That is my experience as well. There are small subtle differences between a good aluminum dome (SB26CDC) and a good beryllium dome (Satori TW29BNWG and Bliesma T25B), but those differences are not life changing, they are not immediately noticeable... like I said, small and subtle. The large difference in cost may be worth it in a critical listening application, and for me it is worth it. In a more casual listening situation, the large cost delta may not be justified.
Would you say that is true of the upper end silk domes? I am thinking of ScanSpeak revelator / illuminator series, Satori TW29DN, etc... ?
The problem with most silk tweeters is that they leave piston movement already at 8-10kHz. They don't have wild resonances going on but when you look at frequency response on axis and off axis you know that's not the behaviour of a piston membrane.
And I'm with you that we schould not hear a significant difference between perfectly piston movement tweeter membranes. But there is a transition zone - I really dislike these old Seas tweeters with resonance at 18-20kHz - even if I can't hear to 18kHz. I'm fine with Bliesmas Alu with it's resonance >30kHz but still think the B version is more "natural" (or less "inyourface").
In direct comparison silk always sounded less detailed in my trials - that's the reason to use silk. I really like ring radiators here, I find them very natural too in the upper range. But I dislike the strong beaming at very high frequencies. (I built one Monitor with ring radiator and small ScanSpeak lowmid. When sounding right in nearfiel it was to dull in the room. When good in the room it was painfull at close distance)
Focal - a friend of mine has a Focal speakers and they get very piercing when plaied loud. Seems like they could need a lesson in crossover design ;-). But really - these are from a cheaper series but big speakers and they really should have spent 3-5 bucks more for a better crossover, it's a pitty.
And I'm with you that we schould not hear a significant difference between perfectly piston movement tweeter membranes. But there is a transition zone - I really dislike these old Seas tweeters with resonance at 18-20kHz - even if I can't hear to 18kHz. I'm fine with Bliesmas Alu with it's resonance >30kHz but still think the B version is more "natural" (or less "inyourface").
In direct comparison silk always sounded less detailed in my trials - that's the reason to use silk. I really like ring radiators here, I find them very natural too in the upper range. But I dislike the strong beaming at very high frequencies. (I built one Monitor with ring radiator and small ScanSpeak lowmid. When sounding right in nearfiel it was to dull in the room. When good in the room it was painfull at close distance)
Focal - a friend of mine has a Focal speakers and they get very piercing when plaied loud. Seems like they could need a lesson in crossover design ;-). But really - these are from a cheaper series but big speakers and they really should have spent 3-5 bucks more for a better crossover, it's a pitty.
I agree, especially because of the rather detailed measurements of drivers.
With kind regards
Michael
Does a silk midrange driver of 3 inch size also no longer behave like a piston at a usual crossover frequency of about 2 to 3 kHz?
With kind regards
Michael
@IamJF @hifijim @Endo2112 I agree most silk domes tend to sound very close to all-around identical in overall character. It will come down to details like:
These factors will have some influence on the pistonic range as well as the break up point and severity. The dome material and dampening coating composition will determine the upper cutoff frequency and how detailed the area above breakup will sound. Many of the better domes have large motors with high BL to combat the increased Mms limiting upper cutoff point. What doesn't change is the off axis HF drop-off point, which is mainly affected by dome geometry and dimensions. A flatter profile dome will typically have better off axis behavior. That's the benefit of larger diameter Be domes with greater structural strength than other dome materials, given the same piston diameter.
There are some bad resonances on a few Seas and ScanSpeak domes, caused mainly by standing waves right behind the pole piece and in the VC gap, if not tapered, especially on wide VC gaps. These drivers tend to have foam dampened pole pieces and radial dome profiles. This combination has the tendency to provoke these near ultrasonic resonances, which are just as likely to worsen IMD across the entire pass band of the driver as the HF breakup of the dome material itself on Ti and other hard domes.
HF cavity resonances can be cured by occupying as much of the open airspace directly behind the dome and above the pole piece as possible with dampening material. What's remarkable on some Audax domes is the total lack of pole piece dampening on several of their models, yet they don't exhibit much of any cavity resonance and show very clean CSD. I believe this has alot to do with the catenary dome profile along with a tapered pole piece vent they use on most of their higher end tweeters.
