1,5" wave guide shoot-out

[sebackman]

I'm going to do a measuring run on JBL M2, 4367 and VTX waveguides the next few days.

I will try a few different dia's in 2450SL, and 2451 cores. I'don't have an Mg dia and not sure it will take apart my other system to get to the Be's.

I will prost some findings here and if someone want the raw data just give me a shout.

Kind regards
//Rob

[Mr. Widget]

Very cool... looking forward to your discoveries.


Widget

[1audiohack]

Hello Rob;

Is the little horn a 5006815?

I just bought a couple of these to try for a high SPL super tweeter for use from about 5kHz and up. I am looking for something that screams at least to 15kHz to augment the 2451SL’s on 4338 horns. I was hoping to use D2’s on them but the PWT measurements I took last weekend show the D2 performs much like a regular compression driver in that it is about 26dB down at 15kHz from its peak. :(

What are you specifically looking for? I am interested for sure. I have never had the 4367’s horns at my disposal.

Thank you Rob.
Barry.

[sebackman]

Yes the small ones are #6815 from VTX F15/F12.

My plan is to do a run with 2450SL, 2451SL, 2451/475Nd, 2451Ti and D2's on all WG's. I don't have access to an Mg driver and not sure I will take apart my main system to get to the Be's.

I use MLS pulses as they are rather noise resistant and will record gated runs with 3ms, 6ms, 9ms time window from about 4-500HZ (no smoothing, that can be applied later) at both 50cm and about 100cm distance.

I hope to do 90, 60 30 degrees on the 6ms runs.

That will be quite a numer of measurements on that many driver pairs. I will run a Focusrite III USB card and do 2 channel measures where the amp signal (Alesis RA300) is subtracted for better resolution. The mike is the usual Earthworks M30. I will not apply any XO, just a capacitor in series for protection and run at maybe 80-85db max.

I will also try run impedance runs for all driver/WG combinations with a DATS3 box.

I hope that would give sufficient data to see what is the room and what is driver/WG and how the 4367 fairs against the M2. I hope the 4367 could be a good choice as the M2 is really 100x100 degrees which is fine in a dampened studio environment but may require some adjustments at home.

I have a space in the workshop where I can apply 8cm egg crate padding to all walls which should help. It's cold over here right now so no possibility to go outside.

This is to decide what route to go with my DSP active 4367M clones that will use a 2451 core instead of the D2's. Unless this shoot-out shows that the D2's are indeed superior on the 4367 WG.

Barry, we routinely use 2450SL or 2451SL on the little #6815 in our two way systems. Albeit they are DSP systems but we use the combo from about 1kHz to 20kHz without any problem. Normally we use a 2206 or a PRO1400 under the horn/WG + a sub.

I also have a system with the same combo and a 2217 which also works fine and does provide bass to be used as party speakers without sub. PM email address and I can send some measures.

All input much welcome on changes to the plan or additional thing to check out.

Kind regards
//Rob

[sebackman]

Here is the driver lineup. :-)

[Mr. Widget]

Oh boy!

And oh boy do you have your work cut out for you!!!


Widget

[Robh3606]

Hello Rob

"I hope to do 90, 60 30 degrees on the 6ms runs."

Just a suggestion but I would add a 15 degree measurement as well. Figure the listening window should be between 0 and 30 and it wouldn't hurt to have a look in the middle so to speak. Just to see how much variation there is between the 3 measurements. You could also average them to see what the "window" curve looks like.

You certainly do have your work cut out!

Have fun Rob :)

[srm51555]

This is going to be fun and exciting! Thanks Rob

[JuniorJBL]

Looking forward to your observations Rob!! Thanks!

[engineerjoe]

Thank you for doing this and I hope I can follow along.I have to ask though, what is a 2451Ti?special diaphragm?

[Mitchco]

Very cool! Looking forward to your data. Are all the waveguides constant directivity? Any thoughts on accounting for constant directivity eq or is it not applicable for your scenario? https://sound-au.com/project173.htm Btw,I am a huge fan of JBL CD's and waveguides. I use the 2384 waveguide with 2453h-SL cd. I really like the sound of the large format JBL cd's. And the Aquaplas coating smooths out the response. But i have been eyeing the M2 and 4367 waveguides for sometime as the 2384 is a little too narrow in directivity. So, I am very curious to see and hear about your results. Thank you for doing this!

