Things to consider JBL 2226H + Faital 6FE100 + B&C DE250

[mga2009]

Hi,

I am planning a speaker for home theater use in a small room using the JBL 2226H, using a 3E Audio ADAU1701 as active crossover (or maybe a chinese ADAU 1452).

Already have the JBL woofers, the horn (QSC H152i clone) and the B&C DE250, and I have a couple of Faital 6FE100 6.5" woofers laying around that I could use...

Keep in mind these speakers will be used as left and right channel, and I am almost decided that I will go sealed, in a smallish cabinet, aiming for an extension 50hz+, as I already have 2x 15" subwoofers to take the work from there.

I do not aim for 110db+ just a nice sounding speaker, with nice dynamics and presence.

So the thing is, should I go for a 2-way or a 3-way? Any special thing to consider that cannot be solved afterwards with a ton of measurements and active crossover?

I have build several DIY speakers, modified a few, but never designed one from the ground, thats why I am starting with an active XO (things easier to implement), but I know designing a 2-way is hard and a 3-way is way harder, but I think the Faital would give me a nice meeting point with the 15" at around 800hz and with the CD around 1.5-1.8khz; I will loose a ton of possible power usage, but as I said I am not aiming for high sound pressure.

Looking forward to your opinion

Cheers!

 

[KCHANG]

Since you are asking for free opinions, let me put in my two cents. Hopefully some other people who are much more experienced/knowledgeable will chime in.

I think your 3-way idea with your selected drivers is quite promising, but I don't see how the 2-way idea would work unless you get a compression driver and a matching horn that could play below 1000Hz. For example, the crossover frequency of JBL M2, from its 15" woofer to the big waveguide, is 800Hz. I have not used the DE250/QSC-horn-clone combo, but I doubt that you can push the crossover frequency low enough to match the radiation pattern of the 15" woofer, even if the 2226 driver can comfortably play that high.

In contrast, the use of a mid-bass driver will comfortably solve the integration problem. The mid-bass driver can match the dispersion patterns of the 15” woofer and the horn. Moreover, the critical frequency range covering fundamentals of human voices and instruments, or at least a good part of it, can be covered by a single transducer – the mid-bass driver. Perhaps it is just me, but I always cringe when I see/hear people talking about crossing over at or around 800Hz. This “inconvenient” crossover point is a main thing in my mind against doing a two-way like M2 using a big woofer and a waveguide/horn.

I don’t see any serious negative point of going 3-way. I also think the drivers you picked should be good enough. With a careful selection of the crossover frequencies and the use of DSP to EQ the speaker to give the desired frequency response, conceptually this 3-way has the potential of being as good as many high-end speakers, DIY or commercial.

One change to your 3-way plan I’d suggest is to drop the crossover frequency between the 15” woofer and the 6” mid-bass to 300Hz, or even lower. This is for the reason mentioned above – the desire to use the mid-bass to cover a good part of the critical range of fundamental frequency.

Another comment is that you might want to consider using a modular structure, with the woofer and mid-bass each in its own box. The horn can just perch on the mid-bass box. The modular construction would allow you to explore a lot of things, such as adjusting the baffle width for the mid-bass, trying different schemes of edge rounding, switching to a different mid-bass or a different horn/CD combination, etc.

Recently I slapped together a pair of 3-way speakers, with each speaker comprising a bass module with two 8” woofers, an Audax PR170M0 in its own box, and a B&C DE10/ME10 combo. Even with just an analogue active crossover and no DSP, the speakers sound quite good.

Cheers,

Kurt

 

[mga2009]

Kurt,

Thanks for your reply.

I do have the same concerns as you mention, will the DE250 with the QSC horn will be enough to meet the woofer? maybe not as I still don't test it, but according to the DE250 spec sheet it should go down to 1khz - 1.1khz without too much problem, and the QSC horn (B52 clone) specs show a usability down to 800hz.

On the other side, the idea to have only one (or almost) transducer in that important frequency range it's a big advantage.

Your idea of separated boxes / modular structure is perfect. I will proceed with it, and start with a 2-way and test it as is, and with time compare it to a the 3-way.

