To me, 150-200Hz is well above subwoofer response, with the higher figure getting into lower midrange, but yes, run a sub too high and cone flexure (and beaming) can become more significant.
This having been said, I have distortion plots of an 18" driver where distortion from 30Hz to 2kHz remains below -40dB from fundamental, so it's not a hard and fast rule that large drivers suddenly distort above a certain frequency through cone flexure.
The driver in question actually has lower 3rd order distortion at 300Hz than it does at 30Hz!
In my experience and opinion, not the harmonic distortion (especially low order) in itself usually is the main problem if we are talking about cone or edge or spider resonances.
My latest try with 15" and 18" drivers up to about 400Hz doesn't showed high 2nd, 3rd HD but one driver sounded worse than the other and real bad distortion was heard, not just some harmonic distortion as slight coloration.
Although I corrected you that the XXL has 92 dB sensitivity, the comparison is interesting that said 3 x XXL 12" vs one (or two) 18XL1600
My XXL 12" in my system are playing up to 340 hz active filter ( DBX 234 Xs ) and MTM are 99 dB Audax HM210Z10 and Mundorf AMT with passive xover crossed at 1850 hz.
Sd is 1398 for 3 x XXL 12" and 1212 for one Faithal 18"
FS 21,8 hz for Peerlees and 32 for Faithal
Peerlees 100 liter each tuned to 23 hz and Faithal 125 liter each tuned to ca 28-30 hz.
3 x XLL 12" driver by a Krell F.B.I, and the Faithal by a Crown XTI 6002 with active filters.
Krell leaving about 700 class A watt in 5,66 ohms load, the Crown leaving 1200 watt/8 ohm ( and 2100 watt/4 ohm )
And i also tried 2 x Faithal on top of eachother = 4 ohm and an Sd of 2424
If I play them both separately 3 x XXL "hold together" the soundstage from very low SPL up to maximum SPL., with "high quality" sound.
And you hear the low frequencies "all the way".
The 18 Faithal perceived almost never to "hold togheter" the soundimage, and sounds "skinny" and a little "strange" up to about 100 dB, from there it plays "better" (but that is a definition).
But the Faithal don´t fill "the floor" with bass until real high SPL, and when it does it´s more of a "party-sound" ( hard to explain)
Belive me, i wanted the Faithal to work and in the basement i have more room for big speakers so no problem there.
Maby the Faithal was "unused" when i bought them, so no hour´s on them, but i give them around 30 hours....And still stiff as hell.
regards John
Hi, impedance sweep / TS parameters should show if TS parameters do not match datasheet. You could stretch them by hand or by playing tone. It is possible Fs and Q off what they should be, as was with my 15fh520. Mine might still be out of spec after years but haven't bothered remeasure other than frequency response.
Did you measure frequency response and match it to the XXL stack? TS parameters in FaitalPro datasheet are often measured after few hours of work so with heat in the system. This might be one reason you are not happy with the driver, if it settles to spec only after some heat buildup. This would be a thing that is not good for home speaker, but very good for pa speaker. It's better to drift it into spec, than out of spec, with full day workout.
You can compensate for this if you know the low temp TS params and build enclosure for that and adjust the response with DSP. Basically if you design with measurements, not with datasheet data.
Did you measure frequency response and match it to the XXL stack? TS parameters in FaitalPro datasheet are often measured after few hours of work so with heat in the system. This might be one reason you are not happy with the driver, if it settles to spec only after some heat buildup. This would be a thing that is not good for home speaker, but very good for pa speaker. It's better to drift it into spec, than out of spec, with full day workout.
You can compensate for this if you know the low temp TS params and build enclosure for that and adjust the response with DSP. Basically if you design with measurements, not with datasheet data.
Last edited:
Jawen, it's Faital, not Faithal. Just in case someone is searching for it.
At the time i haven´t got measurements possibility, so i only use the Crown XTI with DSP tried to boost the "20-ish" region 6-8 dB and listening.
But Crown XTI is a "complicated thing", so I was very bad at different settings, and had not the patience to learn.
I am sure that it could have been much better results, IF one had been used to using XTI functions
But ive give up that time after moving them big boxes around, and i like "simple" things...A preamp to DBX 234 Xs and from there to the amp and to the speaker.
