you seem to ignore that music is composed of many simultaneous frequencies playing through non perfect amps and speakers , take a simple example of two very strong tones dialed up near compression of a tweeter. say one at 8KHz and another at 10Khz they are combined and subtracted in an non-perfect motor which produce combination of series of tones which may indeed be attenuated in one smooth driver and amplified in a extended range (metal cone) with 18KHz peaks. now consider another non-linearity at resonance* which is purely mechanical and 'mixes down' or is subtracted by other terms of the fundamentals harmonics. this is basic signal theory stuff , frequency up conversion and down conversion in radios /transmitters? *'break-up' is a form of mechanical resonance due to non-linearity and/ lack of damping in speaker cones / materials right.
Last edited:
or why the 1st harmonic is creating hash/ congestion in band with a 10dB rise while the other with no breakup or -10dB down still sounds OK or why does one driver with similar motors and different cones can be easily 20 dB differences of inband hash. one Hi-Fi tweeter fatigues and one doesnt
Last edited:
Yes and those tones are a product of the drivers gross non-linearity, just like like the single tones are. The mechanism is the same. Input signal, driver motor creates distortion and the cone/dome rings, amplifying them if the distortion products are of the same frequency as the resonance.
Single tone non linear distortion products and IMD products, will both excite the resonance and be amplified if they land at the correct frequency but the kicker here is that it doesn't matter because we cannot hear them!
Where the real problem with hard domes lie is with IMD products but not quite of the same nature. Lets take a dome with a 25kHz resonance.
Lets feed that dome 8.33kHz and get that 25kHz resonance excited good and proper from the third harmonic. This doesn't matter because we can't hear it, but it's there ringing away. Now lets feed the dome a 8.66kHz tone too and again via the third harmonic produce a tone at 26kHz. Again we cannot hear this and it's more than likely going to be amplified a little depending on the Q of the 25kHz resonance. So now we've got the tweeter producing 25kHz pretty strongly (as distortion products go) and 26kHz less intensely. Now these tones are going to combine to produce something audible that will fold down into the audible range.
That hasn't been my experience if the drivers are correctly used. I don't find metal tweeters fatiguing either.
no how about my 1st example E.g much worst case > puts a 1st order harmonic consisting F0+f1 at the resonance peak which now effectively boosted by the peaking gain, can down mix readily with other inband signals. This down mixing is all acoustic and highly non-linear / chaotic. when I say in-band I'm referring to amps input signal bandwidth.
consider Fo, Fi in-band at -3dB max power with gain at their sum!
gain is calculated compared to a soft cone that has lower BW and minimal break-up. edit cone and dome are interchangable
which you seem to find solace in saying before "only true Be dome and no junky drivers used here"
consider Fo, Fi in-band at -3dB max power with gain at their sum!
gain is calculated compared to a soft cone that has lower BW and minimal break-up. edit cone and dome are interchangable
whishy washy qualifiers and subjective conclusion
which you seem to find solace in saying before "only true Be dome and no junky drivers used here"
Last edited:
Of course you can mix a fundamental with a harmonic that is being amplified by the resonance, I assume this is what you're referring to?
What I was referring to here was that I've never experienced this terrible problem you attribute to metal domes so it clearly isn't that much of an issue. The same could be said for the entire market otherwise metal domes would be never used.
Obviously metal dome resonances are measurable and the increase in harmonic output measurable too, along with IMD products being easily predictable based on how the driver behaves. But the big question here is is it really audible? Or does it fall below the threshold of audibility. Clearly this question depends heavily on the tweeter used. The higher up the resonance and the more well damped the resonance the less effect it's going to have. The same is true for how linear the motor is, if it's harmonic output is 20dB lower than another device then this will reduce any issues too.
Of course we're not going to discuss junky drivers that are inherently faulty, either due to poor design or misuse, because they do not belong in HiFi. The same being true of false Be domes. The whole point of Be is to push the resonance higher up in frequency to make it less of an issue, if the Be dome isn't actually Be then there clearly is no point in using it vs something else.
Last edited by a moderator:
I've never did see why metal is so great on any speaker driver.
ignoring my early adventures, i'm reluctant to go back into that mine field. Maybe you can list all the advantages and convince me, who knows.
