Well here's a topic that might cause some controversy.
I'm sure all of you know about the Zaphaudio's Ribbon Tweeter tests
and some tests done by Mark K. The measurements show that Ribbon Tweeters have significantly higher distortion than their dome counterparts. I'm curious why do ribbon tweeters have more distortion, what is the inherent flaws in the ribbon design that contribute to this higher distortion. I'm very curious and can't find any information on the topic. Also are there any redeeming qualities to ribbons that merit their relatively higher cost.
Thank You,
Angsuman Roy
I'm sure all of you know about the Zaphaudio's Ribbon Tweeter tests
and some tests done by Mark K. The measurements show that Ribbon Tweeters have significantly higher distortion than their dome counterparts. I'm curious why do ribbon tweeters have more distortion, what is the inherent flaws in the ribbon design that contribute to this higher distortion. I'm very curious and can't find any information on the topic. Also are there any redeeming qualities to ribbons that merit their relatively higher cost.
Thank You,
Angsuman Roy
Well, I don't know about the cause of distortion in this case, but I have a few ideas... which I'll keep to myself.
For the benefits of a ribbon- there is generally good horizontal dispersion due to the narrow width, and limited vertical dispersion due to the tweeter's height. You could argue that due to psychoacoustics, the proportion of reflected sound from the walls positively affects the perception of good sound (sound stage, etc), and reflection from the floors and ceiling detracts from it.
In the right room and design, it might be true that even though the ribbon tweeter measures worse (using standard techniques), it actually sounds better due to its polar response. Also, don't forget that people often like a little distortion!
I had some Infinity Qa speakers for a long time (my dad's), and there was definitely something to be said for their EMIT tweeter. I know they didn't have a huge psychoacoustic advantage since I gave them really horrible room placement- but it makes me wonder if there's something else going for them. (disclaimer- next to cutting edge stuff now, it's obvious that they weren't the last word... they were 25 years old)
For the benefits of a ribbon- there is generally good horizontal dispersion due to the narrow width, and limited vertical dispersion due to the tweeter's height. You could argue that due to psychoacoustics, the proportion of reflected sound from the walls positively affects the perception of good sound (sound stage, etc), and reflection from the floors and ceiling detracts from it.
In the right room and design, it might be true that even though the ribbon tweeter measures worse (using standard techniques), it actually sounds better due to its polar response. Also, don't forget that people often like a little distortion!
I had some Infinity Qa speakers for a long time (my dad's), and there was definitely something to be said for their EMIT tweeter. I know they didn't have a huge psychoacoustic advantage since I gave them really horrible room placement- but it makes me wonder if there's something else going for them. (disclaimer- next to cutting edge stuff now, it's obvious that they weren't the last word... they were 25 years old)
Hmm, yea i love the wide dispersion of ribbons and that's their main selling point. Having very wide horizantal dispersion makes the soundstage more 3d. But honestly I'm curious on why or any speculation as to what makes them distory more. On paper they seem like a great idea; a nearly weightless diaphram suspended between very strong magnets. Does the distortion have anything to do with the excursion of the ribbon?
I'll say this, I have had the pleasure of listening to Genesis 501s.
They are the first speaker I would place above my recently departed thors.
There was air and prescence, and detail, but my god it sounded so smooth. I have never heard metal drivers that sound like that.
And it was my first experience with a ribbon driver.
Suprisingly the drivers they use aren't all that expensive 120ish for the ribbons.
Amazing drivers, definately give them a listen.
Nes
They are the first speaker I would place above my recently departed thors.
There was air and prescence, and detail, but my god it sounded so smooth. I have never heard metal drivers that sound like that.
And it was my first experience with a ribbon driver.
Suprisingly the drivers they use aren't all that expensive 120ish for the ribbons.
Amazing drivers, definately give them a listen.
Nes
sorry, I have to play the advocates devil again:
I have not heard any ribbon tweeters sofar that really convinced me,
and I have listened to a lot:
1.) often they sound a bit bodyless, there is no bite
2.) they often sound very "light" and brillliant
3.) the emit are like the phillips I think: Isodynamic, which is physically not linear at all, therefore produces quite a bit of nonlinear distortion
4.) they often sound boring to me, no kick, but that is a matter of taste
5.) I have tried the Jordanow, the Phillips (all of them), Raven, Swans etc, none of them really exiting. Also heard a few of the plasma tweeters, same thing, boring
OK some people like some distortion (like in valves), but I do not like distorting tweeters, just sounds harsh.