The surround width will have a large effect on linearity around the fundamental Fs area and also the off axis breakup FR profile / HD spectrum. It won't affect the character in the pistonic range much unless you increase SPL to the point the dome progressively deforms under acoustical load. That's when things change much more and get ugly, especially if the rear dome enclosed air volume is relatively small. The center of the dome can start to snap back and forth rapidly, creating an out of phase situation between the dome center and perimeter along with other acoustical artifacts. You can see this on the FR plot as abrupt amplitude swings and it can be at any frequency in the driver's pistonic range. This also wreaks havoc on phase linearity.
If you look at most ring radiating soft domes, they are very clean up top not having an unsupported dome center to go out of control. The surround on these drivers becomes the exclusive radiating surface area, but you're no longer dealing with a dome at this point. The dirt cheap Peerless XT25 ring radiator has such clean and extended top end compared to most soft domes, but at the expense of higher HD around Fs. You can obviously get around this with a WG.
HF domes without properly dampened rear chambers often exhibit a secondary resonance of the even smaller air volume behind the VC gap. This worsens odd order HD around the lower midrange just above Fs. Better drivers will have a vented VC former to couple the two separate volumes together to form a larger one. As long as the venting is sufficient in allowing free airflow between both cavities, it will eliminate the issue. SB Acoustics does a great job with this on even their least expensive units, which ScanSpeak could learn a thing or two about with even their most expensive soft domes. Not sure why this is the case.
So this turned into a long winded post which I thought would at least explain why some domes sound different than others for various reasons other than their diaphragm material.
That being said, I tend to favor larger HF domes mainly due to their higher sensitivity and ability to cross lower without much strain. The downsides of this being an earlier HF cutoff point is worth it to my older ears not hearing much past 15k. I haven't compared any other larger soft domes than thr Seas T35C002 and Audax TW034. I certainly prefer the Seas over the Audax for obvious performance differences. I haven't heard the silk dome T34S yet. I doubt it will be significantly different than the Seas. The question is how low can you push the T34S compared to the others.
- motor design (underhung vs overhung)
- overall dome size and proportion to surround width
- dome profile geometry
- rear dampening and chamber volume
- type / blend of fabric material
- type of dampening and sealing coating applied to dome (this is by far the biggest influencer IMO)
These factors will have some influence on the pistonic range as well as the break up point and severity. The dome material and dampening coating composition will determine the upper cutoff frequency and how detailed the area above breakup will sound. Many of the better domes have large motors with high BL to combat the increased Mms limiting upper cutoff point. What doesn't change is the off axis HF drop-off point, which is mainly affected by dome geometry and dimensions. A flatter profile dome will typically have better off axis behavior. That's the benefit of larger diameter Be domes with greater structural strength than other dome materials, given the same piston diameter.
There are some bad resonances on a few Seas and ScanSpeak domes, caused mainly by standing waves right behind the pole piece and in the VC gap, if not tapered, especially on wide VC gaps. These drivers tend to have foam dampened pole pieces and radial dome profiles. This combination has the tendency to provoke these near ultrasonic resonances, which are just as likely to worsen IMD across the entire pass band of the driver as the HF breakup of the dome material itself on Ti and other hard domes.
HF cavity resonances can be cured by occupying as much of the open airspace directly behind the dome and above the pole piece as possible with dampening material. What's remarkable on some Audax domes is the total lack of pole piece dampening on several of their models, yet they don't exhibit much of any cavity resonance and show very clean CSD. I believe this has alot to do with the catenary dome profile along with a tapered pole piece vent they use on most of their higher end tweeters.
The surround width will have a large effect on linearity around the fundamental Fs area and also the off axis breakup FR profile / HD spectrum. It won't affect the character in the pistonic range much unless you increase SPL to the point the dome progressively deforms under acoustical load. That's when things change much more and get ugly, especially if the rear dome enclosed air volume is relatively small. The center of the dome can start to snap back and forth rapidly, creating an out of phase situation between the dome center and perimeter along with other acoustical artifacts. You can see this on the FR plot as abrupt amplitude swings and it can be at any frequency in the driver's pistonic range. This also wreaks havoc on phase linearity.
If you look at most ring radiating soft domes, they are very clean up top not having an unsupported dome center to go out of control. The surround on these drivers becomes the exclusive radiating surface area, but you're no longer dealing with a dome at this point. The dirt cheap Peerless XT25 ring radiator has such clean and extended top end compared to most soft domes, but at the expense of higher HD around Fs. You can obviously get around this with a WG.
