Help me choose a horn and driver

[iansr]

So here I am with a blank sheet of paper attempting to design my ultimate 2 way. I like the presentation of dipoles so on the bottom I will have 2 x 15” woofers on OB per speaker; the AE dipole woofers will certainly be on the shortlist. Up top I want a horn and that’s where I’d like some input.

Things I value and would ideally want to have:
1. A nice linear on-axis FR between 800 or 900Hz (or lower) and 15KHz, or thereabouts.
2. A related off axis FR that Mr Toole would regard as pretty good.
3. Open and transparent sound with a holographic phantom image.
4. Bi-amping via an analogue active crossover. ( I’ve just bought the Nelson Pass active crossover kit from the DIYA store.). I do NOT want to use DSP.

Oh yes and I’d like them to sound as good as the Vox Olympians. No? OK forget that one then ;)

So if I’m to believe what I’ve read then maybe the ultimate would be the TD4003 plus TH4003, but that would cost me deep in the purse (who spotted the Fairport Convention reference there?) and I’m wondering if I could get 90% of the way there at lower cost? Maybe a 2450J-Truextent with a A290FL, or a 2450SL - Truextent with a H9800?

So please pitch in guys, I want to hear your thoughts and recommendations :)

[Robh3606]

I would go with a large format 1.5" throat driver like a 2452 and one of the PT wave-guides. I use PTH1010 and they are basically a newer version of the 2344. They image very well and it's not difficult to do the networks for them.

[Mr. Widget]

I would go with a large format 1.5" throat driver like a 2452 and one of the PT wave-guides. I haven’t messed around with the PT wave guides. Can they image as well as the vertical Array type horn? How about a large format driver on the Array type horn.


Widget

[Robh3606]

Hey Widget

Well I can only speak for the 1010 it's the only one I have used. I would say it's close and the 1400 Array probably has the edge.

Rob:)

[iansr]

I would go with a large format 1.5" throat driver like a 2452 and one of the PT wave-guides. I use PTH1010 and they are basically a newer version of the 2344. They image very well and it's not difficult to do the networks for them.
Thanks for responding. By coincidence I’ve just come across an old post of yours showing the FR of a 476mg on a PTH1010. It was damn impressive; almost flat from 1k to 20khz!

[iansr]

Hey Widget

Well I can only speak for the 1010 it's the only one I have used. I would say it's close and the 1400 Array probably has the edge.

Rob:)

i have to say I’m not keen on the aesthetics of the vertical orientation horns.

[Mr. Widget]

i have to say I’m not keen on the aesthetics of the vertical orientation horns.They are not my favorite aesthetically either, but they are currently in my living room sounding amazingly good!


Widget

[Robh3606]

i have to say I’m not keen on the aesthetics of the vertical orientation horns.


Well you get use to them and they sure sound good!

Rob:)

[mortron]

how hard are the vertical horns to come by? quite pricey?

[grumpy]

The H9800 horn mentioned in the initial post works well vertically... if you have access.
Also, SAM1HF can be used.

[iansr]

The 1400 array is certainly appealing because of it being a mid /tweeter combo. Just wondering if anything can be done about the look. Showing the depth of my ignorance here but would it be sonically detrimental to mount it on a OB ? ( I think that might make it look less odd.)

[withTarragon]

The 1400 array is certainly appealing because of it being a mid /tweeter combo. Just wondering if anything can be done about the look. Showing the depth of my ignorance here but would it be sonically detrimental to mount it on a OB ? ( I think that might make it look less odd.)

I am not sure what your skill set is (or with the help of a friend). Could you build a "cage" with the same footprint as the bass bin. Then cover the cage with grille cloth. Probably a number of ways to tackle this.
Good luck,
-Tom

[Odd]

There are several members who have built it into a box.

http://www.audioheritage.org/vbulletin/showthread.php?34265-Jbl-Array/page2

http://www.audioheritage.org/vbulletin/showthread.php?33151-DIY-1200-Array

[ivica]

So here I am with a blank sheet of paper attempting to design my ultimate 2 way. I like the presentation of dipoles so on the bottom I will have 2 x 15” woofers on OB per speaker; the AE dipole woofers will certainly be on the shortlist. Up top I want a horn and that’s where I’d like some input.

