has a natural roll off from app. 15 kHz. Interestingly the sound of the 4003 is much better than the 4001-4002, the density of information and ability to articulate, also in the treble, is on a different level. The sound of the TAD 4001-4002 might impress for a few notes until you realise that everything they do in the high frequencies have the same monotonous character and is not able to express the "smell" of the instruments very well. If one looks at the AES paper for the 4001 it can be seen that their response over 8-10 kHz is due to deliberately placed resonances. The JBL 375 was also made with the same trick, it is generally more coloured than the TAD 4001, but in a pleasing way. All these drivers where designed by Bart Locanthi, the 4003 was, if I am not mistaken, his last completed driver before he died, apparently he realised his mistake with the others.

Very interesting comments about TAD 4003, thanks, Be. I had in past the 2001 and 4001, did not like them for their “over oily salad sound” and I dismissed TADs all together. 

The concept of the deliberately placed resonances is complicated and I would never feel that any driver might not suffer from it. They are all hugely resonating, some of them do it in “right way” and some of them now, go figure…  From what I remember when I was trying the 2440 and 375 (had them both) I god a feeling that they probably were the most neutral but at the same time the most capable in term of “discrimination” among the common drivers.

There is another aspect that I did not research in 2000-2001 what I was doing over many compression drivers – the control the drivers by amplification. In 2001 I used Lamm SET that cared everything above 100Hz. I did try to use different taps on the compression drivers that I used. It changed the harmonic envelop of the MF drivers but it also changed a lot my upperbass channel.  The upperbass channel is a bone of a compression MF driver so, it was not realty intuitive what was doing on. Also, at that time I had less evolved listing assessment skills then I have now and I presume that I did not give to some of the tried drivers the chance that they might needed.

What I feel the MF horn drivers need is to be driven by a dedicated SET where the effect of variable loading of the SET might be observed at the MF driver only. I know that MF compression driver have very little reactance back to amp but still, their sound very much might be moderated by SET’s plate loading. Particularly the drivers that use the “deliberately placed resonances” mish be very much comforted by loading the out stage of SET slightly heavier. Under other circumstances I would try some better drivers again, with my progressed understanding how to deal with compression drivers but I am not on the MF drivers market. If I begin to lose satisfaction and would detect some specific annoyances with the MF driver that I am using then it might open opportunities for further try. I would then try your TAD 4003…

op.9 wrote:

I was going to try a selenium 405 with the phenolic diaphragm, but they discontinued it. Everything else seemed a gamble. maybe I missed a trick... cheers, james

I never had that Selenium 405 driver. Generally the phenolic diaphragms are very tricky and might have some excessive softness and some luck of transients. Selenium also had a habit to load gap with ferrofluid – it also might slightly detransientize the driver. BTW, removing the ferrofluid is a big no-no as it completely changes the damping of the driver. Still, if the Selenium do NOT use ceramic magnet then it might be an interesting to try. Purely speculatively I think it might be a suitable driver to play Debussy’s “La Mer” performed by Philadelphia Orchestra, lead by Eugene Ormandy… :-)

BTW, here is another lead:

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&rd=1&item=200195741737

it is a pro version of 2440, it is not better driver but juts more hipped and more expensive. It will be very different then BMS drivers and foam... :-)

Wojtek wrote:

This is probably the reason I can listen to EV Patrician IV in relatively small space and close to the speaker without much of fatigue.  The other factor must be that it's 4- way design with narrow band-pass for each horn. EV T25 2" phenolic compression driver (alnico) is still cheap on Ebay although JamminJersey list it for $400 a pair, seems like it's gaining popularity. I have infamous JBL2470 1" big magnet alnico with phenolic diaphragm and was thinking about fundamental channel design around phenolic driver >500Hz -1000Hz counting on dia ability to play lower into LF without breakdown. The thing is I don't know if there would be any advantage over cone midbass streching up to 1000Hz without buying big expensive 250-300Hz horn .

You are very much might be correct. Electro-voice and Klipsh used phenolic driver quite aggressively and some of their their speakers had some send of softness and non aggressiveness. I certainly not willing to admit that all phenolic are soft but it would be just a general tendency that metal cones are more … “momentary”.  Another example is Cogent that use phenolic – thy too soft and do not go transiently where I would like them to go. In case of Cogent the softened is also come for the electromagnet that makes then overly polite…

I think a cool direction to go might be to use a naked fibers or soaked cloth, I never tried but I would it I were in driver making business, The best metal cones do fine as they damped with soft suspension, with loading and in some cases with proper coating (like JBL uses aquaplass). I also do not feel that the driver’s resonances is necessarily are bad things, the main question to me is what kind of resonances they are. In the end if one know the driver well and the driver has right characteristics then it is always possible to deal with the specific driver imperfections.

Using horn drivers is in a way like dressing an ugly woman. The woman know what she need to do with herself in order to hide what is not necessary to expose and what to highlight as the most tradable commodity. The drivers are the same… They are all upon our will…

Just another paging stuff, or something serious?
http://cgi.ebay.com/ALTEC-VINTAGE-290E-GIANT-VOICE-COMPRESSION-DRIVERS_W0QQitemZ170189928173QQihZ007QQcategoryZ73372QQssPageNameZWDVWQQrdZ1QQcmdZViewItem

Fascinating stuff. I'm certainly going to try to get my hands on some 2440's when I can see them for a nice price.

That 18 Sound ND2060A looks very interesting to me. And would fit in with my silly philosophy of only using currently manufactured components...  

But it got me thinking about compression ratios in general – since 3":2" = 2.25:1 seems very different from 4":2" - 4:1. This must make for a very different type of sound.

Any ‘rules of thumb’ here?

op.9 wrote:

That 18 Sound ND2060A looks very interesting to me. And would fit in with my silly philosophy of only using currently manufactured components...

That ND2060A does look very interesting. I never heard or even saw it but who knows – it might be good. The type of the diaphragm that they use is very promising… if you lay you hard on this driver then let know… BTW, the “silly philosophy of only using currently manufactured components” is as silly as a desire of a musician to play only currently manufactured instilments, or the music of the currently alive composers. I think the next step would be a desire to use components only manufactured in April, only components in triangle shape and the only components that were touched by a person who saw Mother Teresa between 3 and 4 times… :-)   Manufactures might desire to use the currently manufactured components – it understandable but for us – the people who need juts one set of drivers over entire life - I think this peruse of current components is not reasonable....

op.9 wrote:

But it got me thinking about compression ratios in general – since 3":2" = 2.25:1 seems very different from 4":2" - 4:1. This must make for a very different type of sound. Any ‘rules of thumb’ here?

None that I personally have seen. Generally I always consider that higher compression ratio eats sound’s fine points (there are reasons why) but I also have seen OK sounding higher compression ratio drivers. I think that it is not about the absence of any “rules of thumb” but about our disability to evaluate the relationship between compression and sound. The problem that each compression level require own proper implementation of loading. We can change compression but we do not know if for the given compression the loading is done identically properly as it was for the former compression. We do not make drivers (at least I do not do) we just get whatever is there and use them. I do not think that this approach permits methodologically correct evaluate the contribution of compression. With all said above I would vote not for more or less compression but for the appropriate loading of a given compression…