For starters, would you agree on making the boxes so the vertical distance between drivers is the smallest, with the woofer and mid on the same vertical axis, both in sealed boxes (to avoid flips on the phase at tune freq. as it will offer more difficulty to make the right XO), with rounded edges (1" radius it's the bigger router bit I have).

For the 2226h I am thinking a sealed 1.9cu. ft. with linkwitz transform (f0 87hz and Q0,651 to fp 55hz and Q0,707), without HP filter.

Thanks again!

 

[barryso]

You might pick up a few pointers by reading some of Wayne Parhams postings over at pispeakers.com. Their forum is audioroundtable.com.

He uses the JBL 2226 and the B&C 250 tweeter in his upgraded 4 pi model. The 7 pi corner horn uses those drivers and adds a midrange driver.

There's a lot of good info in the audioroundtable faq's.

AudioRoundTable.com: Pi Speakers >> Pi Speakers FAQ

 

[lousymusician]

I use the same horn (Parts Express B52) with the DE250, and it will play nicely to 1000 Hz or so. That's a good crossing point with a 12" woofer, pushing a 15 that high might create some problems in the polar response as the woofer radiation pattern will be a bit narrow at the crossover. A single 6" could span the gap, but could create a polar mismatch at the higher crossing point, or drive you to a smaller tweeter horn.

I have seen HT speakers that address that by using two midranges side by side between the horn and the woofer. That might be a good approach here. Doubling the midranges would also give a useful gain in sensitivity and output. Keep the MR's as close together as possible. My first stab at xover frequencies would be 400 & 1200, then adjust to get good polars and smooth response.

Never tried it myself, so my opinion is worth the electrons it was made out of, but it could be a fun project.

 

[mga2009]

You might pick up a few pointers by reading some of Wayne Parhams postings over at pispeakers.com. Their forum is audioroundtable.com.

He uses the JBL 2226 and the B&C 250 tweeter in his upgraded 4 pi model. The 7 pi corner horn uses those drivers and adds a midrange driver.

There's a lot of good info in the audioroundtable faq's.

AudioRoundTable.com: Pi Speakers >> Pi Speakers FAQ

Thanks, I will take a look there!

I use the same horn (Parts Express B52) with the DE250, and it will play nicely to 1000 Hz or so. That's a good crossing point with a 12" woofer, pushing a 15 that high might create some problems in the polar response as the woofer radiation pattern will be a bit narrow at the crossover. A single 6" could span the gap, but could create a polar mismatch at the higher crossing point, or drive you to a smaller tweeter horn.

I have seen HT speakers that address that by using two midranges side by side between the horn and the woofer. That might be a good approach here. Doubling the midranges would also give a useful gain in sensitivity and output. Keep the MR's as close together as possible. My first stab at xover frequencies would be 400 & 1200, then adjust to get good polars and smooth response.

Never tried it myself, so my opinion is worth the electrons it was made out of, but it could be a fun project.

Thanks for your opinion.

I have 4 Faitals laying around, so I can check 1 woofer and then 2 side by side. I am still learning the basics here, so, when I have the "modules" assembled I have to check the polars of the MW so they match the polars in the horn, so that I have a constant directivity across all the frequency range?

Cheers

 

[KCHANG]

For starters, would you agree on making the boxes so the vertical distance between drivers is the smallest, with the woofer and mid on the same vertical axis, both in sealed boxes (to avoid flips on the phase at tune freq. as it will offer more difficulty to make the right XO), with rounded edges (1" radius it's the bigger router bit I have).

For the 2226h I am thinking a sealed 1.9cu. ft. with linkwitz transform (f0 87hz and Q0,651 to fp 55hz and Q0,707), without HP filter.

In my view vertical alignment of the drivers is nice to have, so that you'll get a predictable horizontal polar pattern. If, for example, you offset the woofer from the horn in a 2-way, the axis is tilted and I don't think the resultant horizontal (relative to the floor) radiation pattern will be desirable or easy to manage. For instance, you'll see what I mean if you go to AudioScienceReview and read the review of the JBL HDI-3600 speaker and imagine the mixing of the good horizontal directivity pattern with the not-so-great vertical directivity pattern due to the axis-tilting.

