We, the horn people, know what the compression drivers could do. The controlled dispersion, low distortions, low mechanical dependency, low magnetic anomalies, high efficiency, higher maximum dynamics and many-many more advance are well known. The term “compression driver” is kind of bogus as we do not rely from compression anymore but it is the only name that we have. The most remarkable part in the compression drivers is how they response to crescendos or to any increase in volume. Compression drivers have insultingly linear increase of intensity with constantly-present extra pool of dynamic headroom. It is very hard to make a good compression driver properly driven and properly used to hit a point where it stop gain volume and begin to compress. Well, where did you see a properly driven compression driver?

This is the subject that I would like to talk about: the compression drivers do not compress and do not dive into the typical “hardness” only if they driven by a “clean signal”. If the signal is “clean” then a compression driver is not stressed by anything – very seldom happen. Unfortunately the industry (and the compression drivers are unlucky, as their industry is pro audio industry) do not investigate or invest into “clean signal” but rather they make the compression drivers less sensitive to the “dirty signals”. All those JBL’s aquaplass cone treatments, increase the compression ratio to 10:1 or 20:1, making barbarian suspensions, use electromagnets (it might be controversial), employing of none-reactive cones and many other “solutions” do mask out the reactance of the drivers to the “dirty signal” but along with it they weaken the sonic characteristics of the drivers, “denobleize” the driver’s Absolute Tone ™ and reduce the driver’s transient capacity.

So, what make a signal “clean” and what would not stress the high vulnerability of a fragile, low power, good compression driver?  I would not look at internal characteristics of the compression drivers: magnets, diaphragms, suspensions and many other and will look only at the external variables. Without pretending for the complicity of the answers here are my runners:

1) Bad electricity (bid subject!!!)
2) Presents in signal path of any direct or indirect phase constructors.
3) Presents, position within the chain and type of capacitors in signal path.
4) Quality of grounds
5) Having <18dB in signal’s dynamic headroom

When those 5 reasons are optimized then a compression driver has opportunity to demonstrate the max capacity of its design. So, in many cases: do not blame the compression drivers and if the fish begin to die in your pound then look up the river for the reasons that poison the pond’s water…

Romy the Cat wrote:

direct or indirect phase constructors.

Markus wrote:

I usually manage to read you quite well, but this   Romy the Cat wrote: direct or indirect phase constructors. has me baffled. What do you mean, please?

I have to admit that the said about this true only in context of “good” compression drivers. A good compression driver is a driver that has a consequential reaction to the “clean” signal. Since the drivers that I am experienced are only Mid frequency compression drivers then I would without any hesitation state that good compression drivers employ ALNICO magnets. The Samarium Cobalt, Neodymium, injection and electromagnets are a different subject and they might have own place but they are not well explored by me in reference to Actual Sound.  The Ferrite or the Ceramic – the most commonly used nowadays materials, are absolutely not usable in context of MF compression drivers. In order any HF/MF Ceramic driver  to be able to  give away anything remotely resembling to  Sound the driver should be driven with enormous amount power that, in own turn, has many other negative  Sonic consequences.

So, in the end, in my warm world I see at high frequency compression driver as an ALNICO-centric driver with amplification that amplifiers THE WHOLE SIGNAL ONCE AS A WHOLE, without any tricks, the way how it is done in single-ended triode amps. Saying it I do not demonstrate a loyalty to a specific topology but rather I just report my past disability to get the desirable sound from compression drivers using other type amplification.

Terry,

I did found that PP amps were good to drive BAD compression drivers (with ceramic magnets). The irony was that the better amps to drive the ceramic magnet were not just the PP amps but the solid state PP amps. Do not ask me why – I do not know why.

Still, with the use of the average better compression drivers (let for a sake of reference point consider the JBL 2440 as a reference “average compression drivers”) there is nothing that could be less “clean” then SET.

BTW, I’m not a big fun of OTL because their very fraudulent harmonics, but I might only presume that if someone would wind very high impedance, still light, voice coil and make the OTLs do not drive DC then OTL might be a good candidate to try defeating the SETs… Still, something is VERY special "lives" sonically in the core of the SET’s transformers…..

Paul S wrote:

Now I am wondering:  If we go to such trouble to have transformers wound for our particular applications, maybe we could just as well be winding our narrow band voice coils and welding them to our/their own output tubes.

Isn’t is what some elescrostas do coupling their transducers directly to the plates of some high voltage transmitting tubes? I do not think that we have this option in compression drivers as compression drivers have a very limited voice-coil mass and (derived form a typical ultra-narrow gap design) the voice-coil-heat dissipation problem.

Sure, the OTL are very “clean” and it is imposable to take it from them. It would be interesting to hear more harrow-bandwidth D-OTL (dedicated) with compression drivers. Vincent Brient  did something in that direction:

http://vincent.brient.free.fr/otl3w_eng.htm

If I were him I would not go for 6C33C and that tube are crazy in OTL mode. I would rather go for 6C19P that have less power and do not “blow DC“ when it dies (as the 6C33C does).  The 6C19P sound muck less interesting in SET mode however

I do not know. I never experimented with OTL idea dedicated to compression drivers, perhaps I should. I always took OTL evaluation in a full range context and it never was successful (people do not know but I had Lamm ML2, Atmasphere and Tenor in the same room)…

Still, as I said there is something interesting lives in the “core” of output transformers. The “core” adds some viscosity and some creaminess to sound, in the same way as the coffee made by excrete of the  Palm Civet adds own “allure” to a fry water….

Hi Paul,
My understanding is that the placette has a unity gain buffer. At full CD ouput levels of 2Vrms into an S2 - it wouldn't quite get that into the S2s 16 ohms - but say it did, you'd be looking at about 109db from a pair...
I'm not saying it would or wouldn't sound good, it's just me being curious given that Melquiades is losing an output stage...
 
The point is you don't really need any gain from the amp when driving such high sensitivity. You might successfully construct an OTL using just a 6C33 with no other stages... anyway, I'm taking things off-topic for this thread so please excuse me.

Best,cv

cv wrote:

Here's a crazy thought... have you every tried driving the S2 directly from the Placette?! I know the 6E5p does some unique things but I wonder if it might be educational. The placette has a 7 ohm Zout right? Don't forget that that doesn't translate into current delivery so best stick to low volume...

I was right before but now I really review my rightness for a new perspective. Yes, the compressions drivers are the subject of clean signal but it what I witness with the new 6-chenals DSET put what I said above in another extreme dimension. I do not know what it was: the singe-stage 6E5P driving, the lower intermods, different power supply but the definition of cleanliness of signal move way out there. It moves in such an extreme dimensions that some “sleeping god” issuers with of my Water Drop tweeter become the open issuers.

What the new amp for my compression drivers suggest that there is another level of dynamic (as opposite of static) signal cleanliness. In context of that moving cleanliness the changes with driver sound express a totally superlative level of majestic delicacy when sound moves across frequency and dynamic range. I think it comes from a new v of signal as there was nothing changed in my loudspeakers.