Things I value and would ideally want to have:
1. A nice linear on-axis FR between 800 or 900Hz (or lower) and 15KHz, or thereabouts.
2. A related off axis FR that Mr Toole would regard as pretty good.
3. Open and transparent sound with a holographic phantom image.
4. Bi-amping via an analogue active crossover. ( I’ve just bought the Nelson Pass active crossover kit from the DIYA store.). I do NOT want to use DSP.

Oh yes and I’d like them to sound as good as the Vox Olympians. No? OK forget that one then ;)

So if I’m to believe what I’ve read then maybe the ultimate would be the TD4003 plus TH4003, but that would cost me deep in the purse (who spotted the Fairport Convention reference there?) and I’m wondering if I could get 90% of the way there at lower cost? Maybe a 2450J-Truextent with a A290FL, or a 2450SL - Truextent with a H9800?

So please pitch in guys, I want to hear your thoughts and recommendations :)

Hi inasr,

I would think about JBL PT-H95HF with 2450-1.5 driver with 2441AL or 2445 diaphragm. I think it can be used from 800Hz

https://www.jblpro.com/ProductAttachments/tn_v1n31.pdf
(https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=2ahUKEwijgNXW98zeAhUIFywKHeekBzIQFjAAegQICBAC&url=https%3A%2F%2Fwww.jblpro.com%2FProductAttachme nts%2Ftn_v1n31.pdf&usg=AOvVaw3G7TFaH4za_6uwgwJM1zf7)

JBL Pro Parts 443495-001: WAVEGUIDE, PT-H95HF-1

Product # JBL Pro Parts 443495-001
look at:
http://www.audioheritage.org/vbulletin/showthread.php?34470-Replace-2265HPL-with-2226H&p=349176&viewfull=1#post349176

http://www.audioheritage.org/vbulletin/showthread.php?36528-Is-PT-F95HF-a-meaningful-sonic-upgrade-from-2332&p=371066&viewfull=1#post371066 (http://www.audioheritage.org/vbulletin/showthread.php?34470-Replace-2265HPL-with-2226H&p=349176&viewfull=1#post349176)



If over 1.2kHz is acceptable , then STX825 with 2447 drives can give very good results

Regrads
ivica

[iansr]

Hi inasr,

I would think about JBL PT-H95HF with 2450-1.5 driver with 2441AL or 2445 diaphragm. I think it can be used from 800Hz

https://www.jblpro.com/ProductAttachments/tn_v1n31.pdf
(https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=2ahUKEwijgNXW98zeAhUIFywKHeekBzIQFjAAegQICBAC&url=https%3A%2F%2Fwww.jblpro.com%2FProductAttachme nts%2Ftn_v1n31.pdf&usg=AOvVaw3G7TFaH4za_6uwgwJM1zf7)

JBL Pro Parts 443495-001: WAVEGUIDE, PT-H95HF-1

Product # JBL Pro Parts 443495-001
look at:
http://www.audioheritage.org/vbulletin/showthread.php?34470-Replace-2265HPL-with-2226H&p=349176&viewfull=1#post349176

(http://www.audioheritage.org/vbulletin/showthread.php?34470-Replace-2265HPL-with-2226H&p=349176&viewfull=1#post349176)http://www.audioheritage.org/vbulletin/showthread.php?36528-Is-PT-F95HF-a-meaningful-sonic-upgrade-from-2332&p=371066&viewfull=1#post371066



If over 1.2kHz is acceptable , then STX825 with 2447 drives can give very good results

Regrads
ivica



Many thanks ivaca, those links are really helpful. Any particular reason you suggest the H95HF rather than the H1010. Also I’m starting to think a 3 way is worth considering, in which case the K95MH looks interesting!

[ivica]

Many thanks ivaca, those links are really helpful. Any particular reason you suggest the H95HF rather than the H1010. Also I’m starting to think a 3 way is worth considering, in which case the K95MH looks interesting!