JBL HDI-3600 Speaker Review | Audio Science Review (ASR) Forum


As to having smallest distances between drivers, in a two-way you have no choice but to keep the 15" woofer as close to the horn (which is likely big) as possible so that the center-to-center distance is manageable. In a 3-way, it is still important to keep the mid close to the horn to keep the C-C distance reasonably short compared to the wavelength at the (likely higher) crossover frequency. The C-C distance between the mid and woofer, however, might not be that critical now due to the much longer wavelengths involved, if you pick a low crossover frequency such as 300Hz. This would give you the flexibility to position the woofer closer to the floor, which has the advantages of getting more boundary enhancement and reducing the problems of floor bounce.

Regarding the use of sealed boxes, it seems a natural choice to put the mid in a sealed box. I have no experience, however, in putting a big 15" woofer in a sealed box. The JBL 3-way cinema speakers I had a while back used 2226 woofers and the speakers were ported. I think the sealed box with a Linkwitz transformation should work fine, but I have not done that before.

I have been using plastic pipes cut longitudinally in halves and taped to the front-baffle edges of speaker boxes to form "rounded edges." They surely look ugly, but it is a quick-and-dirty way to try the effects of rounding the edges of a speaker enclosure. My speaker projects are all "experimental" and it has been >20 years since the last time I tried to make beautiful speaker boxes.

Kurt

 

[hweb]

The 2226 is a pretty special driver. I think you would be OK with it out to 1kHz where the DE250/horn would cross over.

The problem with a 6.5" mid in this design is that you would really want to cross it over at about 2 to 2.5kHz. The QSC horn you have is waaaay too big for that. True, that you would have a polar match at 2.5kHz, but the center to center distance from the mid to the horn would be needlessly large and cause some vertical pattern problems.

Here's what I would do: Build a box adequately tall to allow for a baffle that would accommodate the 2226, a 6.5" mid, and a reasonable horn for 2.5kHz crossover (not the QSC). Then you can try both options (2 way and 3 way) with a simple switch of the front baffle.

That said, since you already have the 2226, DE250 and QSC horn, I would build that first (as I'm pretty certain you'll be happy with that - particularly if you keep the sub-50Hz stuff off the 2226).

 

[barryso]

The 2226 is a pretty special driver. I think you would be OK with it out to 1kHz where the DE250/horn would cross over.

That said, since you already have the 2226, DE250 and QSC horn, I would build that first (as I'm pretty certain you'll be happy with that - particularly if you keep the sub-50Hz stuff off the 2226).

The 2226 can go into the 1.5 or 1.6k range without breaking up the way many big woofers do. It's part of what makes it a unique driver so it'll mate pretty well with the B&C tweeter. You're right about below 50 hz, it won't go real low. If you need bass down low add subs.

 

[mga2009]

In my view vertical alignment of the drivers is nice to have, so that you'll get a predictable horizontal polar pattern. If, for example, you offset the woofer from the horn in a 2-way, the axis is tilted and I don't think the resultant horizontal (relative to the floor) radiation pattern will be desirable or easy to manage. For instance, you'll see what I mean if you go to AudioScienceReview and read the review of the JBL HDI-3600 speaker and imagine the mixing of the good horizontal directivity pattern with the not-so-great vertical directivity pattern due to the axis-tilting.

JBL HDI-3600 Speaker Review | Audio Science Review (ASR) Forum


As to having smallest distances between drivers, in a two-way you have no choice but to keep the 15" woofer as close to the horn (which is likely big) as possible so that the center-to-center distance is manageable. In a 3-way, it is still important to keep the mid close to the horn to keep the C-C distance reasonably short compared to the wavelength at the (likely higher) crossover frequency. The C-C distance between the mid and woofer, however, might not be that critical now due to the much longer wavelengths involved, if you pick a low crossover frequency such as 300Hz. This would give you the flexibility to position the woofer closer to the floor, which has the advantages of getting more boundary enhancement and reducing the problems of floor bounce.