HF domes without properly dampened rear chambers often exhibit a secondary resonance of the even smaller air volume behind the VC gap. This worsens odd order HD around the lower midrange just above Fs. Better drivers will have a vented VC former to couple the two separate volumes together to form a larger one. As long as the venting is sufficient in allowing free airflow between both cavities, it will eliminate the issue. SB Acoustics does a great job with this on even their least expensive units, which ScanSpeak could learn a thing or two about with even their most expensive soft domes. Not sure why this is the case.
So this turned into a long winded post which I thought would at least explain why some domes sound different than others for various reasons other than their diaphragm material.
That being said, I tend to favor larger HF domes mainly due to their higher sensitivity and ability to cross lower without much strain. The downsides of this being an earlier HF cutoff point is worth it to my older ears not hearing much past 15k. I haven't compared any other larger soft domes than thr Seas T35C002 and Audax TW034. I certainly prefer the Seas over the Audax for obvious performance differences. I haven't heard the silk dome T34S yet. I doubt it will be significantly different than the Seas. The question is how low can you push the T34S compared to the others.
The Bliesma silk material is pretty thin and made for creating the sound. I had a switching impulse during my tests and as the membrane "jumped" I saw it deforming already. Of course that's not a signal when listening to music but these silk membranes are thin and leightweight.
Looking at th HiFiCompass off axis data it seems to perform flawless until the resonance (dip in frequency response). You normally see a change in off axis fall off when the membrane is not in piston movement any more. (it breaks up in different phase radiation parts - off axis strange things happen. Or the surface of in phase radiation get's smaller)
So I don't think there is any problem up to 3,5kHz at least. But of course the membrane is beaming narrow there, you would get a pronaunced jump in radiating when crossed over so high.
@profiguy The Bliesma A membranes have great off axis radiation and have a very high dome. The flat B's are not as good in that regard but way better for waveguides. This simulation flies around the internet - sorry I don't know the creator.
Looks like one of Gaga's images, he has made many such simulation comparisons
https://www.diy-hifi-forum.eu/forum...y-wie-geht-das&p=243693&viewfull=1#post243693
The B's are easier and more forgiving in waveguide designs for sure, but they cannot be as wide from 10 to 20K as the A's if that is a design objective.
Unless the speaker is intended to be used in an infinte baffle, those simulations start to lose applicability to wide radiating domes in waveguides, where the entire cabinet in 3D has to be included to have any hope of getting an accurate simulation. I fell into that trap for quite a while myself.
I can't do anything at all with such descriptions. Has anyone ever thought about how exactly the midrange drivers have to be integrated so that they all have the same conditions? Were they all midrange drivers of otherwise identical loudspeakers, of course with respective precisely matched crossovers? Did the test listener know which midrange driver he was listening to?
Questions about questions.....
With kind regards
Michael
Well Michael I won’t comment on these drivers specifically. But I can comment on something that you may find useful, in terms of why different materials may sound different.
First off I do all my designs with full 360 degrees measurements, usually at intervals of 10 degrees. I feel that this is the current best practice, enabled by tools like (but not exclusive to) VituixCAD and Klippel NFS.
Now the part I can comment on is the bit about “precisely matched crossovers”.
Because I matched, with DSP, a hard cone driver to a soft cone driver (Purifi alu 6.5” to Purifi 6.5” fibre) so that two different speakers, with exactly the same cabinet and tweeter, had exactly the same on axis frequency response.
However, I found that the off axis responses are NOT perfectly matched.
To be clear for other reader following, it’s not possible to control all of the other off-axis responses for the driver- it’s predetermined by the cone/surround and cabinet.
But this, to me, is one reason why different cone materials will always sound different. And no amount of crossover tuning can make them sound the same.
Simply because they just don’t measure the same in 3D space.
I will have to dig my data out somewhere to illustrate this. But in the meantime, here’s a taster of what I mean.
Courtesy of www.Hificompass.com
Relative to the on axis response (akin to what would happen if you could massage the on-axis frequency response to perfectly flat +/- 0.0 dB), this is what a few of the off axis responses would look like at 15, 30, 45 and 60 degrees, for the PTT6.5X fibre cone.
and aluminium cone PTT 6.5X:
You can see that in just 4 different angles, there are differences in the typical pass band eg. 40Hz to 2-3KHz; as well as the stop band (beyond 2-3KHz)
Last edited:
Thanks for the source! Woha - these guys do great stuff!
https://www.diy-hifi-forum.eu/forum...y-wie-geht-das&p=243693&viewfull=1#post243693
Therefore I use the B's in waveguide (front) and A's for wide radiation (rear/ceiling) in my next project. Looking forward to it!