Hi iansr,

I have no experience with H1010, but for me, PTH95HF is quite good with the mentioned driver.

regards
ivica

[iansr]

@robh3606
i just came across a post from you on WBF where you said :

The system may be DSP but the M2 waveguide is very good and works very well with a passive crossover. Here is a raw measurement of a 476Mg on the M2 and a Leap crossover solution based on Greg Timbers "Key Filter" design. With this compression driver all you need is a series filter and HF bypass and a single parallel notch filter to do the compensation to flatten the response on the waveguide. This waveguide is actually very similar to the 2344 and PTH1010.

Thats really interesting! The M2 DSP has such a high number of EQ points that I thought it would be really dificult to create a passive crossover for either the woofer or the CD/waveguide. Does using the 476Mg makes it less complex?

[Robh3606]

Thats really interesting! The M2 DSP has such a high number of EQ points that I thought it would be really dificult to create a passive crossover for either the woofer or the CD/waveguide. Does using the 476Mg makes it less complex?

Well I guess it depends on the compression driver. Like any good CD horn it's un-equalized on axis output should match the power response of the driver. Tale a look at the difference between the stock M2 compression driver and the 476Mg. The 476 is much better behaved IMHO as are a 2451/52. I am also not going for ruler flat response There is some ripple between 2-4k that you could DSP out but I wouldn't bother.

Rob:)

[Earl K]

http://www.audioheritage.org/vbulletin/attachment.php?attachmentid=82769&stc=1&d=1542030734

That certainly looks real nice Rob.

Reality-wise, can the OP ( Ian ? I suppose ) even buy these 476Mg drivers?

Or are they embargoed ( only to existing owners of speakers that contain this driver ) ?

:)

[iansr]

Thanks Rob. How do you think a 2452 + Truextent compares to a 476Mg or even a 476Be?

[marco_gea]

Thanks Rob. How do you think a 2452 + Truextent compares to a 476Mg or even a 476Be?

Actually, a 2450SL (http://www.jblproservice.com/pdf/Discrete%20Sales%20Models/2450SL.pdf) + Truextent BeX4008 (https://materion.com/products/beryllium-products/truextent-bex-diaphragm-assemblies) diaphragm would be the best candidate.

The 2451 > 2452 > 2453 are progressively cheaper versions of the original 2450 (https://www.hifisentralen.no/forumet/attachments/diy-og-utvikling-ha-yttalere-forsterkere-etc/450193d1500915872-jbl-2450sl-2451-sl-hva-er-forskjellen-comp-driver-front.jpg)s...

Please note that due to the Truextent's PEEK (i.e., plastic polymer) surround, though, the top octave (>10kHz) will be more attenuated (but without any peaking, which IMHO is a good thing).

[Robh3606]

Thanks Rob. How do you think a 2452 + Truextent compares to a 476Mg or even a 476Be?

Hello iansr
Well that depends on how much roll off there will be in the last octave. The Be sounds great but it also rolls off with the Magnesium and Ti diaphragm's they actually roll off but peak a bit out in the 10-20K region. That peak is not as clean like the Be. That peak actually makes them easier to do the network for if you are not planning on using a Bat Killer out to 40K. I was fine with the roll off using a 2344/2425 again OK with a 2435/PTH1010 you may not be. It's a personal choice. Take a look the attached plots.



Reality-wise, can the OP ( Ian ? I suppose ) even buy these 476Mg drivers?

Hello Earl

I doubt it I was lucky to get mine that's why I was suggesting another available driver with a similar response.

Rob:)






Rob

[iansr]

I’m strongly attracted to using the M2 waveguide. As far as the Driver is concerned, I’m considering Marco’s suggestion of a 2450SL + Be, but the Radian 951 + Be is also very tempting, and cheaper! Guido reckons the Radian is as good as the 476Be.

[Robh3606]

Hello iansr

Are you going active or passive?? Active it really doesn't matter but if are trying to go passive make sure you are comfortable with the HF roll off using the Be drivers. Watch the mounting bolt pattern on the Radian. You are probably going to need an adaptor plate going from 1.4-1.5 throat diameter.

Rob:)

[iansr]

Hello iansr

Are you going active or passive?? Active it really doesn't matter but if are trying to go passive make sure you are comfortable with the HF roll off using the Be drivers. Watch the mounting bolt pattern on the Radian. You are probably going to need an adaptor plate going from 1.4-1.5 throat diameter.