I know why they (audiophiles) went away from poly woofer cones... "boring"
BTW intersting some of the newer pro compression drivers have shtcanned metals and are using mylar. The sound reinforcement guys love em.
all the guys doing tweeter testing w/ simple sweeps @ 1 Watt somehow make a great leap of faith that their higher order harmonics will be equal or some such wording to IMDs.( might be at a few points.) its just too hard to test it right and then you might wreck some parts if you don't take care. I found Zaph data to be useless to plain wrong when selecting drivers, 'the catalog readers' eat it up I reckon.
ignoring my early adventures, i'm reluctant to go back into that mine field. Maybe you can list all the advantages and convince me, who knows.
I know why they (audiophiles) went away from poly woofer cones... "boring"
BTW intersting some of the newer pro compression drivers have shtcanned metals and are using mylar. The sound reinforcement guys love em.
yes two tones produce a harmonic ( the strongest one ) in the break-up zone, which when mixed* (acoustically) with a third tone folds back in-band. these products when done in the real world E..g musically create "hash" in band. Now imagine all this near driver compession levels. *assumes the breakup zone is treated as a non-linear/chaotic element.
all the guys doing tweeter testing w/ simple sweeps @ 1 Watt somehow make a great leap of faith that their higher order harmonics will be equal or some such wording to IMDs.( might be at a few points.) its just too hard to test it right and then you might wreck some parts if you don't take care. I found Zaph data to be useless to plain wrong when selecting drivers, 'the catalog readers' eat it up I reckon.
Last edited:
If used appropriately and are correctly designed they give you complete freedom from resonances/breakup within their pass band. Most soft cones suffer from a cone edge/surround resonance that causes a dip in the frequency response. This is usually between 500-2000Hz, which is where you don't want it to be. It is also usually accompanied by a peak in distortion too which I'd rather avoid.
Metal coned drivers also usually show lower distortion throughout their usable passband, at least the SEAS drivers do. Just compare measurements of the Nextel and Magnesium SEAS Excel driver, the mag cones always come out better.
The trick is using them correctly, which isn't always easy. For example I've just finished a design with the W15CY001, it's breakup is a tad over 8kHz, but I cross it at 1.8kHz. For midrange/woofers etc I'd say a good rule of thumb would be cross over at 1/4 the frequency of the breakup, or lower. Of course this is with 4th order acoustic filters.
Well metal will 'sound', as it were, if used incorrectly, and most designs do let some of the metal character through when using xovers that are too high - at least in commercial designs. Monitor Audio etc tend to cross FAR too high up and let through elevated distortion peaks.
That depends, the breakup itself is linear (although I am sure it's cone dependent), it's just a region of higher mechanical efficiency and when equalised flat ceases to be just that and ends up in line with the rest of the cone. You can equalise the linear aspect of the resonance flat, via a notch filter, but you cannot stop the distortion products from exciting it.
Well Zaph's data has correlated well to my own in terms of measuring, it's how you choose to interpret it and then how you choose to use said driver that's important.
Running a test simply at 1 watt can be a fairly good representation for how a driver is going to perform, especially if it's sensitive, which is why it's more useful in Zaph's measurements to measure at a fixed SPL. Either way 1 watt in most tweeters is around 90-95dB, this is loud and actually way above the volume that I listen at, so in a sense, if a tweeter is linear enough with that 1 watt, it will be okay for 99.9% of my listening.
Of course measuring higher than that is always a good idea to really test how capable your tweeter is. Most of the time the higher orders will remain extremely low anyway and it's only the 2nd and 3rd orders that rise. If you've got 4th and 5th shooting up then you know you need a better tweeter or a different design!
I have only now read this thread, so forgive me if these comments are a bit late. I spoke to a respected local speaker designer who build a pair of standmount speakers with Raal ribbons, and then built some floorstanding units with the Scan Be tweeters. I asked him why he did not use the Raal ribbons in the bigger speakers. His answer was simply that the Raal and the Scan Be sounded very much the same, both offer unparalleled high frequencies. He chose the Scan for the bigger speaker because the dispersion pattern was better suited to that application, but he rated both equally highly. I must admit that his comments surprised me a lot (I always thought of Raal as the best of the best), but his speaker designs speak for themselves (in a manner of speaking), so I accepted and respect his opinion. Just some food for thought.
DenonC,
I'm not really surprised. As I have posted before, the big benefit of Be tweeters is the time domain. The super-sharp impulse and waterfall response devoid of ringing. This is also the benefit of the best ribbons and AMT tweeters.