My favourite are still dome tweeters (but very few of them). There are some made by Focal that are nice (a lot of them are horrible though, especially the old ones, very agressive) and the Scan speak range is very smooth. My personal favourite is still because of brilliance, livelyness and dispersion (better than ribbons, it is almost hemispheric) http://www.exclusivehifi.com/en/hifi-stereo_tech2.html (suprise -
...)
I have not heard any ribbon tweeters sofar that really convinced me,
and I have listened to a lot:
1.) often they sound a bit bodyless, there is no bite
2.) they often sound very "light" and brillliant
3.) the emit are like the phillips I think: Isodynamic, which is physically not linear at all, therefore produces quite a bit of nonlinear distortion
4.) they often sound boring to me, no kick, but that is a matter of taste
5.) I have tried the Jordanow, the Phillips (all of them), Raven, Swans etc, none of them really exiting. Also heard a few of the plasma tweeters, same thing, boring
OK some people like some distortion (like in valves), but I do not like distorting tweeters, just sounds harsh.
My favourite are still dome tweeters (but very few of them). There are some made by Focal that are nice (a lot of them are horrible though, especially the old ones, very agressive) and the Scan speak range is very smooth. My personal favourite is still because of brilliance, livelyness and dispersion (better than ribbons, it is almost hemispheric) http://www.exclusivehifi.com/en/hifi-stereo_tech2.html (suprise -
...)Well said Kea. I don't really understand the Isodynamic concept its planar magnetic but if i understand there are magnets in the diaphram itself. Apparently planar magnetic can suceed for example all the B&G tweeters which beat domes in certain aspects. But I understand that they generally measure worse than dome tweeters. I'm curious if ribbons really respond "faster" or handle transients better? But mainly i really want to know as to what physically makes the ribbon tweeters distort more.
angsuman said:
Thats easy..
Its the same that it is for all drivers - the mechancal resonance of the tweeter. (though note that other drivers like domes often have problems at the top of their passband where they start to "ring" - not a problem for the ribbon.) Even if you are only reproducing a 5 kHz tone and the resonance is say 1.5 kHz - to produce it at all will in some measure "trigger" the resonance. (..particularly easy for a normal pleated ribbon because it has near none existent mechanical dampening.) This in turn has distortion on the distortion that "bleed's" in across the drivers bandwidth - sometimes at relativly high levels.
And yes, most lower mass drivers - particularly those without much of suspension (like normal pleated ribbons), will usually be faster provided nothing interferes with the the driver's motion (i.e. no acoustic resistance). They will be faster in the "physics" sense in that they can oscilate faster and thus reproduce higher freq.s at a higher output. They will also be faster in the subjective sense in that they respond more quickly to a change from the signal (..shown in the initial rise of an impulse graph).
btw, *most* of the problems kea posed are likely due to an overdamped output because of amplifier pairing with very low output impeadance. Match OR almost double the ribbon's impeadance (i.e. for an 8 ohm ribbon either 8 ohm output or up to 16 ohm output for the amplifier), and you should find that the sound changes rather dramatically in favor of the low mass driver. The tangible air compression (or "kick") IS a product of the drivers mass and its transfer impeadance to the air. The less mass, the less "kick". This is a subjective valuation, BUT generally the lower in freq. you go the more mass your diaprham should have - otherwise it does indeed provide a "lifeless" character (..which incedentally is more accurate to the source but likely less accurate to the event - but this is definitly an art in balancing these two aspects, not a science).
Finally, it seems that everyone is *certain* that the likes or dislikes for ribbons and other low mass drivers are due to harmonic distortion. I find that laughable. Chances are that most people wouldn't even be aware of an increase in 3rd order up to 5% and 2nd order up to 15%. At the lower % levels (below 1%) I would be looking more at 5th order and above to cause any sort of audible problem. So - if you are looking for a problem in most drivers - I wouldn't be to concerned with driver THD unless the higher order (5th+) started moving much past .1%. This doesn't mean that you shouldn't try to minimize THD, but rather that you should be careful in that a reduction in THD may well lead to an increase in distortion that we either cannot measure yet OR can measure, but so far do not understand its importance.