[sebackman]

Busy day with about 200 measurements done. All impedance measures done and all 50cm measures done.

Tomorrow I will do 100cm measures to compare. Annother 160 or so.....

4367 does load the drivers deeper than M2 and 4" does provide better coverage than D2 below 1kHz. But we guessed that, now we know.

M2 is smoother of axis comp to 4367 and VTX. Maybe 50cm is just to short for the 4367 off axis, I will see tomorrow.

More data tomorrow.

Rock on
//Rob

[sebackman]

I have now completed the measurements and albeit not the same quality as anechoic they do give a reasonable picture of the various JBL 1,5" drivers on the VTX, M2 and 4367 WG's, on and off axis. At least the settings (HW/SW/room/settings/timing/temperature) are identical in all measures, which makes them direct comparable.

The result is about 400 measurement files at 50cm and 100cm distance for 30, 60 and 90 degrees. The files are done without any smoothing and include phase. I did do 3ms, 6ms and 9ms time windows for all combinations to make sure that I got reasonable low end data with a long window but also a short window to see the high frequency with limited room intervention. I used pairs of 2450SL, 2451SL, 2451&475Nd, 2430K and 2451h (Ti, only one). 9 drivers all in all.

I have long argued that the 4" drivers do sound good on these WG's and numbers seem to support that. Please bear in mind that these are raw files and would look significant "nicer" with some smoothing, as most (also JBL) manufacturers show them. It is also clear that 4" goes lower than D2 in these WG's and that 4367 provides horn load to a lower frequency that M2. M2 sure is nice off axis. Horizontal and vertical dispersion is NOT the same in the M2 as could be anticipated from looking at it.

Below are some graphs. You can see measurement data (distance, time window aso) in the graphs. SW is LspCAD 6 PRO with MLS pulses at max resolution. HW computer + Focusrite SOLO V3 soundcard, Earthworks M30 mic, and an Alesis RA300 studio power amp. All walls (and ceiling) had at least 8cm thick sound padding (with egg crate surface absorbers). Measures are 2-channel, meaning that one channel measures the power amp out and second channel the mic and we only project the delta. No capacitor, only a protection resistor in series, so phase is not tuned 90 degrees as in many other measures with protection cap.

Kind regards
//Rob

[Mr. Widget]

Lots to absorb there… thanks for sharing!


Widget

[1audiohack]

Hello Rob;

I don’t have the 4367 but have everything else you have tested and your findings mirror mine.

Thank you for showing the phase response as well. That is one solid key to establishing a drivers truly linear usable range.

Thank you for taking the time to post and share all of this.

Barry.

[Wardsweb]

That is dedication. Thanks for the info.

[jmpsmash]

Incredible work. Thanks for sharing.

[sebackman]

If there is any particular data of interest just shout. All of it is too big to email/post (450Mb) but I can supply parts to interested members, however not for general distribution for obvious reasons.

Just finished the first of two pairs of "Hotrodded" Array 4892 for my workshop (PRO1400 & 2450SL). With new diaphragms in the 2450SL's and new JBL wave guides they are just wonderful compared to just about anything. Crisp! Albeit I have to admit I cheat using BSS DSP in an active setup. Subs are 2 pairs of Array 4893 with dual 2217 in each and lots of power. Can't stop playing. :-)

Next stop will be to finish the clone "Hotrodded" 4367 with 4" driver and BSS DSP in a M2-type active setup. That should be interesting. I do like what I have seen in the measurements when it comes to the 4367 wave guide. I will post in the other thread when I get there.

Now, where to put it all....

