I don’t get this one. Can you post this back a step before you truncate it?
"Patrick" I am always fascinated by your speculations, 3D modelling and (sometimes) even testing results 
Perhaps in the future we will not just design audio and other stuff in 3D, but like in a Sci-Fi story, we will able to digitize ourselves and listen to a modeled audio system in a simulated room in a simulated universe
Perhaps in the future we will not just design audio and other stuff in 3D, but like in a Sci-Fi story, we will able to digitize ourselves and listen to a modeled audio system in a simulated room in a simulated universe 
You say 'impulse response'. Do you mean frequency response by that? A lot of biradial horns are not linear, they usually dip more and more to the upper end. Not all but especally with horns which start to load at low frequencys and work over a very wide frequency range do. That's usually the trade-off you have to pay. To counter that, the horn can be shortened, which leads to constructions like the JBL Progressive Transition Waveguide (or Dayton copy of it). They work excellent but got also a trade-off, the gain is much lower. With modern drivers that's usually not a problem, especally if you use them for HiFi at home, PA use can be different though.
I know. 3d printing is in many cases ideal for that in development. Well, how much would it cost for you to print a bi-radial horn with a cut off frequency of, let's say, 1200-1500Hz? I'd really like to build a kinda clone of one of the JBL bi-radial monitors like the JBL 4430.
No, you don't. At least not in europe. And if they do once in a decade, they want a kidney and a heart for it.
Yes, you can get them but they are still very expensive. TBH, there are much better mid/bass drivers you can get nowadays for a lot less money, I don't see the 2235 as that legendary and irreplacable, just because there's the JBL label on it and they are 'vintage'. And because you can get the better, more recent woofers new, they are then in a excellent condition, which is doubtful for most 2235. Well, I don't mind reconing or some optical flaws, I just don't see why I should pay more for something which does not give me any benefit sound wise. I thank you for your suggestions, it's just not what I want. I don't want to build a 4430 with old drivers and horn and a new enclosure, my goal is to build a clone, a speaker with most of the characteristics and for a more reasonable price.
So which 15" driver would you suggest ? The advantage of the 2235 was that it is midway between a HiFi driver and a PA driver. It canbe tuned quite low and used up to approx 1 kHz. There are not many of these nowadays and those that are I would not call cheap. And most of the cheap ones go quite low but not very high or quite high but not low. The new variant of the Beyma Studio woofer does look interesting but it doesn't have a lot of x-max.
Regards
Charles
Regards
Charles
If you are talking about the JBL 2235 and what's comparable, you should not complain about a low Xmax, it only got +-3mm pure geometrically, (coil height-pole plate) / 2. In praxis it can do more, but so do most drivers. Damage starts at 11mm excursion (Xdmg 22mm peak to peak), so that's not an excursion monster at all.
There are few drivers, that's true, but there are some. One of them is the Celestion TF1525, nice midrange, practically resonance free in the usable rane (-> Test Selbstbauprojekt - Monacor K+T Klonwall - Bildergalerie - Bild 9) and the best part is, you can get them for 105€ (Thomann or Strassacker i.e. over here in Germany). There are others but that would be my first choice.
There are few drivers, that's true, but there are some. One of them is the Celestion TF1525, nice midrange, practically resonance free in the usable rane (-> Test Selbstbauprojekt - Monacor K+T Klonwall - Bildergalerie - Bild 9) and the best part is, you can get them for 105€ (Thomann or Strassacker i.e. over here in Germany). There are others but that would be my first choice.
That looks like a good budget choice but it's not at all up to the standards of the 2235h, which offers a much more robust bottom end, with 6mm one direction mathematical, not sure where you pulled 3 from, and is usable to reasonable LF in a reasonable box size. The celestion requires a much larger box for any reasonable vented tuning, and doesn't work sealed in realistic enclosures.
The very flat celestion published curve is almost certainly a poor reflection of reality- the waterfall on your link shows a substantial rising response which is much more likely.
I agree with the assessment that the 2235h prices are an absurdity though. Better results could be had for less money with a number of drivers, though you might have to add weight to drop the efficiency and Fs down in many cases (the 2235h uses a pro-style motor with a ring to add mass).
The very flat celestion published curve is almost certainly a poor reflection of reality- the waterfall on your link shows a substantial rising response which is much more likely.
I agree with the assessment that the 2235h prices are an absurdity though. Better results could be had for less money with a number of drivers, though you might have to add weight to drop the efficiency and Fs down in many cases (the 2235h uses a pro-style motor with a ring to add mass).
Well, there are different Xmax out there and most are exaggerated. JBL adds 10 up to about 45% in some cases on their datasheet specs depending on model, use (hint: car subwoofers) and year, compared to the strict mathematical value. I admit my Xmax might be wrong because of rounding errors, I got mine data from a german forum. Do you have any information about the winding height and pole plate thickness?