Your designer sounds very smart, and the dispersion is the next thing to consider.
Motor quality can really affect dynamic range and the frequency response plots, but I would expect the very best tweeters, regardless of form and material, to go towards the same point of neutrality and transparency. It's a really good thing for us we have so many choices.
Zaph and others have argued that ribbons tend to have a lot of distortion products in their bottom end, but given how well liked Raal products are, I doubt it's meaningful, and I would trade a little distortion for tight dispersion in my humble abode any day of the week. (Just a personal preference, no flames please!).
Thanks for that!
Erik
I'm not really surprised. As I have posted before, the big benefit of Be tweeters is the time domain. The super-sharp impulse and waterfall response devoid of ringing. This is also the benefit of the best ribbons and AMT tweeters.
Your designer sounds very smart, and the dispersion is the next thing to consider.
Motor quality can really affect dynamic range and the frequency response plots, but I would expect the very best tweeters, regardless of form and material, to go towards the same point of neutrality and transparency. It's a really good thing for us we have so many choices.
Zaph and others have argued that ribbons tend to have a lot of distortion products in their bottom end, but given how well liked Raal products are, I doubt it's meaningful, and I would trade a little distortion for tight dispersion in my humble abode any day of the week. (Just a personal preference, no flames please!).
Thanks for that!
Erik
Very nice Face!
I really think those tweeters are underappreciated in the home (vs. car) application. I wish more tweeter makers would offer great drivers with small face plates. It kind of sucks when your tweeter has the same effective baffle size as your 5" mid-woofer.
Post a pic!
Erik
Or a move to combine the Be tweeter with a waveguide like this in order to control the directivity.
Hi Erik
According to Raal a lot of that distorsion is due to break-up and bell modes in the ribbons themselves. That is why their (Raal's) ribbons look a lot different from normal ribbons (not pleated), and that is why they sound so much 'cleaner' than other ribbons. Pity Zaph never got to test one of the Raals.
Deon
Others have measured them and there are distortion plots scattered around the internet. The RAALs are quite nice and do measure reasonably well but they are subject to the usual issues that ribbons have - cannot be crossover over particularly low if you want to get the most out of them.
This is the biggest crux of ribbons imo as they tend to be rather large, so the usual necessity for a high xover point is at odds with the physically large designs. The RAAL 140-15D really requires a 3kHz xover to perform at its best, which is very high considering it's size.
Exactly and give it more useful output down low to allow for a low xover to help combat the high C2C distance.
This does bother me quite tremendously. Although I am a large advocate of waveguide designs, if you're going to use a standard tweeter why go with a large faceplate? The usual 104mm faceplate seems quite antiquated these days with FEA (to optimise air flow and geometries on the small scale) and high strength neo. Compact motors should be all the rage with smaller face plates. Not that one needs to go to extreme lengths of compactness all the time, such as the D3004/604000 down to 62mm, but at least shrink things by a cm or two.
This AMT is basically distortion and resonance free down to 1.2kHz according to HobbyHifi...
Should be possible to cross at 1.7kHz...
https://www.lautsprechershop.de/pdf/audaphon/audaphon_amt_1i_data.pdf
Hmm, I have money burning in pocket....
Should be possible to cross at 1.7kHz...
https://www.lautsprechershop.de/pdf/audaphon/audaphon_amt_1i_data.pdf
Hmm, I have money burning in pocket....
Looks identical to the Mundorf AMT25CM, wonder what the arrangement is? Down to the funky speaker connectors. They use holes suitable for banana jacks.
If it is the same tweeter, it is technically and audibly faultless. See separate thread which Raal helped me with here:
http://www.diyaudio.com/forums/multi-way/284527-mundorf-amt-ss-7-compression-distortion.html
Speakerdoctor recommended them as well, and uses a smaller model in a version of his "Intimates."
Be warned, most of my audiophile friends don't like them. They are too neutral for the current speaker trends being followed. No notch at 2.4kHz, no uptilt in the last octave. No resonances to call attention to themselves or obscure detail. Crossed over right they sound just like Gravesen describes the diamond Jantzen tweeters. They just completely disappear and become impossible to localize. They sound too mellow until you go and hear a live horn, trumpet or guitar.
Best,
Erik
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.