[mefisto]

Hi Rob, let me join the chorus of praise for your efforts. Since you offered, could you provide the measurements of only the 2450SL on the M2 and 4367 WG? If this is still too much of data with all the different time windows/distances, what would be the best combination for ascertaining the on-axis and of-axis performance from let us say 800 Hz to 20 kHz? The intention is to build a two-way speaker.. Kindest regards, M

[Mitchco]

Hi Rob, Nice job! Love the large format driver on the M2. Makes me want to purchase a pair of those waveguides. I am not a user of this tool, but I understand one can import measurement data to show some nice polar response charts and other views: https://kimmosaunisto.net/ Not suggesting a make work project or if it is suitable for your purposes, just in case you hadn't seen it before... Kind regards,Mitch

[Robh3606]

Hi Rob, Nice job! Love the large format driver on the M2. Makes me want to purchase a pair of those waveguides. I am not a user of this tool, but I understand one can import measurement data to show some nice polar response charts and other views: https://kimmosaunisto.net/ Not suggesting a make work project or if it is suitable for your purposes, just in case you hadn't seen it before... Kind regards,Mitch

Hello Mitchco

I am running M2 waveguides with 476Mg's. Using a large format driver is highly recommended. They sound effortless and clean.

Rob :)

[Mitchco]

Hello Mitchco

I am running M2 waveguides with 476Mg's. Using a large format driver is highly recommended. They sound effortless and clean.

Rob :)

Hi Rob, nice! Thanks for confirming.

I will be giving the them a try with the 2453h-SL drivers. I don't think I need an adaptor or spacer...?

Looks like they are going for $269 at: https://reconingspeakers.com/products-page/jbl-m2-horn-lens-5025594/

Kind regards,
Mitch

[Robh3606]

Hello Mitchco

I like the 2453-SL use them on PTH1010's Did not need a spacer on the 1010's. Don't forget the mounting plates. You will need them if you don't have an alternative.

Check the OD of the 2453 against the 435Be which is 4 1/4.

2453 S/B bigger with a 4" diaphragm might interfere with the mounting plate hardware to the horn. I don't have access to a mounting plate or a waveguide or 2453 to check the hole pattern diameter against 2453 OD.

I countersunk the mounting plate hardware and added helicoils to the waveguide so I could flush mount the 476Mg. PITA getting the 1/4 20's screwed into the driver through the plate.

Rob :)

[Mitchco]

Hi Rob,

Thanks for the info!

Kind regards,
Mitch

[sebackman]

Hi,

I made my own aluminium adapter plates with both the big and small JBL pattern. Screws into the WG are counter sunk and I use pin-bolts & nuts for the drivers. No problem to use either JBL 1,5" driver as the holes are offset.

And I would agree with Rob, the 4"drivers seem to like these new WG's. And no doubt the 4" drivers will be a better match for XO below 1kHz on either the M2 or the 4367 WG. In a 3-way (4-way) it may be less sensitive. In a 2-way it may prove important as few 12", 14" or 15" units can go above 1kHz.

476Mg is probably the best choice as you get almost the resolution of the Be (is the difference really audible ??) but no need a UHF as with Be. The Be's do fall off quite steep, but you can use DSP or passive filters (using the superior sensitivity) to compensate. My findings are that an introduction of an UHF (045) XO @11-12kHz sounds more "effortless". And I do use DSP.

2353SL does ned an adapter/spacer on the VTX WG but much less so on the M2. There is a pick-up from 10-12kHz instead. I have not measured them on 4367. Good alternative.

Hence I will not go Be in the new builds. Now if I could only find a set of Mg diaphragms...

[Mitchco]

Thanks for the additional info! The M2 will be quite a change I suspect from the 2384's I am using now.

On another forum, user notnyt measured a few JBL CD's on the 2384 and had this to say about listening impressions:
https://www.avsforum.com/threads/just-purchased-a-pair-of-jbl-4722n-speakers.1925586/page-194#post-42167329

If you scroll back through the thread there are several measurements to go with the comments. Don't know if the comments make sense, but it was enough for me to remove the 2432H CD that were on the stock 2384's (sounded harsh) and try the 2453H-SL which sounded much smoother.

Thanks again for your measurements!