Regarding the use of sealed boxes, it seems a natural choice to put the mid in a sealed box. I have no experience, however, in putting a big 15" woofer in a sealed box. The JBL 3-way cinema speakers I had a while back used 2226 woofers and the speakers were ported. I think the sealed box with a Linkwitz transformation should work fine, but I have not done that before.

I have been using plastic pipes cut longitudinally in halves and taped to the front-baffle edges of speaker boxes to form "rounded edges." They surely look ugly, but it is a quick-and-dirty way to try the effects of rounding the edges of a speaker enclosure. My speaker projects are all "experimental" and it has been >20 years since the last time I tried to make beautiful speaker boxes.

Kurt

Kurt,

Thanks for your reply. Tons on valuable info here. What diameter are you using on those plastic pipes?

The 2226 is a pretty special driver. I think you would be OK with it out to 1kHz where the DE250/horn would cross over.

The problem with a 6.5" mid in this design is that you would really want to cross it over at about 2 to 2.5kHz. The QSC horn you have is waaaay too big for that. True, that you would have a polar match at 2.5kHz, but the center to center distance from the mid to the horn would be needlessly large and cause some vertical pattern problems.

Here's what I would do: Build a box adequately tall to allow for a baffle that would accommodate the 2226, a 6.5" mid, and a reasonable horn for 2.5kHz crossover (not the QSC). Then you can try both options (2 way and 3 way) with a simple switch of the front baffle.

That said, since you already have the 2226, DE250 and QSC horn, I would build that first (as I'm pretty certain you'll be happy with that - particularly if you keep the sub-50Hz stuff off the 2226).

Thanks! As recommended, I will build modular boxes for testing the 2-way and a 3-way with 2x side by side mid woofers and make some tests.

To circumvent the CtoC distance between mid woofer and horn, would it help to angle the horn a little? Also, do you have any recommendation for a good horn / waveguide for the 2.0khz+ range?

The 2226 can go into the 1.5 or 1.6k range without breaking up the way many big woofers do. It's part of what makes it a unique driver so it'll mate pretty well with the B&C tweeter. You're right about below 50 hz, it won't go real low. If you need bass down low add subs.

Great, so there is a chance the 2-way will work, meaning easier design (for a noob like me) and less DSP and amplifier capabilities.

 

[hweb]

For a >2kHz horn you might look at the Eminence WG10 (Eminence WG10 1" High Frequency Waveguide 70x60 2-Bolt) or the Dayton H6512 (Dayton Audio H6512 6-1/2" x 12" Waveguide 1-3/8"- 18 TPI) which is a knock off of the JBL PT horn.

For the WG12, this Peerless CD (Peerless by Tymphany DFM-2535R00-08 1" Compression Horn Driver 2/4-Bolt 8 Ohm) would be a great place to start (actually you might not look any further). For the Dayton/JBL horn there is a JBL CD made for it (I think it's the 2408, but I'm having a senior moment). Patrick Bateman (on this forum) has posted lots of measurements of this combo and has written at length about it.

 

[mga2009]

Today I received my QSC clones. Still waiting forthe B&C's to arrive.

I made some quick and dirty (uber dirty) measurements with the QSC horn and Denovo DE150 CD (from www.diysoundgroup.com) and also the Celestion CDX1746 (which is said it's the poor man's DE250).

These are taken with a Cross-Spectrum Labs calibrated MiniDSP UMIK in my lightly treated room.

Measurements are 1/24 smoothing. Red is Celestion Blue is Denovo.

 

[mga2009]

I just wanted to update the thread with a piece of information regarding the JBL 2226H.

This driver is used in the STX series with the 2432H CD and JBL 5006815 Horn.

In the 2-way design (STX815) the JBL 2226H is crossed @ 1.8khz to the CD, althou i've read this is an acoustical crossover, but IDK the difference between electric and acoustic XO, shouldn't be the same?

 

[peterbrorsson]

In the 2-way design (STX815) the JBL 2226H is crossed @ 1.8khz to the CD, althou i've read this is an acoustical crossover, but IDK the difference between electric and acoustic XO, shouldn't be the same?

