and a pipe (slotted) with compression driver may be my favorite tweeter (within its LF limitations) - the highpass filter needs no contouring and output sufficient for lower dB systems (like 100dB or so sensitivity one watt - 1 meter)
a baffle at the driver's exit plane could help on the low end
a baffle at the driver's exit plane could help on the low end

Forum member Javad's speakers are a very bad exaple for a bad loudspeaker (sorry for the double negation).
I.e they definitely sound nicer than the other speakers that you linked to playing Diana Krall. On all the records I know of her she has a full and "velvety" timbre (I guess I am allowed to use high-end BS terms once in a while) but in this exaple she sounds quite thin. So much for comparisons.
Regards
Charles
I.e they definitely sound nicer than the other speakers that you linked to playing Diana Krall. On all the records I know of her she has a full and "velvety" timbre (I guess I am allowed to use high-end BS terms once in a while) but in this exaple she sounds quite thin. So much for comparisons.
Regards
Charles
An amazing number of misconceptions in this thread.
Displacement and/or deformation does NOT mean distortion. If the displacement and/or deformation is linear then there is no distortion.
A compression driver is a speaker like any other. They look different, but the moving parts are identical. The reason for the "compression part is to couple a larger moving surface to a horn. Horns are a requirement for any Constant Directivity (CD) design and so putting a source on a horn often requires a compression aspect on the driver if a larger diaphragm is to be used - hence a Compression Driver (CDr). Compression is not distortion either as all sound is a compression of the air and its not distorted. Very high compression levels can create very low order distortions, but these are inaudible anyways.
One serious downside to using a compression driver is it is many times more complex to do a system design with them than just a tweeter. And getting it wrong, as was basically the norm in the past, makes for some seriously bad sound quality. But a decent CDr on a good waveguide is unparalleled in performance by any other device.
As to the "ripples" in the pass band of the impedance and/or the SPL, this is because of poor horn design. A well designed waveguide does not have any passband ripples in either the impedance or the SPL. What a proper waveguide does do that is difficult to correct is to drop the efficiency of the device at HFs, i.e -6dB/oct, which must be corrected. But this correction lowers the required displacement at the LFs which is a god-send for small drivers. That is why so many small tweeters use them, and it works, just not as well as a CDr on a properly sized waveguide.
And for higher SPLs there simply is no other option.
So the bottom line is that CDrs are just much more difficult to design right and that has seriously limited there adoption. They also tend to be big if done right and this is also a turn off to a lot of people.
Displacement and/or deformation does NOT mean distortion. If the displacement and/or deformation is linear then there is no distortion.
A compression driver is a speaker like any other. They look different, but the moving parts are identical. The reason for the "compression part is to couple a larger moving surface to a horn. Horns are a requirement for any Constant Directivity (CD) design and so putting a source on a horn often requires a compression aspect on the driver if a larger diaphragm is to be used - hence a Compression Driver (CDr). Compression is not distortion either as all sound is a compression of the air and its not distorted. Very high compression levels can create very low order distortions, but these are inaudible anyways.
One serious downside to using a compression driver is it is many times more complex to do a system design with them than just a tweeter. And getting it wrong, as was basically the norm in the past, makes for some seriously bad sound quality. But a decent CDr on a good waveguide is unparalleled in performance by any other device.
As to the "ripples" in the pass band of the impedance and/or the SPL, this is because of poor horn design. A well designed waveguide does not have any passband ripples in either the impedance or the SPL. What a proper waveguide does do that is difficult to correct is to drop the efficiency of the device at HFs, i.e -6dB/oct, which must be corrected. But this correction lowers the required displacement at the LFs which is a god-send for small drivers. That is why so many small tweeters use them, and it works, just not as well as a CDr on a properly sized waveguide.
And for higher SPLs there simply is no other option.
So the bottom line is that CDrs are just much more difficult to design right and that has seriously limited there adoption. They also tend to be big if done right and this is also a turn off to a lot of people.
Last edited:
"anachronistic dead end" Not So!
"not great compared to more modern horn designs" applies now as well.
Expect to see the reemergence of the acoustic lens in the not to distant future. You do not have to sacrifice driver loading (efficiency) for directivity control. WHG
I had the JBL 2391/2420 and 2392/2440 Acoustic lens and drivers, they were ok but not great compared to more modern horn designs. The 2392 got replaced with 2380A which was a major improvement in clarity and distortion. Same 2440 driver in both cases.
The 2390 is better but very expensive on the vintage market.
http://www.jblpro.com/pub/obsolete/Acoustic_Lens_Family1.pdf
My feeling is that they are an anachronistic dead end. I loved them for a while, but ultimately their lack of detail, significant distortion and coloration made their allure fade for me. JMTCW..