[ivica]

I have now completed the measurements and albeit not the same quality as anechoic they do give a reasonable picture of the various JBL 1,5" drivers on the VTX, M2 and 4367 WG's, on and off axis. At least the settings (HW/SW/room/settings/timing/temperature) are identical in all measures, which makes them direct comparable. The result is about 400 measurement files at 50cm and 100cm distance for 30, 60 and 90 degrees. The files are done without any smoothing and include phase. I did do 3ms, 6ms and 9ms time windows for all combinations to make sure that I got reasonable low end data with a long window but also a short window to see the high frequency with limited room intervention. I used pairs of 2450SL, 2451SL, 2451&475Nd, 2430K and 2451h (Ti, only one). 9 drivers all in all. I have long argued that the 4" drivers do sound good on these WG's and numbers seem to support that. Please bear in mind that these are raw files and would look significant "nicer" with some smoothing, as most (also JBL) manufacturers show them. It is also clear that 4" goes lower than D2 in these WG's and that 4367 provides horn load to a lower frequency that M2. M2 sure is nice off axis. Horizontal and vertical dispersion is NOT the same in the M2 as could be anticipated from looking at it. Below are some graphs. You can see measurement data (distance, time window aso) in the graphs. SW is LspCAD 6 PRO with MLS pulses at max resolution. HW computer + Focusrite SOLO V3 soundcard, Earthworks M30 mic, and an Alesis RA300 studio power amp. All walls (and ceiling) had at least 8cm thick sound padding (with egg crate surface absorbers). Measures are 2-channel, meaning that one channel measures the power amp out and second channel the mic and we only project the delta. No capacitor, only a protection resistor in series, so phase is not tuned 90 degrees as in many other measures with protection cap. Kind regards //Rob HI Rob, Nice work. Many thanks that You have shown such work to us, Best Regards Ivica

[emilime75]

I'm going to do a measuring run on JBL M2, 4367 and VTX waveguides the next few days. I will try a few different dia's in 2450SL, and 2451 cores. I'don't have an Mg dia and not sure it will take apart my other system to get to the Be's. I will prost some findings here and if someone want the raw data just give me a shout.Kind regards//RobRob, from your testing and in your opinion - how low could a 2450SL on a VTX waveguide be crossed over if in an active system? Do you think the combo would work well as a 2 way with a 2216ND?

[sebackman]

Hi emilime75,

I typically XO 2450SL om the VTX at about 1Khz LR24. Somewhere between 950 to 1200Hz dependent on woofer. A 10 or a 12" can go higher which may prove beneficial in some setups. In the Franken HiFi Array's we do 1050Hz above PRO1400 with LR24 both ways. 0,271ms delay on woofer and 135 degrees phase on driver.

On the 2216 I would go for 1kHz and I think that would be an excellent combo. The 2216 can go up there and is probably one of the few 15" in the world that can, while offering good bass and mid range. You can't go wrong with that combo.

In my new 4367 clone I XO 2451SL at 750Hz because the horn can do it (much wider) over 226Nd-1 and that is very promising so far.

Go for it!

//Rob

[emilime75]

Hi emilime75,I typically XO 2450SL om the VTX at about 1Khz LR24. Somewhere between 950 to 1200Hz dependent on woofer. A 10 or a 12" can go higher which may prove beneficial in some setups. In the Franken HiFi Array's we do 1050Hz above PRO1400 with LR24 both ways. 0,271ms delay on woofer and 135 degrees phase on driver. On the 2216 I would go for 1kHz and I think that would be an excellent combo. The 2216 can go up there and is probably one of the few 15" in the world that can, while offering good bass and mid range. You can't go wrong with that combo.In my new 4367 clone I XO 2451SL at 750Hz because the horn can do it (much wider) over 226Nd-1 and that is very promising so far.Go for it!//RobThank you for the reply, Rob. Guess I now have a big decision to make, but I should probably start a separate thread and not derail this one.

[mefisto]

Greetings all,

it appears the the 4367 wave-guide is no longer available. Is thee any other one that would allow for a 600-700 Hz cross-over with 2450SL driver? I have the Aquaplassed diaphragm is it matters. I have build an enclosure for TAD 1601A and that is where it is crossed-over in the Exclusive models.

Kindest regards,

M

[dn92]

Greetings all,it appears the the 4367 wave-guide is no longer available. Is thee any other one that would allow for a 600-700 Hz cross-over with 2450SL driver? I have the Aquaplassed diaphragm is it matters. I have build an enclosure for TAD 1601A and that is where it is crossed-over in the Exclusive models.Kindest regards,MTAD TH-4003 clone will correspond to this. There is a good one manufactered in Hungary available through eBay.