Hmm, interesting response so mainly mechanical resonances add the distortions in ribbon tweeters. While a lot of distortion might be barely audible or not audible at all, the point is that on paper dome tweeters are performing better in terms of distortion. So if i get this straight, the main redeeming qualities of ribbon tweeters is the transient response, the horizantal dispersion, and extended frequency response. What about linear distortion in ribbons? I ask all these questions because while i love planar magnetic speakers such as magnepans and some designs incorporating ribbons, I plan on building a new set of speakers this summer and have WAY too many ideas floating around in my head.
kea said:sorry, I have to play the advocates devil again:
I have not heard any ribbon tweeters sofar that really convinced me,
and I have listened to a lot:
1.) often they sound a bit bodyless, there is no bite
2.) they often sound very "light" and brillliant
3.) the emit are like the phillips I think: Isodynamic, which is physically not linear at all, therefore produces quite a bit of nonlinear distortion
4.) they often sound boring to me, no kick, but that is a matter of taste
5.) I have tried the Jordanow, the Phillips (all of them), Raven, Swans etc, none of them really exiting. Also heard a few of the plasma tweeters, same thing, boring
OK some people like some distortion (like in valves), but I do not like distorting tweeters, just sounds harsh.
(suprise -
I mostly agree and have had a few expensive ribbons.
The Heil Air Motion Transformer, a bit different, has much more life and 'kick' in my experiance.
angsuman said:
"linear distortion".. thats a something of peeve with me (exactly what should be described as linear or not linear with respect to transducers). For that look to the CSD of the driver in question.. most of the so-called non-linear distortion is not represented here because at most the graph is down to 36db and often "non-linear" distortion isn't reaching that high (i.e. -20 db = 10%, -40 db = 1%). Also look at the native sample (i.e. the impulse response) for the rise-time character - but be carefull, sometimes the scaling for graphs is VERY different.
BTW, there are a few other reasons why a ribbon could well be better than another driver for the same passband (.. most of it relating to the use of a transformer). Additionally, there are gains to be had when specifically using a PASSIVE crossover with a ribbon as opposed to an active crossover (unlike most drivers).
When used properly - a true ribbon is simply better sounding (and it is not a "coloration" that helps it to sound better).
I don't know if this helps you at all.. but if I had the money this is the tweeter I'd purchase (at Solen.ca):
http://www.raalribbon.com/products_flatfoil_140-15.htm
Yea the very flat impedence of ribbons does help with crossover design. So basically it comes to down to the way things are measured? and what point of view they can be tilted to?
Dang. the raal's look amazing a ribbon on steroids. I like the companies honesty in showing they have nothing to hide with their measurements.
Dang. the raal's look amazing a ribbon on steroids. I like the companies honesty in showing they have nothing to hide with their measurements.
Tenson said:
MUCH better..
They are very similar in bandwidth, so there are Five major (known) differences between the two drvers:
1. Mass and ribbon construction.. Search for the subjective impressions people give compairing a normal pleated ribbon to the fountek Neo versions - its not kind to the Fountek. The RAAL is pure aluminum.. a little higher in mass, but it doesn't have the pleats so the total mass is actually less then even a normal pleated ribbon of this size.
2. That dispersion waveguide on the fountek (and most ribbon drivers) is essentially a "micky-mouse" solution - that often is an audible problem (though minor).
3. The foam guides represent some pretty inovative thinking to increase vertical dispersion while allowing the extended low freq. response of a larger ribbon and neither increasing cavity resonances nor decreasing horizontal dispersion. Still - its a solution you could easily implement for the fountek with likely similar results.
4. The eff. of the Fountek is better and this usually provides better sound overall. (..a "plus" then for the Fountek.)
5. The CSD isn't even close. (..the decay of the Fountek goes beyond 1.5 milliseconds at level below 1.5 kHz vs. the RAAL which is down 15db in the same range before it even appears - and even then its within a quarter of a milisecond.)
Really, I've never seen a ribbonn as good as the RAAL before. The Raven R3 or the larger Alains may be better in some respects - but both will suffer far worse vertical directivity, and I seriously doubt their spectral decay will be better. On the other hand they can prob. go a little lower in freq..