DIY 1200 Array

[4313B]

4313B, is it your contention that the 2206H is superior midrange to the drivers used in the Array series?

Yes.

But I agree with Greg's thinking that the fifteen would be the best choice. Case in point, the S4700, S4800, 4338, 4365, K2-S9900, DD65000, DD66000, DD67000. "If I had it all to do over again (with respect to the 1200 Array) I would just run the 1500AL right up to the horn and leave out the twelve."

[Robh3606]

With a proper termination on the horn, it'd have a more consistent off axis pattern, rather than as-shown, which has a power response peak at about 900 Hz. The horn is clearly not controlling pattern to 750hz in the horizontal, hence my gripe.

Help me out here. Take a look at the directivity curves for the system. The first is on axis the second is first reflections. Why doesn't the pattern flip show up here?? You can see the woofers directivity increasing then at the crossover point you can see the transition to the horn. What I don't see it a blip in the curve at 900 hz. Shouldn't there be one if the horns directivity is abruptly changing??

Rob

[DavidF]

Attachment 59268



Hi Rob:

If you look at the response family just above 1k, you'll see that the dispersion rapidly increases with falling frequency, and that there's a peak and a dip in the off-axis response family below 1khz, a telltale sign of a poorly terminated horn. With a proper termination on the horn, it'd have a more consistent off axis pattern, rather than as-shown, which has a power response peak at about 900 Hz. The horn is clearly not controlling pattern to 750hz in the horizontal, hence my gripe. It's not as bad as some other horns but given the cost of these systems and the quality of them in other respects, I hate to see such an easy fix being left out.

Best,
J

J, I am thinking I would want to eliminate the woofer as a source of resonance in that 500-800 Hz lump. As to whether the horn could benefit from a different termination I would hate to loose the soundstage/imaging abilities by tampering with the radiation pattern if that negates some of what I am liking.

[grumpy]

I think part of the apparent difference is one plot set baselines the center, zero-axis response, showing differential measurements, while the JBL measurements do not. That said this system was of the better ones JBL has tested on their "spin-o-rama" testbed. Further improvements would be cool to see, and icing on the cake.

[ivica]

Attachment 59268



Hi Rob:

If you look at the response family just above 1k, you'll see that the dispersion rapidly increases with falling frequency, and that there's a peak and a dip in the off-axis response family below 1khz, a telltale sign of a poorly terminated horn. With a proper termination on the horn, it'd have a more consistent off axis pattern, rather than as-shown, which has a power response peak at about 900 Hz. The horn is clearly not controlling pattern to 750hz in the horizontal, hence my gripe. It's not as bad as some other horns but given the cost of these systems and the quality of them in other respects, I hate to see such an easy fix being left out.

Best,
J

Hi Rob,

May be I do not understand attached FR response but something not expectable happened around 6kHz i off-axis response. Is it UHF driver influence, or some kind of driver-horn 'ringing' phenomena.

Have You ever compare array 1400 dispersion to (2312 or 2307) 2311+2308 with appropriated ("light year behind" drivers). I was amazed how wide horizontal dispersion 2308 lenses 'had produced'....


Regards
Ivica

[ivica]

I know you are directing this at Rob because you and I have a problem, ........

Hi 4313B,
My English is fare to be good to understand your words, but rally I have no idea what kind of problem WE have ?
I was, just, wondering what would be the reason for 1400 horn-driver behavior off axis around 4-8 kHz, as can be seen on the attached horizontal FR dispersion responses I can say may be more then +/-6dB, and is visible on the "cumulative spectral-decay plot" on the same presentation

http://www.audioheritage.org/vbullet...8&d=1371763519

not caring about what kind of dispersion the 2311+2308 has because you simply want it turned off[/B]. That is the peril in listening to anything Greg has done more recently compared to what he did thirty plus years ago.

And please don't make it a point to make fun of one of Greg's favorite phrases. He has earned it and he owns it.

Well I can understand that, may be, that such driver+lenses behavior is(WAS) the mean reasons why JBL 4333, 4343, 4345, 4350 ,4355 were "rejected" from the mean audio monitoring studios in the 'favor of' ...??????......not to mention what JBL studio monitors have made JBL name in studio-monitor branch.

I can accept that there are more evolution in improving audio reproduction in last thirty+ ( or better to say sixty+ ) years, where renowned Mister Greg Timbers, an engineer of great honor and respect, has significant and unforgettable place, but sometime 'over exaggeration' can blear the real life.