The Celestion is ofcourse not a direct replacement but it's so much better in the midrange than the JBL which got a lot resonances and delayed decay. I don't mind the enclosure size and I've got already 4 150l 15" subs, so if it's not deep enough, I could use them. But I doubt I'd need that, 40-45Hz is low enough - for me - and I would use/plan them fullrange anyway.
Yes, a rising response is normal. I don't claim Celestions (or most other manufacturers) published measurements are reliable, there are a lot of professionally made measurements available, i.e. from the magazines Hobby Hifi and Klang+Ton (German). For me the most important thing is the extremely clean midrange regarding response, decay and distortion and that's something exceptional, even three to four times more expensive drivers rarely deliver all three combined - and that's what makes the Celestion a rather good choice for a 2-Way 15" speaker.
That's the wrong way to go for a 2 way. Added weight is poison for the midrange. You don't need it to reduce the level either because if it's not that deep, you don't have to reduce the level that far. In the bass the level is lower anyway, that's the reason you can't use the full spl of a potent 15" if you're using it fullrange. In the midrange the reduction done by the crossover.
The 2235h does not have a ring to add mass in the recone kit - which would be required to be included. I think you confuse it with the centering alignment helping shims. These shims/tube is inserted over the pole core to grant perfect alignment while the reconing and is removed in the second last step before the dustcap is glued on.
See, I'm not saying the 2235h is a bad driver, it's built for a different purpose, to allow smaller enclosures and go deeper. To me, the midrange is much more important and I prefer to trade some Hz of the lower end for a higher spl, 2dB more is a lot of power if I might get the weird idea I have to go tube or something like that.
And if it's not deep enough I could still use my subs. My preferences and priorities maybe are just weighted differently.
The first thing which hurts the midrange is, it is much harder to move and control the mass. The smaller the mass, the easier it's possible to follow the signal. High mass in motion is harder to stop and to accelerate, mass - once in motion - 'wants' to stay in motion. That leads to overswing and the electromagnetic force influences the amplifier 'back', like a motor can also be used as a generator. The amp has to correct more etc. Or simply put: If you give the driver a signal and switch it off, it swings wider and longer till it comes to a halt again.
In a certain range the higher mass can be compensated if you improve the motor (magnet). That isn't possible here, you can't make the magnet stronger, rather the opposite is the case: The mass/motor ratio gets worse, the suspension is 'softer for the higher mass'. As a result small movements become swallowed by the mass inertia, less resolution, less fine-dynamic, fine impulses get weaker, details get lost.
The 2235H got already a quite high mass for a 2 way/mid-bass 15" driver, ca. 125g while the TF1525 only has to move 75g. The motor is in roughly the same ratio to the mass on both drivers (BL 20,5 vs. 13,5) but for the higher mass and, additionally, the lower Vas (needed volume), the suspension has to be stiffer also and that means there's also more losses and higher forces needed for initial movement. Last but not least, the JBL got a noticable higher inductance than the Celestion (1,2mH vs. 0,82mH), that translates into a higher filter effect on higher frequencies, higher impedance and less power from the amplifier.
None of this for itself is a huge deficite but all in all it does add up. I know, it's not the same but compare it with tweeters, the lighter the membrane, the more details they can reproduce, that's why foils (ribbon, magnetostats, electrostats, plasma-tweeters etc.) got such a good resolution and that's why they aim for alway less moving mass, stiffer materials (metal domes, ceramics, diamond, beryllium) to make the membrane lighter while keeping or even improving the stability.
This is frustrating, I like making pretty waveguides lol
But if I'm understanding this correctly, the impulse response will be excellent when the frequency response is excellent. Because they're related.
IE, when I saw excellent impulse response in the JBL waveguides and LeCleach horns, what I was actually seeing is the narrower beamwdith of the devices. IE, if your horn flattens the on-axis response via geometry, it's impulse will measure better.
Therefore, one could get excellent impulse response from an OS waveguide by simply equalizing it.
With that out of the way, what you'd want to do next is 'clean up' the impulse response. This can be attacked on a number of fronts:
1) improved phase plug design
2) treatment of reflections at the throat and at the mouth
3) things like the HOM plug which absorb higher order modes
In other words, once you were done with this process of optimization, you'd wind up with something that looked a lot like your loudspeakers
I know this post will come off as kind of "fan-boy-ish" but I think that it's difficult to appreciate good waveguide design unless you actually understand how loudspeakers work. This is something that I've noticed when I go to audio shows; you'll see some rich dentist look at a set of speakers, and then the salesperson will get into their spiel about what makes the speakers special. But I've long noticed that the salespeople are frequently bringing up features that are completely irrelevant to sound reproduction. I've been building and studying loudspeakers for nearly two decades and I'm still learning things.
Basically, if I've put in this much effort, and I still misunderstand things like impulse response, the average dentist who has $10K to spend on loudspeakers will certainly be prone to getting swindled by made up BS.
I was afraid that is what you were going to say.