[Ian Mackenzie]

Greetings all,

it appears the the 4367 wave-guide is no longer available. Is thee any other one that would allow for a 600-700 Hz cross-over with 2450SL driver? I have the Aquaplassed diaphragm is it matters. I have build an enclosure for TAD 1601A and that is where it is crossed-over in the Exclusive models.

Kindest regards,

M


Look at the Joseph Crowe radial horns from Canada

He has measured these horns in his blogs. They are impressive!

https://josephcrowe.com/collections/see-all

https://josephcrowe.com/products/es-600-bi-radial-wood-horn-no-1978

He matches the horn throat to your driver and the crossover.

I think you will find this horn a better option than the 4003 which as specifically designed for the 4003. The throat sections of that horn maker were not acceptable. He may have improved since. The M2 wave guide is another consideration.

[sebackman]

Hi mefisto,

I don't know what could work beyond the JBL M2, 4367 and the VTX wave guides. Sorry can't help there.

Do ask for specific measures with the 2450SL/2451SL on whatever horn you decide to use. The distance between the phasing plug and the mouth does matter. Horns that may look very similar may not always perform the same way.

I have tried many horns/wave guides but in reality few of the expensive and nice looking horns can match the rather simple and optically not convincing VTX horn. Even if the newer JBL horns are really made for the D2 they seem to love the 4" drivers.

I would ge a single of the el cheapo plastic VTX waveguides and compare to whatever else you find. Its only $50 in the US and bolts straight up to the 2450SL. Coat it on the outside with bitumen helps a bit with making it more "dead".

I have no idea on the horn recommended above and maybe they are just the ticket. But do buy a "wave guide" and as oppose to the older diffraction horns. For many of them there is a dramatic improvement for HiFi use. There are probably excellent diffraction horns for the 245X drivers, but he hassle to find them made me give up and use the proven combos.

Ian knows his stuff so maybe start there and I agree that the older TAD 4003 horn may or may not suit the JBL drivers. And they seem to have a diffraction mouth which would be a no-go for me. But they sure look nice. Each to their own.

If you need any driver horn data/measurements on the JBL wave guides just give me a shout

Kind regards
//Rob

[toddalin]

It's easier to demonstrate off axis using an RTA, a protractor, and piece of string. I was easily able to easily do 5 degree increments with mine:

https://youtu.be/pbbS0QYK2_o

[mefisto]

Hi dn92,

thank you for your reply.

I know about Athos Audio, I even corresponded with a gentleman who was contemplating ordering from them, but the quoted price was beyond my means.

Hi Ian,

thank you for your reply.

As with the Athos Audio, I am aware of Mr. Crowe, but again the price is way beyond my means.

Hi Rob (sebackman),

thank you for your reply.

If I am misinterpreting your measurements from post 13 correctly, the VTX would not support 600 Hz cross-over, but the M2 might.

Regarding the measurements, I have a dim recollection that we corresponded about it, but I cannot find any such e-mail. So if you have a measurement of the 2450SL (Ti Aquaplassed diaphragm) on the M2 wave-guide and you think that it could be reasonably equalized to about 600-700 Hz, could you please e-mailed it to me?

Kindest regards,

M

[Ian Mackenzie]

Hi Rob,

I had some time to look closely at your measurements posted earlier in the thread.

A few thoughts for users at home contemplating these VTX wave-guides

Rob’s work here is an excellent foundation for a nice diy project.


The crossover point and crossover characteristics
Within reason l don’t see an issue with crossing the VTX wave guide over at 750-850 hertz with the named driver variants.

Attempting to go down to 700 hertz without a steep high pass filter is best avoided. In theory the match of the hortizonal and vertical low end pattern control is likely more audible. Wave guides unless large like the 4367 typically unload and loose pattern control below 750-850 hertz. Thats why they are wide. What’s needed is ideally a match in the pattern control with the woofer and the wave-guide and a smooth pattern control up the sensitive 2,000 hertz region. This is to avoid pinching the polar response from 100+ down to 80 or less too quickly. Those discontinuities are bad if you read the 4430-4435 white paper.

The woofer properties
A light cone at the apex of the dust cap (no mass ring) is important to the continuity of the woofer and the wave-guide. A crossover point going of hundred hertz is not going to change that with removal of the mass ring in the case of the 2235H. The 2234 in the 4435 has no problem keeping pace with the 2344 with a 12 db crossover. Removal of the mass ring is not that difficult (Google) in preference to shelling out on a pair of 2216nd which are more challenging to set up. It’s then a fairly inexpensive diy loudspeaker with cheap drivers and horns.

