To try and get existing specs and expect to make a decision on those is really a shot in the dark. If you can get SPL/Phase measurements from at 60db, 80db, 100db, and 120db levels, that would be a first good indication of dynamics. Also take note of the power used to accomplish those levels.
Designs with goals as Shin has is really a dedicated engineering effort. Rather than talk about things in general, it's really much better to get data and put it on the table to actually discuss it like a group of engineers.
I wish I had a project like that.🙂
Designs with goals as Shin has is really a dedicated engineering effort. Rather than talk about things in general, it's really much better to get data and put it on the table to actually discuss it like a group of engineers.
I wish I had a project like that.🙂
This time I hope you build a plain simple box for testing...and listening.... before building your special thing
Re: very good ?
I've lost which page I posted the measurements I took but here they are again:
Note the huge distortion increase around driver Fs. At 500hz and above the performance is very good for a 3" dome. Measurements taken at 90dB/2m.
I've got quite a few XO presets, some with the highpass on the mid set low and some higher.
Bratislav said:
I've lost which page I posted the measurements I took but here they are again:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Note the huge distortion increase around driver Fs. At 500hz and above the performance is very good for a 3" dome. Measurements taken at 90dB/2m.
I've got quite a few XO presets, some with the highpass on the mid set low and some higher.
tinitus said:
You only get to see the stuff that's half respectable, I do build rectangles too 😀
Re: Re: very good ?
Impressive indeed (for a 3" dome). And a lot better than the graphs I've seen sub 1KHz (maybe Bill improved them yet again?).
But point is, there was a lot of talk about 120dB and like. Even as good as your measurement is, it will look very different when you start pumping double digit watts into it. They are only 3" dome after all - they need to move a lot of air even at 500Hz. In that respect, a driver with more area will be superior (and that is why Sigfried used W22 in Orions).
The comparison between Orions and ATCs is actually quite ineteresting. I'm yet to organize a direct shootout (side by side) but even my non-scientific guestimates put Orions ahead in pretty much all areas - except one: SPL.
Orions are exceptionally fine transducers - like a well honed chamber orchestra they can be spellbounding.
But only ATCs will really rock the party.
In any case, I don't really think a 120dB goal is relevant for domestic environment. Unlike large venues (concert hall), "realistic" levels in my home sound way too loud. Would you really tolerate a drummer for an extended period in your living room?
But that is only me I guess.
ShinOBIWAN said:
Impressive indeed (for a 3" dome). And a lot better than the graphs I've seen sub 1KHz (maybe Bill improved them yet again?).
But point is, there was a lot of talk about 120dB and like. Even as good as your measurement is, it will look very different when you start pumping double digit watts into it. They are only 3" dome after all - they need to move a lot of air even at 500Hz. In that respect, a driver with more area will be superior (and that is why Sigfried used W22 in Orions).
The comparison between Orions and ATCs is actually quite ineteresting. I'm yet to organize a direct shootout (side by side) but even my non-scientific guestimates put Orions ahead in pretty much all areas - except one: SPL.
Orions are exceptionally fine transducers - like a well honed chamber orchestra they can be spellbounding.
But only ATCs will really rock the party.
In any case, I don't really think a 120dB goal is relevant for domestic environment. Unlike large venues (concert hall), "realistic" levels in my home sound way too loud. Would you really tolerate a drummer for an extended period in your living room?
But that is only me I guess.
Re: Re: Re: very good ?
I wouldn't! The ones I know drink far too much to keep in the house. 🙂
Bratislav said:
I wouldn't! The ones I know drink far too much to keep in the house. 🙂
Shin,
The spectral decay looks like it's extending too long for an accurate sound, and the ridge at around 5K is really not good. If you can post the spectral decay for the first 1ms or less I could have a closer look.
The spectral decay looks like it's extending too long for an accurate sound, and the ridge at around 5K is really not good. If you can post the spectral decay for the first 1ms or less I could have a closer look.
Re: Re: Re: very good ?
