6 Db slope filter for 2441 ???????

[Bruno GINARD]

Hi all

Does anybody ever tried a 6 Db/ slope filter to use a 2441 around 500 Hz in a DOMESTICAL application ?


In spite of the risque of dammaging with high power ( and that's not the case at home....) i think it'll be better for phase rotation..............

Thank you for your answer if you have tested this configuration

[spkrman57]

Even with only 5 watts or less(tubed), I usually start at 1 octave higher than the drivers recommended crossover.

Even with not damaging the driver/diaphram, I have read from others that distortion is high in the lower ranges.

Earl uses Altec 1.4" drivers(as I do also) and he has some novel ideas on how to use several 1st order crossovers together but not have the compounded order result.

Search posts for "Edgarhorn" and he replied several times with his ideas!

Regards, Ron

[Mr. Widget]

Does anybody ever tried a 6 Db/ slope filter to use a 2441 around 500 Hz in a DOMESTICAL application ?

In a home application playing reasonable SPLs I think you could get away with it without damage.

That said, I doubt you will find a horn that will work well in that application. From some recent experimenting I've done it seems that much of the "honk" and harshness that horns are famous for stems from the fact that we use horns too close to their cut off... even with a steep slope they tend to sound better when you "de-rate" the horns... for example use a 500Hz horn at 800Hz and an 800Hz horn at 1200Hz... This is actually an extension of what JBL has been doing if you look at their line of horns historically... if you take the HL91/2307 initially they recommended it's use to 500Hz, then 800Hz and ultimately 1200Hz. For the 2397 they originally used it in the Cinetron system at 500Hz, years later when they offered it as a component they suggested an 800Hz cut off, and in my experience it is best above 1KHz. The more recent JBL horns that I have heard a Zilch's house all sound like they might be improved if they were crossed over a half to a full octave higher... what may measure fine or be acceptable in a PA speaker isn't necessarily the ideal in a home system used primarily for critical listening.

I suppose if you are interested in using a first order network and want superlative sound you could try a large tractrix or some other large horn designed for 350Hz or there abouts... that would hopefully take care of the horn issue... the distortion caused by pushing the driver that low might become an issue though... I personally have never tried it so I have no idea how they perform at lower frequencies


Widget

[Earl K]

From some recent experimenting I've done it seems that much of the "honk" and harshness that horns are famous for stems from the fact that we use horns too close to their cut off... even with a steep slope they tend to sound better when you "de-rate" the horns... for example use a 500Hz horn at 800Hz and an 800Hz horn at 1200Hz... This is actually an extension of what JBL has been doing if you look at their line of horns historically... if you take the HL91/2307 initially they recommended it's use to 500Hz, then 800Hz and ultimately 1200Hz.

FWIW; I'd agree with that assessment of "Best Practise Loading" for Radial, Exponential, Tractrix, True-Conical & Conical/Hybrid-Waveguides . IME, they all start to prgressively unload the diaphragm at the lower frequencies. The unloading starts & just gets worse. I don't necessarily follow this "best" practise with my exponentials / but then I run my home system with a maximum SPL limitation, always kept in mind .

OTOH; properly designed , true Constant Directivety horns ( which I suspect Bruno is not contemplating using down to 500hz ) offer a sort of builtin BrickWall Brake-effect for the loading of the lower frequencies. Ie; they seem to offer true loading right down to high side of the lowest, most significant impedance peak . At least that's my interpretation of all the various horn/driver impedance plots I've studied ..

regards

[Earl K]

Earl uses Altec 1.4" drivers(as I do also) and he has some novel ideas on how to use several 1st order crossovers together but not have the compounded order result.

Search posts for "Edgarhorn" and he replied several times with his ideas!

Some clarification is in order .


- Actually, nothing that I articulated is my own idea. I do use 2-pole hipasses where the two F3 points are set closer than 1 octave apart ( this should produce filter ringing and a lot of group delay ) . But, ( this is a big "but" ) there's always at least 8 ohms of resistance separating the capacitor and inductor so the negative effects, that the degrade transient behaviour of the filter, are minimized or avoided . It's the resistance value that dictates how much interaction one gets within the crossover area ( ie; for example, producing Butterworth style "ringing" ) . Buried somewhere in the Project May archives is information ( from Giskard ) that states that around 20 ohms of isolation between the two poles, provides "full" isolation ( no interaction between the reactive elements ). He must have run some values through a simulator to arrive at this conclusion.
- The filter approach that I use has simply been "lifted" directly from many of the JBL schematics that Don McR. so graciously put us onto ( located at the Harman Website ) back in early 2003 . Giskard continues to highlight and underscore where these schematics can be found for download ( though many here seem a bit unawares as to their real importance in the evolution of crossover design theory ). Anyways, ultimately, the 2 poles do combine ( even with the presence of some interfering resistance ) to give a full 12 db per octave of attenuation down in the "stop-band" .