Aim for least a 24 db slope on the woofer. Like wise for the wave guide.

The details
The how part is the key for those at home.

The effect of a fixed L pad in preference to a protection cap? What to do?
Greg Timber’s advocated a 6 db pad when l last met him. With an active crossover Greg said 6 db was okay. I would even look at a 10 db pad too keep noise down if going active.

The effect of the changing phase response of the wave-guide needs to be accounted for in either FIR dsp or a passive crossover to ensure phase tracking.

The horn EQ
The challenge for those at home is defining the EQ for these wave guides either in DSP or a passive crossover. Then having a simple way of bending the high frequency (brilliance) to work with their preferences. The same with the 2,000 region (presence) which l traditionally found sensitive in setting up the 2344 100x100 wave guide You need to listen or measure because it’s quite sensitive in that region. Compare to a reference loudspeaker as Robert H suggested.

REW and your ears are your friend

A Simple Passive DIY M2 network.

A passive network is certainly a simple approach if it works correctly. A while ago l did a simple passive but very effective steep acoustic slope 2 way passive network that could be adapted to any of these M2 or VTX wave guides and drivers with adjustments.

The attached screen dumps are the on axis response, the simplified schematic and a graphical representation of my simple passive network for the M2 driver and horn components. The waveguide variable EQ shifter was recommended by Greg Timbers who said the M2 wave guide when flat can be too hot in some acoustic settings.

The response is not a magazine curve and the final response will need to be measured in free field conditions.

The aim was to develop a passive network for diy audio amateurs at low cost and complexity that effectively tracks the M2 active system response except for the low.

The network can be easily revised for other woofer and horn driver combinations including the VTX wave guides. A seperate active low frequency boost module will be developed for the diy audio amateur to emulate the M2 low frequency specifications.

To clear the aim here is not to clone an exact passive version of the M2 but to offer a working diy passive M2 monitor without all the dsp hubba and expense.

Once l have tested the network with some different drivers l will post an update and talk about the next steps.

[Ian Mackenzie]

Hi dn92,

thank you for your reply.

I know about Athos Audio, I even corresponded with a gentleman who was contemplating ordering from them, but the quoted price was beyond my means.

Hi Ian,

thank you for your reply.

As with the Athos Audio, I am aware of Mr. Crowe, but again the price is way beyond my means.

Hi Rob (sebackman),

thank you for your reply.

If I am misinterpreting your measurements from post 13 correctly, the VTX would not support 600 Hz cross-over, but the M2 might.

Regarding the measurements, I have a dim recollection that we corresponded about it, but I cannot find any such e-mail. So if you have a measurement of the 2450SL (Ti Aquaplassed diaphragm) on the M2 wave-guide and you think that it could be reasonably equalized to about 600-700 Hz, could you please e-mailed it to me?

Kindest regards,

M

I understand.

[Robh3606]

Hi dn92,

If I am misinterpreting your measurements from post 13 correctly, the VTX would not support 600 Hz cross-over, but the M2 might.

Regarding the measurements, I have a dim recollection that we corresponded about it, but I cannot find any such e-mail. So if you have a measurement of the 2450SL (Ti Aquaplassed diaphragm) on the M2 wave-guide and you think that it could be reasonably equalized to about 600-700 Hz, could you please e-mailed it to me?

Kindest regards,

M

I have done a 700Hz crossover point in my DIY Passive "M2" I was pushing it a bit but I got the best summation @ 700 and was only about 50Hz lower than the 4700. My measurement is with a 476Mg but any of the 2450 series should be very similar as far as low end response on the waveguide. Look at the upper curve no EQ. You can see the step below 700 so 600??


Also attached is a measurement on the PTH1010 a 100x100 waveguide. Smaller but still loads well and is inexpensive compared to the M2. There is now a -1 version I don't have measurements on.

Rob :)

[mefisto]

Hi Robh3606,

thank you for your reply.