"not great compared to more modern horn designs" applies now as well.
Expect to see the reemergence of the acoustic lens in the not to distant future. You do not have to sacrifice driver loading (efficiency) for directivity control. WHG
Hi Bill
I did my MS Thesis on acoustic lenses, built several and did a model. There is an AES paper on the same. The real problem with them is that they actually lack effectiveness. A high (enough) index of refraction requires a serious step to the plates and this causes a large reflection from the device itself. Hence it is difficult to meet hopeful expectations without serious trade-offs resulting. An acoustic lense suffers the same problems as an optical one and that is reflection at the interface. In optics we have coatings that work well because the wavelengths are so small. In acoustics no such solution is possible.
There is no way that an acoustic lense could ever replicate the directivity control of a waveguide.
I did my MS Thesis on acoustic lenses, built several and did a model. There is an AES paper on the same. The real problem with them is that they actually lack effectiveness. A high (enough) index of refraction requires a serious step to the plates and this causes a large reflection from the device itself. Hence it is difficult to meet hopeful expectations without serious trade-offs resulting. An acoustic lense suffers the same problems as an optical one and that is reflection at the interface. In optics we have coatings that work well because the wavelengths are so small. In acoustics no such solution is possible.
There is no way that an acoustic lense could ever replicate the directivity control of a waveguide.
An amazing number of misconceptions in this thread.Yes that was expected but I still feel the thread gives some good indications of strength and challenges
it is many times more complex to do a system design with them than just a tweeter. Can you elaborate?
What a proper waveguide does do that is difficult to correct is to drop the efficiency of the device at HFs, i.e -6dB/oct, which must be corrected. Will you explain that a little further please?
And for higher SPLs there simply is no other option. That is a plus but so important for most people`s use at home
So the bottom line is that CDrs are just much more difficult to design right and that has seriously limited there adoption. They also tend to be big if done right and this is also a turn off to a lot of people.What are the general rules of how big they have to be?
Last edited:
That is a plus but so important for most people`s use at home
If your room is small and playback music always quietly, dome is better.
The reason why I say this is not because dome sounds better at lower level, but because dome speakers distort quickly, so you do not want to turn the level up. If you have a horn speaker, I bet you want to turn the SPL much higher, and if you can't, you'll be frustrated with horn speakers, and I'm talking about the domestic home audio. I have heard different audio systems in different contexts, and I think I can say there is a clear correlation between distortion level and sweet spot of the playback level.
Yes that was expected but I still feel the thread gives some good indications of strength and challenges
Can you elaborate?
Will you explain that a little further please?
What are the general rules of how big they have to be?
All transducers have a falling velocity above resonance. This means that for a flat radiation resistance they6 will have a falling power response. A piston has a narrowing directivity that exactly matches the falling pressure such that, on axis, the pressure response is flat. But a waveguide is CD and this means that its falling power is spread evenly in its radiation field. Thus its pressure response will fall at 6 dB/oct. This has to be designed into the crossover.
For a direct radiating piston a HP filter is all that is required, simple enough,but for a waveguide one must have a HP filter at the lower end of the pass band and another HP filter at the upper end, and, in general, there will be peaks in the response at resonance which must also be corrected. Thus the crossover for a CDr on a CD device is far more complex than it is for a simple tweeter.
The waveguide will do its magic only down to a frequency such that its mouth is a 1/2 wavelength across. Thus for every octave lower that we want to go it has to double in r diameter, i.e. quadruple in size. To match a woofer at 800 Hz requires a waveguide that is 18-20" across - 400 Hz an its 40" across, get the trend. Now a waveguide that is only 10" across will only control down to about 1600 Hz. That's pretty high for any good two way design. Doable,but not ideal.
If your room is small and playback music always quietly ...
I call that muzak, not music. Its background not serious listening. Serious listening at realistic levels in never quiet.
Here's an adhoc study showing the relative distortions of a Morel CAT 408 tweeter vs a Peerless DCM2535 CDr on a PVR waveguide ( 152i clone ).
Generally, the $42.00 Peerless beats the pants off the Morel ( distortion wise ) except in the 2nd Harmonic above 3500hz > though these are all levels of distortion ( above 1K ) that I doubt I can hear.
The Peerless CDr was HiPassed ( EQed mostly flat by the HP filter ) to facilitate a more direct comparison.
I "get it" that many object to the directional aspects of horns & waveguides ( that's an inarguable/sacred preference IME ).
🙂
Generally, the $42.00 Peerless beats the pants off the Morel ( distortion wise ) except in the 2nd Harmonic above 3500hz > though these are all levels of distortion ( above 1K ) that I doubt I can hear.