- People usually "seem" to be infatuated with low order filters in order to maximize transient response ( occuring in the crossover area ) . A 2 pole Bessel filter ( or some approximation ) will give execellent transient response, as well as ultimatly providing better LF protection for the driver . Most of JBLs' newer hipass circuits ( since the early 1990s ) appear to keep this in mind ( one needs to dissect the schematics to understand this ). They are still mostly 2-pole but have mitigated Group-Delay effects in the crossover area by adopting forced separation of the 2 interacting poles ( by using resistance from the buildout resistor in the Lpad ) . A side benefit of placing the inductor within a low-impedance circuit loop is that there is a reduction in the necessary mH value for a comparable coil ( therefore, there's a cost savings ) .

- Don't ask me to produce or explain the math behind these approaches / I can't because I haven't found it in my math books .

- Re; my reference to the Tom Dawson approach of combining a LCR with a single pole hipass . His approach is quite innovative ( though I haven't tried it ). One caveat ; his series LCR ( strapped as a parallel conjugate across the load ) must be hidden ( from the load that the amp sees ) behind some inline, series resistance ( so the amplifier doesn't destabilize by sensing a low Z load at the LCRs' resonance point ) . This LCR, does its' notching down in the stop-band of the hipass , well away from the crossover point . Tom says he picks the center point for the notch , empirically .

- Bruno may want to try the Tom Dawson approach, since it appears to be closer to his stated goals of giving minimal phase rotation through the crossover area .

[Tom Loizeaux]

I'm no authority by any means, but I think a 6dB slope can only work well if you crossover at least one octave higher then recommended. In your case, that may defeat the purpose.
I've also been told that JBL compression drivers, though "rated" at 500 or 800 Hz, sound significantly better when they work above 2K.

I think this is an argument for a 4-way system that covers the low-mids with a smaller cone driver (4343,4344,4345,etc.)

Tom

[Zilch]

The more recent JBL horns that I have heard a Zilch's house all sound like they might be improved if they were crossed over a half to a full octave higher... what may measure fine or be acceptable in a PA speaker isn't necessarily the ideal in a home system used primarily for critical listening.

Yup. At Mr. Widget's suggestion, I recently loaded 18 dB/octave 1200Hz FFBREQ cards into my 5235 crossover, and all of the PT waveguides sound better.

2235H and LE14H-3 (LE14A also,) seem to play fine that high with them, though I'll need some critical listening to be certain.

I don't think it's coincidence that the best-sounding system here uses horns that load to 500 Hz (2352) crossed at 1 kHz, notwithstanding that they are constant directivity....

[frank23]

Some clarification is in order .

Now, there is something to read!

frank

[scott fitlin]

I had an experience with 1st order slopes, by accident, but I was surprised at what I heard.

When I first installed the BSS 366t, I was playing around with slopes, and I accidentally switched my horns ( 2395,s with 2441,s ) to 6db slopes@750hz, and my woofers too. I didnt realize it at first, but it sounded really sweet, and very transparent. I listened to music this way for about 3 hours, and then when I was playing with the xover, I saw what I had done, and immediately changed it. Nonetheless, even though I would never run my horns this way, it did sound really good.

I was also surprised that I didnt hear audible stress from the comp drivers, but I wasnt playing that loud.

[Bruno GINARD]

very very interesting.................

The 6 db Xover is for a "simple" Hartsfield horn, 2390 lense , and of course for the Hartsfield.

I was thinking about 6 db X over after a "test".....a3115 filter on a speaker and (very) old N400 on the other.

3115 has a high cut for the woofer with 12 db slope, and N400 only a 6 db ( an inductor)

Sound is really better , clearer, with N400.

I am going to try an "all 6 db" filter, on the woofer, the 2441 (between 600 and 800 hz with only a cap.) , and the 2405 ( a 1.5 Mf cap for a 10 khz cut off)....

The trouble i can see is the ability for the E 145 in Hartsfield's horn to go up to 500 Hz................

Last but not least, i still don't understand the value of the caps and inductors for the 2440/2441 on the N 400 Sheme................