As I am not as knowledgeable as the forum members (hence my search for a wave-guide/driver combination that is documented), can you explain the "I was pushing it a bit" because looking at your raw measurements, the 105 dB level is crossed as approximately 650 Hz, so why is the cross-over at 700 Hz "pushing it"?

Regarding the "best summation @ 700", are you talking about the pattern matching as mentioned by Ian?

I will read your thread.

Is there any other source for the M2 but Speaker Exchange?

Kindest regards,

M

[sebackman]

mefisto, if you look at the data files (pls see below) you can see how low you can go with each driver on each horn. All the raw data is there. No caps to adjust for, just plain FR in different forms. Data can be imported to your favorite simulator SW.

The small VTX horn it is not the same as the older PT-XXXX. It is JBL part no #5006815. They look similar but they are not, the 6815 is a better waveguide. Albeit that the PT-series does exist in a 100x100 and the 6815 is 90x40, I think. Narrow dispersion vertically may not be a drawback as it reduces floor and ceiling bounces.

With DSP you can get away with XO very low with steep filters. In my findings I agree with Ian and if you have a decent woofer that can go up a bit it is easier to get the drivers to like each other.

IMHO 700Hz is borderline for M2 and would say closer to 800-850Hz is easier. But is surely can be done as Robh3606 clearly shows. I do 850HZ on my M2/476's from 1400PRO.

4367 can do 700Hz as it is wider but I understand that they are NLA, so that is may not be an option unless you find someone sitting on a pair. I will do 700LR24 DSP to 2216Nd-1 in my 4367 clones.

VTX is probably 875-1000Hz even if also they can be pushed down with DSP. I typically do 950-1000Hz in my builds.

Regarding passive XO's I'm the wrong guy to give any advise. Fortunately many others here can. :-)

For whoever may having a difficulty getting to sleep attached is a link to my Dropbox with all the shout-out data. I will leave the data there for a few weeks and then take it down, so download what you need.

Start with looking at the schedule in the Excel file to see the way the files are named. There you also can see which driver is which as I measured 9 drivers of different models.

Full set of data files for 50cm & 100cm distance for all drivers on all horns. 90, 60 and 30 degrees covered. In the PDF there are screen shoots of all measures. No smoothing except for a few added graphs and then noted in the text.

Impedance data for all combos done on a DATS3. SW free from Dayton even without the HW to view files.

"Naked" means no wave guide, nothing else...

https://www.dropbox.com/sh/57lpmq9ogf3m9na/AABuCnr5thRHk4DssJaTowd_a?dl=0

Have fun.

[mefisto]

Hi Rob (sebackman),

thank you for your response and the link to the measurement data.

I will certainly look at the data, the problem is, I am not sure that I can interpret them correctly. To wit, see my response to Robh3606 who wrote that the 700 Hz was "pushing it", but my reading of the raw data was that 700 Hz could be comfortable.

Regarding your statement "Narrow dispersion vertically may not be a drawback as it reduces floor and ceiling bounces", my understanding is that the size of the wave-guide mouth determines the frequency range within which the pattern is controlled, cf. Keele. Thus, unless one crosses over below this frequency, no reduction of the bounces is achieved.

Regarding the "small VTX horn", I am confused; the JBL part no 5006815 shows as "Set up for 4 or 2 bolt pattern. 1”-1.3” throat"; the 2450 SL is 1.5" throat.

Kindest regards,

M

[sebackman]

Hi mefisto,

Easy part first, the 5006815 is a 1,5" throat waveguide, so that is plain wrong. I have a bunch of them here. The bolt pattern on the flange is the small JBL 4" driver bolt pattern as 2450SL (4-bolt) and I think 2453. They will also take the D2 (aka 2430k, 2-bolt)) direct fit.

If you want to use 2451 or 2452 you need to either make an adapter or just use some spacers and large washers to bolt it down. I use an adapter. Look in the "compact monitor" thread that I did a few year ago, I think the drawing on the adapter plate is there.

On the lowest frequency that is determined by the acoustic load on the driver by the horn and is a function of the combination of the driver and the horn design. At a certain frequency the horn does not represent an acoustic "resistance" for the driver which means that the "controlled" function reduces quickly. It may well be that it can be used below this but many parameters will change and unless you have measuring capabilities and lots of time I would not recommend going below this point.