No, you can add a single pole electrical filter and get acoustically roll off second order or higher depending of driver response. You have to use dsp (or have huge overlap of drivers) and correct response to flat in order to achieve both electric and acoustical slopes...

 

[manninen]

For the WG12, this Peerless CD (Peerless by Tymphany DFM-2535R00-08 1" Compression Horn Driver 2/4-Bolt 8 Ohm) would be a great place to start (actually you might not look any further). For the Dayton/JBL horn there is a JBL CD made for it (I think it's the 2408, but I'm having a senior moment). Patrick Bateman (on this forum) has posted lots of measurements of this combo and has written at length about it.

Grand Comparatif de Compressions 1 pouce - JustDIYIt !
both tymphany and de250 tested here, not the best of the bunch

 

[scholl]

Ive used 2226H to 1.6K but would still strongly recommend the three-way with crossover points at 3-400 and 2-3Khz.

I've used the much better sounding AE TD15S to 1.6K then again at 700hz and experienced a huge improvement going active at 350hz to the 18Sound 6ND430. A nice improvement in Vocal clarity and sound stage and imaging. Crossing the tweeter at 2Khz made it the deciding factor in the off access response.

 

[hweb]

The 2226 is an exceptional driver, but I wouldn't take it much past 1kHz. It will start beaming badly. It is a 15" driver and physics just can't be ignored.

 

[mga2009]

The 2226 is an exceptional driver, but I wouldn't take it much past 1kHz. It will start beaming badly. It is a 15" driver and physics just can't be ignored.

Thanks for your reply! yup, everything points to the 3-way. For a 2-way, the QSC would be better matched to a 10" or 12"?

Ive used 2226H to 1.6K but would still strongly recommend the three-way with crossover points at 3-400 and 2-3Khz.

I've used the much better sounding AE TD15S to 1.6K then again at 700hz and experienced a huge improvement going active at 350hz to the 18Sound 6ND430. A nice improvement in Vocal clarity and sound stage and imaging. Crossing the tweeter at 2Khz made it the deciding factor in the off access response.

Yes, the 3-way it seems the more viable. I've made some early test and the sound it's quite good.

In any case, I have a doubt with the 3-way as In another forum, an experienced user (BFM) told me dispersion with a horn is narrower as the horn is made larger, and he wouldn't recommend use a horn spec'd to go much lower than the crossover frequency. The QSC horn is spec'ed lower than the (around) 2Khz crossover for the tweeter... so maybe I should go with another horn/waveguide?

Considering this, I purchased one the famous MCM 54-580 / Pyle PH65... so maybe I should go for the 3-way with a smaller and shallower horn, like this in a layout similar to the picture I attached (picture shown with an EOS6). The dispersion of the PH65 will match that of the Faital 6FE100 (working between around 400hz and 2.200hz or a little higher).

Grand Comparatif de Compressions 1 pouce - JustDIYIt !
both tymphany and de250 tested here, not the best of the bunch

Great comparison! Too bad it is limited to only one horn. Also, it seems from the test, that no CD is perfect and that the performance is closely related to the horn used and there is no CD perfect for all horns (and vice versa).

No, you can add a single pole electrical filter and get acoustically roll off second order or higher depending of driver response. You have to use dsp (or have huge overlap of drivers) and correct response to flat in order to achieve both electric and acoustical slopes...

I understand. But if I consider using DSP, wouldn't the acoustical XO be the relevant? Way I mean is, If JBL crosses the 2226H at 1.800hz passively, I can also do this with my DSP and a LR48.

 

[peterbrorsson]

Yes! If responses match up!

 

[Sjef]

If you have never built a three way loudspeaker before here is a hint to make life a little easier. One of the reasons why a three-way is so much more difficult than a two-way is because the highpass and lowpass filters on the midrange highly influence each other because of the phase shifts of the filters. In other words, any change you make in the low to mid crossover has a direct influence on the mid to high crossover as well. Try to keep the crossover points at least a decade apart to minimize this influence. So for instance cross at 250Hz and 2500Hz or 300Hz and 3000Hz, Further apart is even easier if the drivers allow it but at least this will be a whole lot easier than trying to get a 500Hz and 1500Hz crosover right.