I have to confess I cannot live with silk dome tweeters they do not sound right because of lack of stability to follow the signal.
I cured cheaper types of them applying DIY very thin aluminum foil. Last time on a mylar dome.
https://www.diyaudio.com/community/threads/tuning-tweaking-of-bp-115-a.375464/post-7403173
Expensive tweeters I would sell.
I managed to put thin foil on headphone speakers, too. On the small surface which should work as a piston it brought diaphragm induced distortion down. On piezo paper cones this tweak works also.
I do not like the sound of silk ring radiators although they work as bending wave transducers where stability plays no role.
Maybe I would lose in a blind listening test but that's my experience so far.
I cured cheaper types of them applying DIY very thin aluminum foil. Last time on a mylar dome.
https://www.diyaudio.com/community/threads/tuning-tweaking-of-bp-115-a.375464/post-7403173
Expensive tweeters I would sell.
I managed to put thin foil on headphone speakers, too. On the small surface which should work as a piston it brought diaphragm induced distortion down. On piezo paper cones this tweak works also.
I do not like the sound of silk ring radiators although they work as bending wave transducers where stability plays no role.
Maybe I would lose in a blind listening test but that's my experience so far.
Last edited:
Oh, and I believe that there are very good fullranges which give you all you need. Some measure perfectly and go up to 30khz or more.
In listening tests with original instruments against the 20cm fullrange I once built there was no difference to hear (=small and big triangle and mouth Harmonika blues harp c)
In listening tests with original instruments against the 20cm fullrange I once built there was no difference to hear (=small and big triangle and mouth Harmonika blues harp c)
I get very good results supporting impulse response by putting a damping textile to the baffle front.
Hi, yeah what ever suits best. I don't use a baffle at all so don't have to do anything to it 😉
Like always everything is interacting together as a complete system and it is impossible to get rid all "problems" because our trasducers are physical objects. Best we can do is get rid of audible issues, espeially the worst ones what ever they are. I'm not sure how to perceive edge diffraction, other than perhaps as kind of effortlesness where listening position doesn't matter and it always sounds kind of open. I do not have a way to AB test so I could be wrong, perception is always combination of multiple things.
Like always everything is interacting together as a complete system and it is impossible to get rid all "problems" because our trasducers are physical objects. Best we can do is get rid of audible issues, espeially the worst ones what ever they are. I'm not sure how to perceive edge diffraction, other than perhaps as kind of effortlesness where listening position doesn't matter and it always sounds kind of open. I do not have a way to AB test so I could be wrong, perception is always combination of multiple things.
I have only "studied" two AMT, one from the Adam F5/7 monitor turned into 5" MTM, and the other a midsize 26x50mm from China Taobao (my first DIY) paired with 5" Eve monitor's fiberglass honeycomb using 1st-order-series XO. I wrote then, its projection reminded me of one of NY's last RKO where I saw Barbra Streisand's Yentl. Air motion indeed. Without any shrillness.
HF material signature defines speaker sound character, I have to agree. I can't stand Ti hardness, for example. There's no alternative but to try them one after another.
My all-time favorite tweeter is the Eton 29 HD 2 Magnesium Ceramic tweeter. Holy grail of simultaneous smoothness and detail. Unfortunately I can no longer find it in the States but you may have better luck in Germany.
The SB26CDC is similar in spirit and much cheaper. Might be worth trying.
The Bliesma T34A-4 is also excellent if you want detail and effortlessness. It sounds almost like an AMT in a dome form-factor.
The SB26CDC is similar in spirit and much cheaper. Might be worth trying.
The Bliesma T34A-4 is also excellent if you want detail and effortlessness. It sounds almost like an AMT in a dome form-factor.
-if you linerize (flatten-out) the response as Allen noted, most of the delay goes away. The resonances that don’t represent the more troublesome resonances of the driver (or to usually a lesser extent diffraction).
Note that the waterfall plot is also a single-axis plot.
When people talk about “attack” or “dynamics” at high freq.s it’s usually referencing efficiency along with moving mass. (..in other words applied force.)
If you want a lot of “attack” then look to a driver with the required bandwidth that is very efficient and has more Mms. Small exit larger diaphragm compression drivers epitomize this.