If you look at the impedance curves (DATS3) in my provided data you can see what happens when the acoustic "connection" between the drive and the horn gets lost.

I agree that the low frequency capabilities is to some extent directed by the throat but also the flare and the width of the horn (and many others). There are certainly others here that can lay out the technical specifics much better than I can.

Regarding dispersion the most important parameter is the physical shape of the flare/horn. A square horn will have a somewhat even dispersion horizontally and vertically while a rectangular horn will have different coverage pattern. A reduced vertical dispersion will reduce sound power going in that direction compared with straight and horizontal energy, which will inherently lower the amount of floor and ceiling bouncing energy. Somehow I don't think that was the question?

kind regards
//Rob

[mefisto]

Hi Rob (sebackman),

thank you for your response and specifically for confirmation that the 5006815 is a 1,5" throat wave-guide. And no, I will not use 2451 or 2452 as I do have the 2450SL.

I do not want to derail the thread too much, but my understanding is that the "reduced vertical dispersion will reduce sound power going in that direction" works only below a frequency controlled by the size of the mouth. Thus the cross-over must be below this frequency for this to work.

I had downloaded all the files, that is an enormous amount of work that you had done.

Kindest regards,

M

[Ian Mackenzie]

We appreciate you chiming in on your project. Detailed discussion of your project is best moved to a new thread.

Some measurements of the 2216nd and the M2 horn on and off axis with REW and some JBL tech sheet data of the 90 x 50 waveguide.

The FR reponse curves for illustration purposes only. The divisions are 5 db and smoothed 1/24 octave as l recall in a room.

The 90x50 waveguide power response
The response graph is a 2216nd and a M2 horn and an unknown driver?

Remember your crossover point will be -6 db so in practice this will look like the attached image (DD67000 horn and driver with crossover and EQ).

Using REW is actually quite straightforward with a Mac PowerBook as you only need the calibrated Minidsp mic with its calibration file. For diy this is more than adequate loudspeaker measurements. REW capabilities had been continuously improved and it now features a host of measurement types for loudspeaker measurements, impedance, voltage drivers with a Focusrite usb interface, RT60, curve averaging, curve smoothing and psycho acoustic curve filters for assessment of room acoustics and loudspeaker voicing.



How you approach this is entirely up to you.

[toddalin]

And where would one even look to find these drivers anymore? Maybe here...

https://orangecounty.craigslist.org/msg/d/west-covina-jbl-drivers-and-horn-for/7667486353.html

[mefisto]

Hi Ian, all,


We appreciate you chiming in on your project. Detailed discussion of your project is best moved to a new thread.

I do not believe that I had mentioned my project in my responses herein, that had been sort-of disclosed here: https://www.audioheritage.org/vbulletin/showthread.php?43605-Question-re-near-field-low-frequency-measurement&highlight=

However, I understand the worry that the thread is moving away form the main topic; consequently, I will respond l the above-mentioned thread.

Kindest regards,

M

[Robh3606]

Hi Robh3606,

thank you for your reply.

As I am not as knowledgeable as the forum members (hence my search for a wave-guide/driver combination that is documented), can you explain the "I was pushing it a bit" because looking at your raw measurements, the 105 dB level is crossed as approximately 650 Hz, so why is the cross-over at 700 Hz "pushing it"?

Regarding the "best summation @ 700", are you talking about the pattern matching as mentioned by Ian?

I will read your thread.

Is there any other source for the M2 but Speaker Exchange?
Kindest regards,

M

Hello

Well if you look at the curve it's fairly linear until 700Hz where it quickly starts to roll off. When it starts to roll off the phase is changing as well. You are better off going for a 750-800 Hz. That said I am using odd order filters 3rd order acoustic woofer and 5th acoustic waveguide. I was not concerned with the exact crossover point. Just adjusted for a smooth summation.

If you look at the directivity plots on the M2 the directivity step change at crossover is actually at 700Hz not 800Hz. That said 50Hz one way or the other I don't see as a game changer as long as it sounds good and you don't have an obvious step in the response. The 4" drivers were crossed over as low as 400Hz in the early designs so they are robust. I would not recommend going that low.



Rob :)

[mefisto]

Hi Robh3606,

thank you for your reply.

Please note that I had moved any further discussion as per my post # 47.

Kindest regards,

M