Hi Bruce,
Looks like I might get around to doing this today, so if you can let me know what you're looking for in particular I can toss in a measurement or two.
As for right now, I'm going to use the curvy front baffle I have, and use 12.5" Sonotube in behind each woofer, and I'll adjust the length/volume for a Q of 0.6.
I have been measuring at 1m distance and tweeter level, so right between both TD10M's. I can also take a few measurements of a single TD10M to see if there is any combing going on here. My initial measurements of the TD10M's in the TL test box don't show any combing until 4.5kHz where regular dips are showing at 4.5k, 5.5k and 6.5k.
I also brought my SPL meter home from work, so I'll be able to get a reference dB reading at 2.83V for the TD10M.
Your idea for the two AE 6.5" drivers and the RAAL would certainly be a pretty staggering speaker. It was the first thing I thought too when I saw they were going to make a 6.5" driver. You could use all 95dB of the RAAL's sensitivity, and still get at least a little bit of bass out of the 6.5" drivers. In that case I would go for a TL for sure and probably not even bother with a sub.
I've been living with the Ariels for many years now, and I can honestly say that even with a roll-off at around 60-70Hz, they still have great bass, and I often prefer to listen to them without a subwoofer. Two TD6.5H's would give you enough sensitivity to max out the RAAL and according to John you could still get down to about 60-65Hz. I have some friends looking for a high end set of speakers, so that exact setup might be my next project.
I'll keep everyone posted on my progress.
Cheers,
Owen
Looks like I might get around to doing this today, so if you can let me know what you're looking for in particular I can toss in a measurement or two.
As for right now, I'm going to use the curvy front baffle I have, and use 12.5" Sonotube in behind each woofer, and I'll adjust the length/volume for a Q of 0.6.
I have been measuring at 1m distance and tweeter level, so right between both TD10M's. I can also take a few measurements of a single TD10M to see if there is any combing going on here. My initial measurements of the TD10M's in the TL test box don't show any combing until 4.5kHz where regular dips are showing at 4.5k, 5.5k and 6.5k.
I also brought my SPL meter home from work, so I'll be able to get a reference dB reading at 2.83V for the TD10M.
Your idea for the two AE 6.5" drivers and the RAAL would certainly be a pretty staggering speaker. It was the first thing I thought too when I saw they were going to make a 6.5" driver. You could use all 95dB of the RAAL's sensitivity, and still get at least a little bit of bass out of the 6.5" drivers. In that case I would go for a TL for sure and probably not even bother with a sub.
I've been living with the Ariels for many years now, and I can honestly say that even with a roll-off at around 60-70Hz, they still have great bass, and I often prefer to listen to them without a subwoofer. Two TD6.5H's would give you enough sensitivity to max out the RAAL and according to John you could still get down to about 60-65Hz. I have some friends looking for a high end set of speakers, so that exact setup might be my next project.
I'll keep everyone posted on my progress.
Cheers,
Owen
Owen, I'm interested in efficiency, frequency response (on axis and off axis), and distortion components.
Ideally once as a stand alone unit and a second time as a pair some distance in between, so that I can sort out the MTM vs MT trade offs.
I've got four different designs in mind ...
A WWMT, with 12s for the Ws, a 6.5 for the M, and a Rall for the T
A WMTMW with the same drivers
A WMTM with the same drivers
vs.
An MTM much like you are doing, but with the Raal rather than the Fountek, supplemented on the low end with dual subs.
I'm trying to get my mind around whether crossing over to a Raal as low as the MTM designs will require will break the "seamlessness" that I so crave, vs., the "power" and "heft" that I also crave. If your measurements continue to show little to no combing below 3.5k then that becomes pretty compelling.
What ever I end up doing will need a "matching" center for when I'm using the system as a theater rather than music.
Ideally once as a stand alone unit and a second time as a pair some distance in between, so that I can sort out the MTM vs MT trade offs.
I've got four different designs in mind ...
A WWMT, with 12s for the Ws, a 6.5 for the M, and a Rall for the T
A WMTMW with the same drivers
A WMTM with the same drivers
vs.
An MTM much like you are doing, but with the Raal rather than the Fountek, supplemented on the low end with dual subs.
I'm trying to get my mind around whether crossing over to a Raal as low as the MTM designs will require will break the "seamlessness" that I so crave, vs., the "power" and "heft" that I also crave. If your measurements continue to show little to no combing below 3.5k then that becomes pretty compelling.
What ever I end up doing will need a "matching" center for when I'm using the system as a theater rather than music.
opc said:
This was almost exactly what I ended up with in mine. I ended up with 1900hz Xover. I used the DCX2496 and EQ'd both drivers flat prior to applying the xover.