I neved ever uses "loudness" or modified frequency response ( exept passive xover), and don´t want to have a "complicated" setup.
My belive is if a driver dosen´t sound good with the "right" xover.....Use another driver
And all Faital drivers ive been in contact with are super-stiff, but maby that doesen´t suit a woofer playing in your livingroom.
Se if i gave them a chance again, but my helth make all these things superhard.
And i haven´t learn how to measure TS parameters
Regards John
T/S parameters usually are all over the place, they very dependent on measuring methods and measuring levels, driver condition etc.
Here is an example of level dependency, JBL 2216Nd, measured at lower and higher input levels. This driver is like came from the marriage of a PA and a hi-fi woofer so not represents a full-blown PA woofer like the Faital 18XL1600 but the change of the parameters are clear. I saw a similar measurement of a typical 18" PA woofer with very stiff suspension where the Fs dropped from mid 30s to low 20s at high level input.
Hi John, yeah if XXL system suits you better thats good, what ever works. In general, PA drivers are optimized with assumption of measurement capability and DSP, which enables the high power ratings for example.
YSDR, yeah TS parameters are low signal parameters. So, if suspension stiffness reduces with high input level thats what jawen here described and true then, a proof.
YSDR, yeah TS parameters are low signal parameters. So, if suspension stiffness reduces with high input level thats what jawen here described and true then, a proof.
From TS parameters in wikipedia, emphasis mine:
"As an example, 𝑓s and 𝑉as may vary considerably with input level, due to nonlinear changes in 𝐶ms. A typical 110-mm diameter full-range driver with an 𝑓s of 95 Hz at 0.5 V signal level, might drop to 64 Hz when fed a 5 V input. A driver with a measured 𝑉as of 7 L at 0.5 V, may show a 𝑉as increase to 13 L when tested at 4 V. 𝑄ms is typically stable within a few percent, regardless of drive level. 𝑄es and 𝑄ts decrease <13% as the drive level rises from 0.5 V to 4 V, due to the changes in 𝐵𝑙. Because 𝑉as can rise significantly and 𝑓s can drop considerably, with a trivial change in measured 𝑀ms, the calculated sensitivity value (𝜂0) can appear to drop by >30% as the level changes from 0.5 V to 4 V. Of course, the driver's actual sensitivity has not changed at all, but the calculated sensitivity is correct only under some specific conditions. From this example, it is seen that the measurements to be preferred while designing an enclosure or system are those likely to represent typical operating conditions. Unfortunately, this level must be arbitrary, since the operating conditions are continually changing when reproducing music. Level-dependent nonlinearities typically cause lower than predicted output, or small variations in frequency response."
So, there very likely is drivers whose suspension related parameters vary more than with other drivers. It would be intuitive to assume the stiffer the suspension the more it could loosen up but it could also be material related stuff so probably not as simple as looking Cms only. There is also stiffening of suspension with excursion as the soft parts strech, Kms(x) in Klippel data. I need to read more on this.
Ugh, this is another variable for driver performance in any particular (bass) application, that doesn't seem to be predictable before actually testing a driver in action.
"As an example, 𝑓s and 𝑉as may vary considerably with input level, due to nonlinear changes in 𝐶ms. A typical 110-mm diameter full-range driver with an 𝑓s of 95 Hz at 0.5 V signal level, might drop to 64 Hz when fed a 5 V input. A driver with a measured 𝑉as of 7 L at 0.5 V, may show a 𝑉as increase to 13 L when tested at 4 V. 𝑄ms is typically stable within a few percent, regardless of drive level. 𝑄es and 𝑄ts decrease <13% as the drive level rises from 0.5 V to 4 V, due to the changes in 𝐵𝑙. Because 𝑉as can rise significantly and 𝑓s can drop considerably, with a trivial change in measured 𝑀ms, the calculated sensitivity value (𝜂0) can appear to drop by >30% as the level changes from 0.5 V to 4 V. Of course, the driver's actual sensitivity has not changed at all, but the calculated sensitivity is correct only under some specific conditions. From this example, it is seen that the measurements to be preferred while designing an enclosure or system are those likely to represent typical operating conditions. Unfortunately, this level must be arbitrary, since the operating conditions are continually changing when reproducing music. Level-dependent nonlinearities typically cause lower than predicted output, or small variations in frequency response."