Regards
Ivica

[ivica]

...
If you turn the 045 off before taking system measurements you end up with a nice enough plot. .....

Hi 4313B,

It would nice if dispersion measurements of 1400 horn-driver (with 045 turned off) can be presented here as an "addendum to"
your respectable work such as:

http://www.audioheritage.org/vbullet...l=1#post335728



Regards
Ivica

[Robh3606]

I think part of the apparent difference is one plot set baselines the center, zero-axis response, showing differential measurements, while the JBL measurements do not.

Hello Grumpy

That makes sense, I was wondering about if they were normalized or not. I knew his vertical plots were I didn't realize that the polars were as well.

Rob

[Robh3606]

Hello Ivica

I went back to some of my measurements of my 1400 clones and ran some plots FWIW. You might find them interesting. First the system waterfall, then the 435be and last the 045be. These were all measured with the network in place. They are not raw driver measurements.

Rob

[Robh3606]

The 045be

[badman]

Help me out here. Take a look at the directivity curves for the system. The first is on axis the second is first reflections. Why doesn't the pattern flip show up here?? You can see the woofers directivity increasing then at the crossover point you can see the transition to the horn. What I don't see it a blip in the curve at 900 hz. Shouldn't there be one if the horns directivity is abruptly changing??

Rob

No way for me to know- but the Stereophile measurement quite clearly shows the off axis peak and dip typical of a horn below cutoff. The slower transition of a bigger/better damped termination doesn't exhibit this behavior, as the "cutoff" is a mild transition to 2pi then 4pi radiation, as opposed to the abrupt change of a freestanding, hard-edged horn.

It's just a variant on baffle edge diffraction, but with the baffle being a horn. The two manifest themselves quite similarly, and are something I'm a bit of a stickler for, since I played with it and made myself sensitive to the effect. Some listeners might not notice, or even find it appealing, but it would manifest itself primarily as increasing vocal as one moves off axis, compared to a better termination which would be more consistent between the axial and off-axis tonal balance. Effectively it's like a 3dB peak right in the midrange, if you're anywhere but on-axis, in a relatively dead (or open) room. In normal rooms with a non-axial listening position it's effectively a midrange peak (based upon the stereophile measurements as well as my experience with this type of horn below cutoff).

The imaging wouldn't be negatively affected by an adjustment to solve this, as the wavelaunch for higher frequencies is effectively further back in the horn- they don't "see" the termination as directionality has already been imposed. It's almost exclusively around cutoff that this effect is to be expected. My typical cutoffs for horns is a little bit higher than some others describe, I would never call the array horn a 750Hz device.

<edit>
I should add that a friend who added foam to his 1400s a la this article:


http://www.enjoythemusic.com/diy/041..._the_mouth.htm

and his subjective result was very positive.

[ivica]

Hello Ivica

I went back to some of my measurements of my 1400 clones and ran some plots FWIW. You might find them interesting. First the system waterfall, then the 435be and last the 045be. These were all measured with the network in place. They are not raw driver measurements.

Rob

Hello Rob,

Thank You for the detailed measurements presented here.
Looking at the measurements and the data shown in the previous posts:

http://www.audioheritage.org/vbullet...l=1#post335677

http://www.audioheritage.org/vbullet...l=1#post335666


it seems to me that around 3kHz some kind of 'ringing' is present (can be seen with 435 or 476 drivers too). As wave length of about 3kHz is about 0.12m ( half is about 6cm and that seems to correspond to the distance from the horn throat to the diffraction slot in the horn).
Such diffraction slot may be allow so good horn dispersion in the off-axis plane perpendicular to the slot 'direction'.
Some of "the old school" acoustic fans said that they do not like the 'sound quality' of horn with slots (examples: 238x, 2344), but here on H1400 and on all H4338 H4365 H9800 H9900 H66000 horns, horn wall expansion after slot are not so 'abrupt' as in 238x or 2344, and that is may be the reason why lot of audiophiles ( I have to mention our members 4313B, renowned Mr. GT,....) enjoy in listening H1400.

Regards
Ivica

[JuniorJBL]

Here is a link to the data from GT's speakers.

http://www.audioheritage.org/vbullet...l=1#post350106

[Champster]

Great looking system and very interesting thread.

Where did you source your Be CDs?

Thanks,
Paul

[Allanvh5150]

Have a look here, Paul.

http://www.vueaudio.com/about/beryllium/

Allan.