BTW, I’ve not heard a compression driver I would describe as “silky”.
Really the classic Dynaudio tweeters typify that “silky” sound. The better evolution of that sound/design is from Morel’s better tweeters. They tend to provide a more “meaty”/present image owing to their lower freq. performance.
Last edited:
Yes, compression drivers are worth mentioning. They can sound like nothing (as it should be) when properly set up, but that takes much cabinet/room and EQ work.
Edge roundovers - yes they need to be bigger as used usually. I used a 25mm radius one and just recently bought a 34mm router.
It depends on the baffle size and in which frequency area you get your reflections! As I build a wider speaker now woth 12" lf drivers i got a bigger router. 25mm is good for "normal", slim speakers but the more the better.
A big disadvantage of AMTs is simply the vertical radiation! I take off axis frequency response more and more serious the longer I do speaker design and AMTs are very uneven in this regard.
Diameter alone does not tell everything with off axis response - T25A ot T34A are exceptions cause of their membrane shape. Also ring radiators beam stronger as their membrane is way bigger as the coil.
It depends on the baffle size and in which frequency area you get your reflections! As I build a wider speaker now woth 12" lf drivers i got a bigger router. 25mm is good for "normal", slim speakers but the more the better.
A big disadvantage of AMTs is simply the vertical radiation! I take off axis frequency response more and more serious the longer I do speaker design and AMTs are very uneven in this regard.
Diameter alone does not tell everything with off axis response - T25A ot T34A are exceptions cause of their membrane shape. Also ring radiators beam stronger as their membrane is way bigger as the coil.
thanks, any models in particular?
compression driver would be horns, right?
would love to try one, but i think these are out of my budget 😀
what about some speakers that have a 1-2cm foam strip around the edge of the baffle? i guess this has a similar effect
since the defraction frequency depends on the distance from driver middle to the edge, it might be worth calculating the frequency and go with the edge rounding "atleast" under the smallest frequency that is defracting?
tho my guess that edge rounding has also a influence on how soundwaves travel in general, they can more easly wrap around the speaker with large edge radius, specially low frequencys
Specially low frequencies ... don't care about edges at all. They are >5m wavelenght 🤓
Hi, misconception here. All sound frequencies diffract! sound going around corner is the diffraction, and inevitable. So why is that a concern?
What the roundover does is it helps with it, simplified, makes the edge disappear for short wavelengths. What "sharp" edge does it represents sudden change in acoustic impedance, and secondary sound source appears at the edge. The secondary sound source at the edge emits in opposite polarity to backward direction where the sound came from.
This secondary sound comes right after the direct sound, delayed by distance from source to any point on the edge, from there to mic, so usually smeared in time if distance from center of the source to edge varies, also depending on direction you measure at.
On a frequency response chart you would see interference ripple, as sound from the edge interferes with the direct sound. And this would happen roughly on frequency bandwidth from bafflestep up, all the way the transducer emits sound toward edge, up until the transducer beams.
If you are interested on this I suggest reading about it, but even more importantly playing with VituixCAD diffraction tool, as it alloes real time adjustment of baffle and ideal source giving lots of insight about it. You'd see interference, and when 2.5cm roubdover works better and when bigger would be more effective. Export data to the main program to see polar graphs. Have fun! 🙂
The very first response to this thread makes an important point.
A good tweeter application/installation starts with a good diffraction mitigation plan, and a good directivity plan. The filter network is key, and getting the tweeter integrated with the low/mid frequency driver is paramount. The kind of tweeter we might choose matters less.
When we perceive harshness, lack of detail, excessive brightness, dullness, lack of sparkle, piercing shrillness, analytical etched sound, lack of air... all these things are rarely the fault of the tweeter itself, and almost always the result of bad filter design, poor management of directivity, or too much high frequency diffraction.
I have found the Satori TW29TXN-B and the TW29BNWG to be excellent tweeters. Very detailed, yet very smooth and relaxed. But I am certain that with a sharp-edged high diffraction cabinet and a poor filter design, I could easily get them to sound miserable. Proper execution is everything.