So, there very likely is drivers whose suspension related parameters vary more than with other drivers. It would be intuitive to assume the stiffer the suspension the more it could loosen up but it could also be material related stuff so probably not as simple as looking Cms only. There is also stiffening of suspension with excursion as the soft parts strech, Kms(x) in Klippel data. I need to read more on this.
Ugh, this is another variable for driver performance in any particular (bass) application, that doesn't seem to be predictable before actually testing a driver in action.
Last edited:
It's late and no simulator at hand so quick thought experiment: if thinking stiff suspension and it loosening with more power, which is bad as sound would change, then for home use where this change might occur it would be better to use small closed box for such stiff driver. Logic would be that the small box would also add high air stiffness, and perhaps dominate if the suspension loosens up, reducing the difference in stiffness between small and large signal. So, if one has stiff PA driver, consider small closed box and DSP and power. Alternatively use driver without such issue.
It seems logical that the more the T/S parameters change of a driver at different drive levels, the more it changes its sound character.
Yeah, and it is kind of weird no one measures large signal except klippel. Hobbyist trying to find driver for a project could try calc TS params with higher input as well, try and scout how luch driver drifts spec with intended SPL range.
Could you elaborate, what's the thought process there? my processor power is consumed for today 😀 vented box limits cone excursion yeah, at tuning, but there is driver resonance now higher up in frequency, which still enables same problem? so small vented box?
A large vented box softens the driver suspension, it's the opposite of a small sealed box. If it's large enough, it start to dominate but in the other way than a sealed box. Or maybe I'm totally wrong here (sorry if so). 😆
Yea, i think this driver must be "treated" like Dr. Floyd E. Toole did when he desgned the Infinity MTS Prelude´s aktive woofer.
The woofer is a 12" with a elips-formed box not many liters, driven by a BASH amp 850 watt/ 8 ohm 2600 watt peak
He said :
What did the trick was 850W of BASH amplification (love those acronyms) and the light, stiff, low-resonance C.M.M.D., tied to a hefty voice-coil/magnet structure. "We drive the **** out of it" was Toole's scientific explanation of how the small box could produce such prodigious, super-clean low bass. I knew it the first time I heard the foundation of this speaker.
Can be read at side 3
https://www.stereophile.com/content/infinity-prelude-mts-loudspeaker-page-3
We must treat the big Faital the same way....Quite "small" box and drive the **** out of it!
Hi, yeah I'm also just thinking about it, few minutes in so we are learning here together I guess. Unexpected twists and turns can be expected, but thats fine as it all rains to bucket of knowledge 😀 Now after night sleep I need to learn more, on a holiday so might take few days to read. Lets get back to this if some data is found.
Overall this is very unfortunate realization at least for me. I was planning to make a system for a friend that is good for proper party and also for quiet home listening. Luckily nothing has been bought yet but the design process will take more time. Either I need to find driver without this issue, and if it is loose suspension one it likely doesn't have good xmax which means multiple drivers which drives up cost and size. Perhaps even two systems, one for home and one for party but then part of the system would always be unused. Luckily it was now and not later as it would have been even more unfortunate with cost. Let's see.
There are pro drivers with relative soft suspension, high power handling and relative high xmax. Eminence Kappalite 3012LF for example and the additional good news is that Eminence defines the xmax with measured 10% THD of the drvers and not just by the geometrical overhang of the voice coil.
The 3012LF is a monster 12" that puts many 15" to shame. I'm using two of them more than 5 years now at home.
Last edited:
| Usable Frequency Range | 46 Hz - 2 kHz |
Can you live with that?
Then price-wise neomagnet driver always more expensive, so maby look for ferromagnet drivers, to get most bang for the buck.
Yes, I am satisfied with them in net 75l vented boxes tuned about 30Hz.
There is no ceramic substitute for this drivrer and most pro manufacturers puts more development effort on neo drivers anyway.
IMO, forget the frequency range that is on the datasheets, they are defined by some criteria that may not your criteria.
There is no ceramic substitute for this drivrer and most pro manufacturers puts more development effort on neo drivers anyway.
IMO, forget the frequency range that is on the datasheets, they are defined by some criteria that may not your criteria.