On the cabinets I've done with standard automotive finishes, most didn't crack later on, the seams just sunk in. With something like the epoxy you are just creating a solid surface that won't move with moisture or temperature. MDF is particular to moisture, but I don't know that you could ever fully get 100% seal on inside and outside to keep the moisture out.
Here MDF is about $25 for a 3/4" sheet while baltic birch is about $45. That comes out to about $.78 per sq ft for MDF and $1.80 for baltic birch so it is a good jump. One thing you can do to save material is instead of making one whole piece the "figure 8" type shape is to do it in halves. That way you can line the halves up next to each other and save a lot of material that gets wasted and cut out the middle. You'd have an additional seam on top and bottom then. Stagger the seams for strength.
opc said:
The best options are going to have the best thermal characteristics. Standard old fiberglass insulation, mineral wool, or the "non itch" types of insulation. The Sonic Barrier, Black Hole 5, etc can work to absorb the higher frequency information, but it is expensive. A fully stuffed enclosure of fiberglass insulation will do much more at the frequencies needed than the thin layer of sonic barrier. Also having the inside edges of the enclosure being very irregular also helps a great deal to avoid issues with standing waves and reflections off the back walls.
For most cases I wouldn't do an 18" in a sealed enclosure. While most people think of sealed enclosures as being more accurate, in many cases a vented or passive radiator subwoofer tends to be much lower distortion for a given SPL level. Much less power is applied meaning less thermal issues, less current to the coil means less flux modulation, and less excursion means the driver operates in a much more linear range. Tuning low enough means any difference in group delay vs a sealed enclosure is a not factor.
With the MTM's I did using the same drivers I followed a similar premise. I ended up doing a vented cabinet tuned in the 45hz range. As I only intended them to play to the 60-80hz range, any issues with a peak in group delay were not an issue and I ended up with some extra efficiency from that range up to 150hz or so.
opc said:
That is one of the benefits of the MTM alignment. At the center position vertically you dont' see those issues until higher up in frequency. However move the mic down at say 3" increments and you'll see where they are. You can then translate this 3" increment to an angle based on the 1m distance. It should correspond quite the same at farther distances. For example 6" at 2m. I determined I wanted only a 22" vertical pattern at about 10ft from the speakers and this was achieved fine with the positioning.
This is something I actually plan to do here myself, only in a vented enclosure to save size. The RAAL could be a good combination. We're also looking at doing a baffle of laminated layers that will have a 6.5" waveguide machined right into the center of the MTM. I haven't done any TL's before but I don't see why it wouldn't work well. The TD6.5H has 8mm Xmax so a pair can offer pretty good output. Whether you'd need a subwoofer or not would likely depend on the levels you want to play to. I used to have a pair of Vifa MTM's that were tuned to 32hz and they did quite well without a sub at moderate levels.
John
brucemck2 said:
I'm currently working on an installation for a large theater room. The front LCR each have a pair of sealed TD10S's, the TD6.5M, and the smaller RAAL. The larger RAAL was planned originally but due to limitations in height of the center channel it wasn't possible. The center uses a vertical MT flanked by the TD10's. These are mated to a subwoofer using 8 IB15's. I used the DCX2496 for processing and power is provided by 2 of the Emotiva UPA7 amplifiers plus a QSCPLX3402 for subwoofer power. Currently the sealed TD10S's are being used without a subwoofer. There was plenty of excursion available to EQ them to be used down to 30hz. For home theater duty the IB15's provide output into the mid teens.
Here is a pic, but it is large so I'll just post the link until I can resize and rotate it later.
http://www.aespeakers.com/pics/nick-theater/PICT2831.JPG
John
John, where do you cross them?
Any advantage to MTM with the 6.5s, or is the Raal the limiter?
Would there be any advantages/disadvantages to using the 12s over the 10s?
What's the physical diameter of the 10s? That will determine whether I could use them horizontally on top of my Seaton Submersive which I currently use as a center speaker stand.
If you wanted no crossover between 100hz and 3.5k could that work with this configuration?
Any advantage to MTM with the 6.5s, or is the Raal the limiter?
Would there be any advantages/disadvantages to using the 12s over the 10s?
What's the physical diameter of the 10s? That will determine whether I could use them horizontally on top of my Seaton Submersive which I currently use as a center speaker stand.
If you wanted no crossover between 100hz and 3.5k could that work with this configuration?
opc said:
Hi Owen,
For a 10" midbass I would use 2" Owens 705 fiberglass attached to the cabinet walls. I like to build bracing behind the speaker that tapers the main rear volume and directs the rear reflected wave away from the rear cone. B&W uses a tapered tube for their 5" midranges. A 10" midbass in a steep tapered folded line would work. You might try adjusting the volume of your proto TL and seal off the rear opening for a quick listening test.
A 98db/watt ring radiator tweeter would be nice.
Sign me up for one one!
One of the biggest challenges for high efficiency speakers is finding a non-compression driver that's anywhere over 94 dB.