The Peerless CDr was HiPassed ( EQed mostly flat by the HP filter ) to facilitate a more direct comparison.
I "get it" that many object to the directional aspects of horns & waveguides ( that's an inarguable/sacred preference IME ).
🙂
Attachments
I "get it" that many object to the directional aspects of horns & waveguides ( that's an inarguable/sacred preference IME ).
"Object to"? Their directional characteristics are one of their most valuable features. And how is it "inarguable"?
Last edited:
If your room is small and playback music always quietly, dome is better.
The reason why I say this is not because dome sounds better at lower level, but because dome speakers distort quickly, so you do not want to turn the level up. If you have a horn speaker, I bet you want to turn the SPL much higher, and if you can't, you'll be frustrated with horn speakers, and I'm talking about the domestic home audio. I have heard different audio systems in different contexts, and I think I can say there is a clear correlation between distortion level and sweet spot of the playback level.
I totally disagree with this. If you like dome tweeters, you should use them in your home system. If you like compression drivers, you will be disappointed using a dome, and should use compression drivers no matter what volume you are listening at.
"Object to"? their directional characteristics are one of their most valuable features. And how is it "inarguable"?
I'm simply saying that others object to their directional properties and I won't bother arguing against their preferences ( though those preferences are in fact contrary to my own ).
🙂
Thread posts can be easily misunderstood - been there, done that.
Virtually all loudspeaker designers agree that CD is desirable. What we don't agree on is what the DI should be. SO, to me, the argument is about the level of the DI, not if it should be flat or not.
Virtually all loudspeaker designers agree that CD is desirable. What we don't agree on is what the DI should be. SO, to me, the argument is about the level of the DI, not if it should be flat or not.
John, I beg to differ.
Many years ago I designed and built a 90x50 Synergy horn that crosses a 4550 at 900Hz to some B&C 8NDL51s. These speakers regularly get driven all night at 1600W/RMS per box and have not failed, nor do they sound harsh at those power levels. I've never seen any dome tweeter in a horn or not that can provide that level of output below 1KHz for any period of time without destruction, let alone doing it as cleanly as the BMS ring radiator.

This one?
I think you were the one that clued me in to Danley using BMS at Sound Physics Labs.
I call that muzak, not music. Its background not serious listening. Serious listening at realistic levels in never quiet.
One of the reasons that I bought Earl's speakers was that I realized I was actually hoarse after listening to them!
Basically we'd been hanging out at RMAF, and he was playing his speakers at what seemed like a modest level. But the next day I was hoarse, and I realized that the Summas were just so darn clean, I didn't even perceive that we were listening at a level that was probably 10dB louder than I'd normally listen at, and I was shouting over the speaker to hold a conversation.
It's really quite a trip when you get the cabinet really inert and you get the distortion really low and you just have a loudspeaker that can belt it out with no drama at all.
I've heard the big $100K flagships from Dynaudio and Focal and Wilson, and every last one of them sound like the tweeter is on the verge of exploding when played loud.
Patrick Bateman said:My measurements indicate that a cheap dome tweeter can outperform a BMS 4552.
I have both at my disposal; I like the BMS 4552.
But if your goal is a low crossover point, a dome tweeter wins because it has more displacement, due to higher xmax. Basically the surface area of the radiator is smaller but the xmax is higher, so the overall displacement of a dome is higher.
There's no free lunch, and a BMS 4552 will smoke a dome tweeter if you need to do 130dB at 2khz.
But if you need to do 110dB at 1khz, the dome wins. The output of the 4552 drops like a rock on the low end of it's bandwidth.
Hi,
Are you able to provide more data to support the words that I've underscored?
I'd like to see the distortion spec ( at 100db ) of a waveguide loaded 25mm or 35mm dome that makes it down to 1000hz ( Hi-Passed to be flat ).
I realize that many ( if not most ) of today's modern small compression driver are truly limited in their lower reaches and aren't comfy at 1000hz.
I find the overlays section within REW excellent for comparisons like these.
🙂
Your comment is interesting. I use a "domesticated" Yorkville Unity U15. It has the BMS4550 "tweeter" I believe. Similarly, many of Danley's Synergy horns use this same, or similar CD for the highs. The BMS4550 is spec'ed to crossover down to 900 Hz or so (I use 1K, Yorkville is 1.5 KHz I think). A point: these are commercial PA speakers that are built to output crazy loud dB levels, far more than is needed at home. Why would performance suffer even down to 1 KHz?
- Status
- Not open for further replies.
- Home
- Loudspeakers
- Multi-Way
- what are the drawbacks of compression drivers?