I'm not going to elaborate on why it is not true other than to say that a mass ring of non conductive material placed at the apex of the cone and voice coil would act just like a lumped mass and neither the signal nor the sound radiation would know that it was any different than just a heavier cone or voice coil. This idea that mass limits "resolution" is true only to the extent that massive drivers operate over limited frequency bandwidths. Unless you are talking about full range drivers, which from context I assume that you are not, the mass of the woofers moving parts is basically irrelevant - as long as the motor has enough BL to control it, which in this case it does.
Yes, they are intimately related. Give me one and I know the other and visa versa.
The limitations come from directivity control because I can only correct the impulse/frequency response with DSP at a single point. Unless all points in the coverage area are similar then correcting this one point does NOT correct the total sound. Hence my claim that acoustical problems cannot be corrected with EQ.
If the waveguide is CD then correcting its response at one point will correct it everywhere. A point that John Eargle made when he did the 4430.
My favorite explanation of this:
Woofer Speed
"Mass isn't the problem - inductance is. So if you want faster transient response, ignore that moving mass parameter that some manufacturers push - look at the inductance! And if they don't list the inductance, ask yourself why - is there something they don't want to show? Inductance is the key to driver transient response - ask for it when transient response comes up!"
And that's what I was afraid you were going to say.
You (and others) obviously did not read my entire post. I actually did say that the motor can compensate the mass. But only to a certain degree and that's the main thing, because with more mass you have to use stiffer and more lossy suspension. That's why it's only compensates it to a degree. But let's ignore that for now.
The JBL got an impressive strong motor, that's true, a BL of 20.5 is actually very good. But the motor alone is irrelevant (just the same like you said about the mass), it's the ratio of motor to mass which matters! And if you look at the numbers, it becomes quite a different picture:
JBL: 125g Mms (other sources speak of about 150g (and a Xmax absolute of 8,38mm which translates in +/-4,19mm) but let's stay with the 125g) and a quite impressive BL of 20.5. The Celestion got a quite medicore BL of 13.5 but just 75g moving mass. Yes, sure, everyone screams 'how could you?' and 'the JBL got a much stronger motor!'. Don't let you confuse with high absolute specs, let's see, how's the ratio, how much 'BL for each g moving mass' do we have?
JBL 2235H: 20.5/125g = 0,164 BL/g
Celestion TF1525: 13.5/75g = 0,18 BL/g
And that shows, the Celestion actually does have more motor per gram moving mass.
And now please remember, in addition, the JBL got higher losses and higher inductance on the downside too!
Last edited:
Fair point, and agreed (BUT, see below).
Actually, to be precise, it's the ratio of effective electrical force factor to mass that determines electrical damping, i.e. the ability of the motor assembly to control the moving mass (*):
(BL^2/Re)/Mms = 2*Pi*(Fs/Qes)
(*) Incidentally, electrical damping also determines low frequency extension, so if you actually want bass out of your woofer, you can't push electrical damping too high...
For the JBL 2235H, we have Fs/Qes = 20/0.28 = 71
For the Celestion TF1525, we have Fs/Qes = 47/0.66 = 71
Lo and behold, they are equivalent, in terms of electrical damping!
Inductance is irrelevant when Woofers are used as intended, i.e. for low frequencies, since there will typically be much larger inductors in series.
As to higher mechanical losses, I agree these are important in determining a Woofer's mechanical damping, and hence its ability to reproduce micro-dynamics and detail at low volumes (the lower mechanical damping, the better).
However, mechanical damping is actually measured as:
Rms/Mms = 2*Pi*(Fs/Qms)
For the JBL 2235H, we have Fs/Qms = 20/2.5 = 8.0
For the Celestion TF1525, we have Fs/Qms = 47/5.9 = 8.0
What?! They are actually equivalent in terms of mechanical damping too!
And, incidentally, neither is particularly good in this regard.
For instance, a TAD T-1601a has Fs/Qms = 27/6.8 = 4.1 which is much better.
Last edited:
In this case my Beymas with a Qms of 21 and an fs of 42 Hz must be king ! Although I doubt that Qms value to be correct.
To stay on topic: I guess it is actually the low frequencies where the 2230 monitor sounds nice and much less the mid/high part. I have a pair of original 2344 Horns with 2424 drivers and I can say that a more modern combination with a Failtal HF146 on a 18S XT1464 waveguide sounds much cleaner than the old JBL combination. The cabinets with the JBL drivers are currently used as stands for the ones with the Faital/18s/Beyma combination.
Regards
Charles
To stay on topic: I guess it is actually the low frequencies where the 2230 monitor sounds nice and much less the mid/high part. I have a pair of original 2344 Horns with 2424 drivers and I can say that a more modern combination with a Failtal HF146 on a 18S XT1464 waveguide sounds much cleaner than the old JBL combination. The cabinets with the JBL drivers are currently used as stands for the ones with the Faital/18s/Beyma combination.
Regards
Charles