I have also found that the SB26CDC, at 1/10 the cost of the beryllium waveguide TW29BNWG, can give an excellent performance. The difference between this $50 tweeter and the $500 beryllium is not that great, when both of them are carefully executed in a low diffraction baffle/cabinet and optimized with a good filter.
j.
A good tweeter application/installation starts with a good diffraction mitigation plan, and a good directivity plan. The filter network is key, and getting the tweeter integrated with the low/mid frequency driver is paramount. The kind of tweeter we might choose matters less.
When we perceive harshness, lack of detail, excessive brightness, dullness, lack of sparkle, piercing shrillness, analytical etched sound, lack of air... all these things are rarely the fault of the tweeter itself, and almost always the result of bad filter design, poor management of directivity, or too much high frequency diffraction.
I have found the Satori TW29TXN-B and the TW29BNWG to be excellent tweeters. Very detailed, yet very smooth and relaxed. But I am certain that with a sharp-edged high diffraction cabinet and a poor filter design, I could easily get them to sound miserable. Proper execution is everything.
I have also found that the SB26CDC, at 1/10 the cost of the beryllium waveguide TW29BNWG, can give an excellent performance. The difference between this $50 tweeter and the $500 beryllium is not that great, when both of them are carefully executed in a low diffraction baffle/cabinet and optimized with a good filter.
j.
There are a lot of replies and I have not read them all.
I get the impression the OP is a newbie to speaker design, so while a waveguide is a good option to reduce diffraction when done right, it is probably a step too far to hope a newbie will do it right.
I think we could make a guess and associate unpleasant treble (excluding on axis response issues) to off axis diffraction peaks and odd or high order distortion.
With that in mind I'd like to suggest a ring-radiator such as xt25 variations. These can be mounted flush on a baffle and the tighter directivity compared to many domes will minimise diffraction without going to a waveguide. Also they have slightly higher 2nd harmonic (sweetness?) with low 3rd and higher harmonics. The raw response is extremely smooth making xover design easier.
Edit: I'll add that the price of XT25SC90 has dropped recently making it almost guaranteed to please, IMO.
I get the impression the OP is a newbie to speaker design, so while a waveguide is a good option to reduce diffraction when done right, it is probably a step too far to hope a newbie will do it right.
I think we could make a guess and associate unpleasant treble (excluding on axis response issues) to off axis diffraction peaks and odd or high order distortion.
With that in mind I'd like to suggest a ring-radiator such as xt25 variations. These can be mounted flush on a baffle and the tighter directivity compared to many domes will minimise diffraction without going to a waveguide. Also they have slightly higher 2nd harmonic (sweetness?) with low 3rd and higher harmonics. The raw response is extremely smooth making xover design easier.
Edit: I'll add that the price of XT25SC90 has dropped recently making it almost guaranteed to please, IMO.
Last edited:
I'm very much in agreement with Jim's last response. Execution matters.
I also disagree with not using a waveguide for novices. Here I'm assuming we're talking about a HiFi dome tweeter and shallow waveguide like my own, not CD's mounted on horns that are 8" deep. The waveguide makes life so much easier for novice. Baffle diffraction is less of an issue, the directivity matches the woofer better, and the acoustic center also matches the woofer. Those three things were the areas I wrestled with in the crossover when I was new to speaker design. Now the crossover just falls into place quickly because waveguides largely prevent those issues. I would also point out that the "difficulties" of waveguides mentioned so far also apply to an unwaveguided tweeter - but are worse. For obvious reasons.
I also disagree with not using a waveguide for novices. Here I'm assuming we're talking about a HiFi dome tweeter and shallow waveguide like my own, not CD's mounted on horns that are 8" deep. The waveguide makes life so much easier for novice. Baffle diffraction is less of an issue, the directivity matches the woofer better, and the acoustic center also matches the woofer. Those three things were the areas I wrestled with in the crossover when I was new to speaker design. Now the crossover just falls into place quickly because waveguides largely prevent those issues. I would also point out that the "difficulties" of waveguides mentioned so far also apply to an unwaveguided tweeter - but are worse. For obvious reasons.
my 1st post on your thread:
https://www.diyaudio.com/community/threads/best-tweeter-for-silky-highs.403802/page-3#post-7466215