This Sounds like a sick Setup.... Would something like this work for studio use ?Btw - I don't really understand you concept. You could easily build a very nice speaker with SDX12, PTT6.5 and T34B as 3-way. Fully active, great amps, smooth off axis behavior. To boring?
In principle - yes.
But for serious use I normally start one step earlier - what's the room, setup, usecase, preferences. Do you need wide radiating speakers or controlled directivity (waveguide etc).
You can solve A LOT of problems with at least a single bass array at the front (and normally need damping at the back anyways).
I will soon upgrade my main speakers to 2x SB29NRX75-8, M74B (A would also work but hey - I do this once
) and T25B. That works for sure for studio use - it's like an updated ATC SCM 200.The main reason for a T34B is the low crossover frequency and extremely high SPL level. SDX12 and PTT6.5 can't deliver this SPL level but need the low crossover frequency. You could do 2x 12", M74A/B and T34B for very high SPL needs. Or do a 4 way with a bass array. (in my room I can do a bass array with the 3-way - depends on the size of the frontwall).
Sorry that's probably a little confuse ... you can write me a pm about your situation.
Yes, I just had a phone call with Soundimports to check.
The (when I am right) American manufacturer of the beryllium foil stopped delivery to Scanspeak, Bliesma and likely SBA.
Besides, Soundimports told that Bliesma has manufacturing / delivery problems with their other units too.
Soundimports is not optimistic concerning Bliesma's future....
The (when I am right) American manufacturer of the beryllium foil stopped delivery to Scanspeak, Bliesma and likely SBA.
Besides, Soundimports told that Bliesma has manufacturing / delivery problems with their other units too.
Soundimports is not optimistic concerning Bliesma's future....
Hi lamJF and fluid,
Thanks for your nice words. Yes, these images were made by me and are taken from the thread on 'How to do a waveguide and achieve constant directivity' here.
At that time, I simply tried to understand what are the best concave and convex membrane forms for optimal (i.e. wide directivity up to the high frequency range) directivity. Starting from there, I learned that a tiny 'step' around a convex tweeter membrane (as we usually use) significantly improves directivity behaviour.
With respect to different shapes of convex domes I learned, that Blisma was spot on for optimal directivity ( I assume they exactly knew what they did here). I was simply surprised about the nice characteristics of the Blisma tweeters and wanted to learn, how they achived this.
I'm following your thread with great interest for quite some time anyway.
Regards,
Christoph
Thanks for your nice words. Yes, these images were made by me and are taken from the thread on 'How to do a waveguide and achieve constant directivity' here.
At that time, I simply tried to understand what are the best concave and convex membrane forms for optimal (i.e. wide directivity up to the high frequency range) directivity. Starting from there, I learned that a tiny 'step' around a convex tweeter membrane (as we usually use) significantly improves directivity behaviour.
With respect to different shapes of convex domes I learned, that Blisma was spot on for optimal directivity ( I assume they exactly knew what they did here). I was simply surprised about the nice characteristics of the Blisma tweeters and wanted to learn, how they achived this.
I'm following your thread with great interest for quite some time anyway.
Regards,
Christoph
Reliable Dutch vendor Soundimports indicates these Bliesma units as "end of life"
Materion, the sole true beryllium supplier, is the bottleneck here. Almost 6 months ago a close source to me advised that ETA was 9 months for a standard product that they'd been using for ages, and MOQ are gone way up!!
First it was Scan-Speak, now Bliesma. Soon it will be... <insert your favourite brand here>
I was writing with Bliesma a few weeks ago. Materion has ended the life of some Beryllium foils which where needed for the membranes of these speakers.
They seem to work together to find an alternative foil/material and parameters of the speakers should not shift too much.
Nevertheless I ordered a bunch of these tweeters for my next projects, I belive it will take a while until they are available again.
And Bliesma seem to be pretty bussy - not sure if they are in trouble. But they can't deliver speaker they don't get parts for ... as ScanSpeak and probably all the other BE tweeters.
They seem to work together to find an alternative foil/material and parameters of the speakers should not shift too much.
Nevertheless I ordered a bunch of these tweeters for my next projects, I belive it will take a while until they are available again.
And Bliesma seem to be pretty bussy - not sure if they are in trouble. But they can't deliver speaker they don't get parts for ... as ScanSpeak and probably all the other BE tweeters.