I'm guessing you looked at distortion figures for mk1 or mk2 mid dome. These are for the latest mk3 introduced in 2001. ATC offer upgrade policies on a trade-in basis to replace older drivers with the newer revisions.
If your judging the ATC mid performance by the SCM100 then you'll find that's a mistake. Take a look at the graphs again and realise that ATC use it in the very non-linear regions. Do you like this colouration? Its not that objectionable but its audible in an AB against an XO that uses the most linear operating range and frankly I much prefer it.
When used incorrectly its so very easy to make it sound harsh, forward and fatiguing. The subtlety is lost and you get a colder sound that makes strings in particular sound strident and brittle rather than warm and involving. ATC are excellent engineers and I have no doubt of their technical competance, they clearly have more experience than any DIY'er or SL for that matter. The success of their products and drivers shows that. But to get the best out of the ATC mid you need steep filters, the likes of which are digital based. This obviously means a radical redesign of the SCM products and lets face it - why should they since they're already so popular.
ATC specs of 120dB are impossible to acheive with the dome mid, you'll likely find that's peak SPL measured at a particular frequency(most likely the bass region) which isn't that impressive considering those are the specs of a pair of loudspeaker playing with a combined total of four 15" drivers.
The ATC dome is only really suitable for playing upto around 105dB/2m with clarity. Above this and my ears start to tell me its time to turn it down because its just noise rather than anything resembling fidelity. This still is still a good way ahead of the likes of anything from Scanspeak, Seas and the other 'polite' sounding hi-end manufacturer's which I'd never consider for a serious loudspeaker now, they just don't have the bite to pull off a realistic sound.
Speaking of the W22, I've used these and its an OK driver. My main problem is that its got bugger all dynamics, instead its soft and detailed but not what I'd call a driver that capable of utter believabilty, at least not when used singularly and not in dipole where efficiency is lessened further. Funnily enough I sold them to Vikash who is now using them in a pair of Orions.
This is all IMO 😉
Transient impact and SPL aren't the same thing. I'm looking for more dynamics not SPL.
Bratislav said:
I'm guessing you looked at distortion figures for mk1 or mk2 mid dome. These are for the latest mk3 introduced in 2001. ATC offer upgrade policies on a trade-in basis to replace older drivers with the newer revisions.
If your judging the ATC mid performance by the SCM100 then you'll find that's a mistake. Take a look at the graphs again and realise that ATC use it in the very non-linear regions. Do you like this colouration? Its not that objectionable but its audible in an AB against an XO that uses the most linear operating range and frankly I much prefer it.
When used incorrectly its so very easy to make it sound harsh, forward and fatiguing. The subtlety is lost and you get a colder sound that makes strings in particular sound strident and brittle rather than warm and involving. ATC are excellent engineers and I have no doubt of their technical competance, they clearly have more experience than any DIY'er or SL for that matter. The success of their products and drivers shows that. But to get the best out of the ATC mid you need steep filters, the likes of which are digital based. This obviously means a radical redesign of the SCM products and lets face it - why should they since they're already so popular.
ATC specs of 120dB are impossible to acheive with the dome mid, you'll likely find that's peak SPL measured at a particular frequency(most likely the bass region) which isn't that impressive considering those are the specs of a pair of loudspeaker playing with a combined total of four 15" drivers.
The ATC dome is only really suitable for playing upto around 105dB/2m with clarity. Above this and my ears start to tell me its time to turn it down because its just noise rather than anything resembling fidelity. This still is still a good way ahead of the likes of anything from Scanspeak, Seas and the other 'polite' sounding hi-end manufacturer's which I'd never consider for a serious loudspeaker now, they just don't have the bite to pull off a realistic sound.
Speaking of the W22, I've used these and its an OK driver. My main problem is that its got bugger all dynamics, instead its soft and detailed but not what I'd call a driver that capable of utter believabilty, at least not when used singularly and not in dipole where efficiency is lessened further. Funnily enough I sold them to Vikash who is now using them in a pair of Orions.
