It would also be interesting to compare the distortion of the RAAL to the NrT
Printable View
It would also be interesting to compare the distortion of the RAAL to the NrT
I am surprised by this. Are you sure that the discussion wasn't a generalization towards ribbon tweeters and that someone did actually mention RAAL?
The term "ribbon" tweeter gets thrown around a lot lately but few are actually true ribbons. A true ribbon MUST have a metal/conductive ribbon material as current flows directly across this diaphragm, which is surrounded by magnets. A planar and folded ribbon have a voice coil imprinted/glued on a non-conductive material and current flows through this imprinted winding pattern and thus forces the diaphragm to move. Manufacturing costs of a true ribbon are far greater and in my opinion, performance is also far superior.
There are planar and folded ribbons with higher than average distortion and there are true ribbons with relatively high distortion as well.
However, non-linear distortion is one of the least important characteristics when I chose a transducer. In addition, many of the various THD graphs you see posted are rather meaningless. These graphs are generated by using a steady state sine wave swept across a range of frequencies while measuring even and odd order harmonic multiples of the fundamental.
Here is a good definition of harmonic distortion: Nonlinear distortion characterized by the output of harmonics in a signal waveform that do not correspond with the input signal waveform. Harmonic distortion is caused by an amplifier, transducer, or other element that malfunctions.
In practical terms, send a tweeter a pure 2 kHz signal, and it will produce a measurable level of both even and odd order harmonics based on that 2 kHz signal.
As we know, music is not a single steady-state sine wave (single frequency at a single level). Music consists of many different frequencies at many different levels hitting the transducer all at once. In my experience, these THD measurement graphs have little real-world value and are not representative of driver performance. Yes, I have seen some cheap domes with very low "THD" measurements, however -- hit this same cheap dome with multiple signals at varying levels and IM (intermodulation) distortion can be as high as 30%.
I once had a company claim they can knockoff our 170/200 tweeter for about $4 each (a HUGE cost savings to us). They said their tweeter had lower THD. I measured this tweeter and THD was quite good but it buzzed so loudly that it was useless. The problem is that THD measurements only look at odd and even order multiples of the fundamental. If I hit the tweeter with a 2kHz sine wave, the THD calculation is based on the spectral measurements of 4kHz (2nd order) and 6kHz (3rd order). Some measurements then include the higher order harmonics, 8kHz (4th order) 10kHz (5th order etc.). It does not consider what is going on in between these frequencies, which might be significant especially if there are mechanical issues.
This sample tweeter buzzed horribly at a frequency that was not included in my distortion sweep so the buzz, even though it would have been audible to anyone, was never picked up on any measurements (I believe it was at about 3.2kHz)
The MLSSA system we use allows for a much more useful test which is extremely capable of revealing problems, like a buzzing tweeter or alignment offset. This is often called a "Rub and Buzz" test. I don't want to get into a full discussion about this but it is a test I rely on when comparing transducers. Unfortunately, the MLSSA system does not produce graphs for this, instead it just returns a number (an index) so unfortunately, there is really nothing fancy I can visually display. If you google rub and buzz test you can learn more.
Back to THD -- there has been a longstanding argument as to what level is actually audible. Some professionals claim 5% and higher is considered bad, others 2% and some even 1%. 2nd order harmonic distortion is extremely difficult to identify even at 10% (tube amps have high 2nd order hd) while 3rd order can sound nasty / grating once it is identified. I prefer to use transducers that exhibit 2% or less 3rd order harmonic distortion measured with a 90dB fundamental at 1 meter. Reason for this is that I was once able to "hear" this distortion at this level with a steady state sine wave.
However, it is impossible for any human to hear this same 2% level of distortion with music. Musical instruments themselves generate far greater levels of both even and odd order harmonics. This form of harmonic distortion is something you WANT to hear as it is produced by the instrument and gives the instrument its character. A small amount of harmonic distortion produced by the transducer would be impossible to distinguish from the harmonics produced by the instrument itself.
Anyone who says driver A is “better” than driver B because driver A has 0.2% THD while driver B has 0.8% has very little real-world experience in this field. The same can be said for amplifiers and receivers, which have THD levels so low that it is a specification only used for marketing.
Back to the RAAL 70-20XR... This tweeter far exceeded all of my requirements and anyone claiming to be an expert and choosing not to use this tweeter simply because it might have higher harmonic distortion than a particular dome, is, quite frankly, not someone I would ever purchase a loudspeaker from. Sure, some cheap domes might actually have lower THD measurements while some very expensive domes might have higher THD measurements. It is not something I am the least bit concerned with as, I hope I have clearly explained, it honestly means next to nothing.
I know this won't satisfy all of you and I also know this post will be referenced by others so I asked Aleksandar if it would be OK to post the harmonic distortion measurements of the 70-20XR. I don't believe this information has ever been posted anywhere as it is not available on the RAAL website as this tweeter and its measured performance is only available to OEM's approved by RAAL.
A few important notes regarding this measurement:
SPL is normalized to 95dB at 1 meter (as you know, 95dB is an extremely high listening level)
There is no high pass filter used on the tweeter (no crossover) thus hd levels from about 3kHz and lower will be quite a bit lower and should be disregarded. In addition, measurements above 10kHz should also be disregarded as 2nd and 3rd order multiples of 10kHz are well above the 20kHz range (this is why THD measurements typically stop at 10kHz).
As you can see, at 95dB from 3kHz to 10kHz, 3rd order harmonic distortion (the nasty one) is at 0.3% or lower (actually off the chart) which is completely inaudible. This is impressive.
2nd order harmonic distortion (what gives a tube amp its pleasing sound) also remains low from 3kHz to 10kHz and stays below 0.8%, also inaudible.
Are there dome tweeters with lower harmonic distortion? Yes, – but the levels of the RAAL are so far below the threshold of detection that it is a meaningless comparison and has no bearing on performance.
Here is the website...It doesnt mention the 70-20XR, rather the other RAAL Tweeter the 70-10
http://www.audioexcite.com/?p=2640
I forgot to post this earlier, but the article doesn't say "much higher harmonic distortion than even cheap dome tweeters".
This is what the author wrote:
"The distortion figures isn’t up to the very best dome tweeter’s performance e.g. ScanSpeak D3004/664000 Beryllium tweeter, but for a ribbon tweeter still a respectable performance."
Thanks for sharing!
Interesting article and I would say the measurements look realistic. I too don't see anything mentioned that the RAAL tweeters have much higher distortion than a cheap dome. The ScanSpeak dome being used for the comparison is $510/each. In addition, the RAAL tweeter being evaluated is the 140-15D. Still, when you properly compare the distortion measurements in this summary and discard the information below 3kHz (no high pass filter was used), the 3rd order harmonic distortion between the two tweeters is nearly identical while 2nd order hd for the RAAL is considerably lower.
In a real world environment, with the tweeter being used with an appropriate crossover -- I really can't say which tweeter has lower distortion -- it is just too close to call. I agree with the author of this article, this is indeed impressive for a ribbon tweeter.
Visited the Ascend offices/warehouse today...
Piano Black Sierra Towers being assembled and tested. Whoever these are going to....very nice!
http://changpics.smugmug.com/Other/A...cktowers-L.jpg
I drool when i'm there, when I see all the enclosures and mass drivers, a small tear forms :D
One pair has to be mine if RAAL:) I had my eyes on them for over a year, finally couldn't pass up the piano black promo. I hope Dave can work something out with his domestic cabinet maker for Sierra Horizon in Piano Black, as far as I know there is at least one other person needs matching Horizon who also ordred towers in piano black.