Rob:)
I’m going to go active, but analogue not DSP. I was going to ask about the 1.4” of the Radian. Is an adaptor plate likely to adversely affect the performance do you think?

[marco_gea]

if are trying to go passive make sure you are comfortable with the HF roll off using the Be drivers.
Rob:)

That, or add a supertweeter. In my experience, the latter invariably leads to superior results, no matter what (compared to using compression drivers that "peak" in the top octave - as such peaking is always linked to parasitic resonances). Others may and probably will disagree, and that's all right with me ;-)

Marco

[marco_gea]

I’m going to go active, but analogue not DSP.

Beware. With direct radiating Woofer(s) & a compression driver+horn, you really need asymmetrical LP and HP electrical slopes (and possibly delay on the HF channel). Make sure that your analogue active crossover of choice provides the necessary flexibility. Textbook-standard symmetrical slopes (e.g., Butt3, LR4, etc.) will NOT lead to optimal results, ever!

This is actually a case in which going passive may provide MORE benefits and flexibility, not LESS.

Otherwise, of course, there is the now in-vogue "brute force" approach represented by DSP with infinitely tweakable slopes and multiple PEQs, etc.

M.

[marco_gea]

I was going to ask about the 1.4” of the Radian. Is an adaptor plate likely to adversely affect the performance do you think?

Not the adapter plate per se. But the (potential) double mis-match in wall angles may well do so.

[Robh3606]

Hello iansr

You ever do anything like this before?? Analog is a great way to go but it's going to greatly complicate things. I would also try to avoid an adaptor plate for a throat transition. If you plan on using a 1.5" horn use a 1.5" driver.


Actually, a 2450SL (http://www.jblproservice.com/pdf/Discrete%20Sales%20Models/2450SL.pdf) + Truextent BeX4008 (https://materion.com/products/beryllium-products/truextent-bex-diaphragm-assemblies) diaphragm would be the best candidate.


Hello Marco

The 2450 is a 2" throat can't use it with the M2 a throatless 1.5" driver is your best bet


Rob:)

[grumpy]

Oddly enough, the 2450SL is a 1.5" exit driver (vs 2450H/J, which are not)).
Definitely watch out for bolt hole pattern differences though, between the various 1.5" exit drivers.

[Robh3606]

Oddly enough, the 2450SL is a 1.5" exit driver (vs 2450H/J, which are not)).

Thanks didn't know that so it's a throatless 2450?? . Odd then why do a 2451??? Did a 2450SL morph into a 2451???

[grumpy]

Hard to keep track of the variations of back-caps, magnets, ribbed/non-ribbed dias, bolt patterns, coated/non-coated, ...
Seems like someone did a photo variation comparo awhile back... subwoof maybe?

[Odd]

82780

82781

82782

[Earl K]

I think these iwata horns look particularly cool ( & apparently sound good too ).

I can't say how they'll work out with your OB experiment, but I'd sure give them a go.

Like Marco says, add a tweeter ( above 10K works for me ).

:)

[iansr]

Yes certainly looks like the 951 would be a bad idea. Hats off to Guido, even though it cost him a potential sale he advised me not to put the 951 on a 1.5 inch horn. Now that’s integrity.

[iansr]

I think these iwata horns look particularly cool ( & apparently sound good too ).

I can't say how they'll work out with your OB experiment, but I'd sure give them a go.

Like Marco says, add a tweeter ( above 10K works for me ).

:)
Those are really gorgeous. You could be fooled into thinking they are modern sculptures.

[pos]

If you plan on using the M2 waveguide, here is an interesting comparison:
http://www.audioheritage.org/vbulletin/showthread.php?38697-Measuring-data-for-D2-(2430k)-on-M2-waveguide-wanted&p=395445&viewfull=1#post395445

Definitly easier to EQ than a D2430K :)
A 2450SL with original diaphragm might be even easier.

Depending on your amplifier choie (well, I already know your choices there ;)) you might prefer a 16 ohm version, to lower hiss and distortion.

[pos]

Beware. With direct radiating Woofer(s) & a compression driver+horn, you really need asymmetrical LP and HP electrical slopes (and possibly delay on the HF channel). Make sure that your analogue active crossover of choice provides the necessary flexibility. Textbook-standard symmetrical slopes (e.g., Butt3, LR4, etc.) will NOT lead to optimal results, ever!