This is all IMO 😉
Transient impact and SPL aren't the same thing. I'm looking for more dynamics not SPL.
soongsc said:
Its a function of the closed back design and the dome itself. What you've got to bear in mind is the time scale vs. frequency and the CSD floor range. Compare the results to a small cone driver:
ATC SM75-150S:
An externally hosted image should be here but it was not working when we last tested it.
SS 15W8530:
Seas W15CH001
Those results are from Zaphs sites.
Also remember that the measurements taken on the dome mid were with driver mounted in the enclosures.
soongsc said:
Yes you can change all the usual aspects of the windowing function, including number of time slices and time slice shift which are the ones it sounds like your interested in.
An externally hosted image should be here but it was not working when we last tested it.
No gating applied to that one. Note that the CSD floor is -25dB as opposed to -35dB I used on the other one. FFT size same as before.
Speaking of the W22, I've used these and its an OK driver. My main problem is that its got ****** all dynamics, instead its soft and detailed but not what I'd call a driver that capable of utter believabilty, at least not when used singularly and not in dipole where efficiency is lessened further. Funnily enough I sold them to Vikash who is now using them in a pair of Orions.
I have the W22. I know it well. It has nasty breakup modes
and requires a lower crossover point, under 2khz steep slope.
Others prefer closer to 1500hz. The ATC softdome can do higher
and it has that advantage. W22 advantage is smooth sound up
to it's low pass limit and offers nice bass. Also, sensitivity is weak,
so to get life like sound and impact, minimum requirement is 160v
headroom amp {min 600w @ 8 ohms}, then the drivers starts to
come alive.
I have the W22. I know it well. It has nasty breakup modes
and requires a lower crossover point, under 2khz steep slope.
Others prefer closer to 1500hz. The ATC softdome can do higher
and it has that advantage. W22 advantage is smooth sound up
to it's low pass limit and offers nice bass. Also, sensitivity is weak,
so to get life like sound and impact, minimum requirement is 160v
headroom amp {min 600w @ 8 ohms}, then the drivers starts to
come alive.
omega87 said:
Hmm..
Less than 5% (2nd order) at 3 kHz for 113 db?
It doesn't scale nearly as badly as I would have thought (.. actually much better than a dome). Perhaps it would still be below 1 % at 97db? Any chance you can obtain a measurement for 1 watt at 1 meter?
Note where the two major ridges are in the CSD plot? The second one is at a x2 frequency which means these are the dome's first and second vibration modes, and corresponds to the diameter of the dome. This happens because of the following reasons:
1. The dome is not stiff enough so when excited from the VC which is at the outer edge of the dome.
2. For cones, you can use the surround materail to absorb the energy, but with a dome, the energy is not so easily transferred elsewhere so it relies on dome flexture to absorb this energy, thus the ringing must exists.
This also prevents the other frequencies from decaying faster. If the decay cannot get below the reference SPL by at least 12db within 0.04ms for frequencies above 1K, the sound will be sort of mushy.
If you can find a stiff dome, preferably metal with some additional stiffening/damping coating, it would be much better. If you can do with less power to the drivers, a cone type mid range can raise this vibration mode to better than twice the frequency. A much better situation.
1. The dome is not stiff enough so when excited from the VC which is at the outer edge of the dome.
2. For cones, you can use the surround materail to absorb the energy, but with a dome, the energy is not so easily transferred elsewhere so it relies on dome flexture to absorb this energy, thus the ringing must exists.
This also prevents the other frequencies from decaying faster. If the decay cannot get below the reference SPL by at least 12db within 0.04ms for frequencies above 1K, the sound will be sort of mushy.
If you can find a stiff dome, preferably metal with some additional stiffening/damping coating, it would be much better. If you can do with less power to the drivers, a cone type mid range can raise this vibration mode to better than twice the frequency. A much better situation.
I've had a brief chat with Simon and he recommends the RAAL over the R3.2.
I think it would be prudent to at least buy one see what its like and then go from there.
I've always thought multiple tweeters was a disgusting habbit but I find myself having to look at this possibility in order to cross lower using lightweight efficient drivers.
If I remove the face plate on the drivers I can get them pretty close, something along these lines:
I think it would be prudent to at least buy one see what its like and then go from there.
I've always thought multiple tweeters was a disgusting habbit but I find myself having to look at this possibility in order to cross lower using lightweight efficient drivers.
If I remove the face plate on the drivers I can get them pretty close, something along these lines:
An externally hosted image should be here but it was not working when we last tested it.
More estoteric tweeters to look at.
LCY ribbons. Reviews are great, wide dispersion, very affordable.
Not a SPL monster.
http://www.e-speakers.com/products/lcy-components.html
Alian ribbons - expensive
http://gallery.bcentral.com/GID4956567P4322677-ALIAN/ALIAN-true-ribbon-tweeter.aspx
ALIAN 24020
97db/w, 300hz - 35khz MSRP: $3999.ea
ALIAN 18015
97db/w, 800hz - 35khz MSRP: $2999.ea
ALIAN 12015
97db/w, 1500hz - 45khz MSRP: $1775.ea
LCY ribbons. Reviews are great, wide dispersion, very affordable.
Not a SPL monster.
http://www.e-speakers.com/products/lcy-components.html
Alian ribbons - expensive
http://gallery.bcentral.com/GID4956567P4322677-ALIAN/ALIAN-true-ribbon-tweeter.aspx
ALIAN 24020
97db/w, 300hz - 35khz MSRP: $3999.ea
ALIAN 18015
97db/w, 800hz - 35khz MSRP: $2999.ea
ALIAN 12015
97db/w, 1500hz - 45khz MSRP: $1775.ea
I've always thought multiple tweeters was a disgusting habbit but I find myself having to look at this possibility in order to cross lower using lightweight efficient drivers.
Two tweeters is signs of line sourcing
re: Raal 140-15d tweeter
http://www.raalribbon.com/products_flatfoil_140-15.htm#
- Gap induction ( mean value over gap width): 0.51 T
- Frequency response: 500 Hz ~ 100 kHz
- Sensitivity: 95 dB / 1 m / 2.83 V
- Impedance: 8 Ohm
- Program power handling: 200 W*
- *Recommended crossover: 4th order L-R @ 1600 Hz
Random thoughts -
.5T isn't that great and it shows in their low 95dB sensitivity
rating. You can find other ribbons with 1T with 98dB - 102dB
sensitivity. They rate the tweeter for 500hz - 100khz, but
the recommended crossover is 4th order 1.6khz. LOL
That design of the Raal seems to imply it that it might have
better than average vertical dispersion, a good thing, but
given the higher crossover point and lower sensitivity, the Raal
might be in the same class as the LCY ribbons.
The Raal mechanical design doesn't look like it's line source
friendly as other tweeters unless you use the 70-10, but it
has it's own unique problems to deal with.
http://www.raalribbon.com/products_flatfoil_70-10.htm
Two tweeters is signs of line sourcing
re: Raal 140-15d tweeter
http://www.raalribbon.com/products_flatfoil_140-15.htm#
- Gap induction ( mean value over gap width): 0.51 T
- Frequency response: 500 Hz ~ 100 kHz
- Sensitivity: 95 dB / 1 m / 2.83 V
- Impedance: 8 Ohm
- Program power handling: 200 W*
- *Recommended crossover: 4th order L-R @ 1600 Hz
Random thoughts -
.5T isn't that great and it shows in their low 95dB sensitivity
rating. You can find other ribbons with 1T with 98dB - 102dB
sensitivity. They rate the tweeter for 500hz - 100khz, but
the recommended crossover is 4th order 1.6khz. LOL
That design of the Raal seems to imply it that it might have
better than average vertical dispersion, a good thing, but
given the higher crossover point and lower sensitivity, the Raal
might be in the same class as the LCY ribbons.
The Raal mechanical design doesn't look like it's line source
friendly as other tweeters unless you use the 70-10, but it
has it's own unique problems to deal with.
http://www.raalribbon.com/products_flatfoil_70-10.htm
- Status
- Not open for further replies.