This is actually a case in which going passive may provide MORE benefits and flexibility, not LESS.

Otherwise, of course, there is the now in-vogue "brute force" approach represented by DSP with infinitely tweakable slopes and multiple PEQs, etc.

M.
100% agree with Marco here.
Crossover slopes are critical, and require specific tricks when done with analog means without resorting to delay for alignment.

Even if the final plan is to use an analog active crossover, it might prove useful to at least use a digital crossover during the design phase.
Of course you would have to make sure you are only using things that can actually be reproduced with your analog active crossover (eg no delays...), but at least it will let you try things.
Using a simulation software would be another option.

[iansr]

If you plan on using the M2 waveguide, here is an interesting comparison:
http://www.audioheritage.org/vbulletin/showthread.php?38697-Measuring-data-for-D2-(2430k)-on-M2-waveguide-wanted&p=395445&viewfull=1#post395445

Definitly easier to EQ than a D2430K :)
A 2450SL with original diaphragm might be even easier.

Depending on your amplifier choie (well, I already know your choices there ;)) you might prefer a 16 ohm version, to lower hiss and distortion.
Thanks Thomas. Amongst their many virtues the Neurochrome amps are dead quiet.

[iansr]

100% agree with Marco here.
Crossover slopes are critical, and require specific tricks when done with analog means without resorting to delay for alignment.

Even if the final plan is to use an analog active crossover, it might prove useful to at least use a digital crossover during the design phase.
Of course you would have to make sure you are only using things that can actually be reproduced with your analog active crossover (eg no delays...), but at least it will let you try things.
Using a simulation software would be another option.

Yup, Initially using a miniDSP unit to determine the best settings is exactly what I plan to do. If I can’t reproduce them using the Pass crossover kit then I have a clever friend who can make me a custom analog crossover.

I may even end up with a hybrid crossover arrangement, part active and part passive. If I decide to go 3 way then just being passive between the CD and the supertweeter might be a good choice. That would still allow me to bi-amp . . .

[iansr]

I think these iwata horns look particularly cool ( & apparently sound good too ).

I can't say how they'll work out with your OB experiment, but I'd sure give them a go.

Like Marco says, add a tweeter ( above 10K works for me ).

:)
Earl
do you know source of the wooden horns in the second photo?

[ivica]

Hard to keep track of the variations of back-caps, magnets, ribbed/non-ribbed dias, bolt patterns, coated/non-coated, ...
Seems like someone did a photo variation comparo awhile back... subwoof maybe?

Hi grumpy,

I think 2447 and 2451 as 1.5" type drivers have the same bolt patterns, while 2450-1.5 (SL), 2452, 53 and D2430K (D2) have the same bolt patters, but all of them are 1.5-inch driver "mouth" size. Some experiments have shown that 2452, 53 would give better results with 0.5-inch thick spacer between the short horns ( such as STX825 horn) and the drivers (almost the same as built-in the 2447 driver).

Regards
Ivica

[grumpy]

... worth keeping in mind as a possible area of optimization for anyone contemplating such a build, such at the original poster.
I'm sure a simple link to the experiment parameters/results would be appreciated.

[richluvsound]

... worth keeping in mind as a possible area of optimization for anyone contemplating such a build, such at the original poster.
I'm sure a simple link to the experiment parameters/results would be appreciated.


3-way JBL project using 1500al , 2450/Be , 045Be .... GEA/Yuichi Horns , 120 litre . tuned to 32 Htz

[pos]

Those 045 horns are so cute! :D

[Earl K]

Earl
do you know source of the wooden horns in the second photo?


Sorry no, I don't know who made ( or sold ) those great looking horns.

The fiberglass ones are available from Autotech Horns (http://horns-diy.pl/en/horns/iwata/iwata-300/) ( I believe it is ).

:)

[Dr.db]

3-way JBL project using 1500al , 2450/Be , 045Be .... GEA/Yuichi Horns , 120 litre . tuned to 32 Htz


Do you have a picture of the fully assembled and finished speaker? :)