I think Lynn mentioned in his "Beyond the Ariel" thread that it's kind of a shame there hasn't been more improvement in the area of efficiency. If anything, drivers are reverting to lower and lower efficiencies lately. The newest Scanspeak stuff is a testament to that.
Owen
The problem is not necessarily in the efficiency, but the output levels needed. A 1" dome tweeter for example could be made to be quite efficient but just doesn't have the displacement required to create the levels desired unless you severely limit the bandwidth. Even if put on a waveguide there is only so much a 1" diameter dome can do. A larger dome 2" or 3" ends up becoming a midrange instead of a tweeter. A compression driver though may have a 3" diameter coil and can displace magnitudes more air while still playing through a 1" throat. Between domes and compression drivers, ribbons or planars are really about the only other option. IMO there is a place for both. I generally prefer ribbons at lower volumes but there are definitely occasions where they are not robust enough and don't offer enough SPL.
John
John
As I have 4 spare TD10M laying around at the moment due to a change in plans in the big system build I might just build a version of this with a JBL PT waveguide and one of my stack of compression drivers. Got a couple of xover boards spare, and no shortage of poweramps.opc said:
It's not like I need more speakers, but I can't have them sitting in the box either.
one thing to consider wrt your proposed sort of dog bone shaped cabinet... While it looks interesting, having all edges a set distance away from the center of the woofers is going to give you a more severe diffraction ripple than a more typical or offset arrangement.
opc said:
I have never done the inside also but I imagine you could. You want it thick enough on the outside so that the wood may move some the epoxy never will. You can pour this stuff on one side at a time if desired.
I could see that working quite well as long as you don't mind up giving up some volume for the extra wood required. I haven't gone to those extremes but have done a short tapered midrange chamber, almost like an axe head coming to a point. The more you can eliminate any reflection off the back of the cabinet the better. Putting a V shaped brace behind the driver about 2/3 of the way towards the back of the cabinet can be effective as well. Quite a few ways to do it.
It would be interesting to compare a standard sealed enclosure tightly stuffed with fiberglass to the tapered cabinet and then to an open baffle.
John
It would be interesting to compare a standard sealed enclosure tightly stuffed with fiberglass to the tapered cabinet and then to an open baffle.
John
Hi Guys,
Well, I finally got around to sealing up the end of that TL and taking a few more measurements. I've measured these drivers in enough different ways that I'm definitely starting to get a handle on exactly what the driver's actual FR looks like.
I've attached an FR graph of the TD10M, and I'm pretty happy with the way it turned out. The response is for one driver, in a sealed box with a baffle width of 14". The mic is 0.5 meters away, and spliced at 250Hz to a near-field measurement. The graph was smoothed to 1/6 octave.
Maybe I'm being a little optimistic, but I'm hoping the shallow 3dB dip between 200Hz and 700Hz is due to the baffle, and the little bump at 1kHz could be due to the driver not being flush mounted. I took several off axis measurements, and they seem to confirm that the dip at 1K is definitely diffraction related.
The spike and subsequent junk above 2.5kHz looks like cone breakup, which is now making me think I need to go a little lower with the crossover. That main peak at 2.5kHz has been present in all setups, even when you get 60 degrees off axis, so it's definitely the driver and not anything else.
Ang:
I thought about what you mentioned when I was drawing the box in CAD, but I figured if I used a large enough radius on the edges, then it wouldn't be much of a problem. On the plus side, the tweeter diffraction will be much better as a result of the round dips on either side. I'll report back on how it turns out once the boxes are done.
LineSource:
That box is very interesting. Like John said, I'd want to see it compared side by side with a normal sealed box to make a real call on it. I have the option of making a half sphere at the closed end of my cylindrical sealed box, so if that would be better than just a flat back wall, then maybe it could be a compromise.
John and Nick:
How does that measurement compare to what you guys have measured in your setups? Any comments/ideas on the dip between 200Hz and 700Hz?
Cheers,
Owen
Well, I finally got around to sealing up the end of that TL and taking a few more measurements. I've measured these drivers in enough different ways that I'm definitely starting to get a handle on exactly what the driver's actual FR looks like.
I've attached an FR graph of the TD10M, and I'm pretty happy with the way it turned out. The response is for one driver, in a sealed box with a baffle width of 14". The mic is 0.5 meters away, and spliced at 250Hz to a near-field measurement. The graph was smoothed to 1/6 octave.
Maybe I'm being a little optimistic, but I'm hoping the shallow 3dB dip between 200Hz and 700Hz is due to the baffle, and the little bump at 1kHz could be due to the driver not being flush mounted. I took several off axis measurements, and they seem to confirm that the dip at 1K is definitely diffraction related.
The spike and subsequent junk above 2.5kHz looks like cone breakup, which is now making me think I need to go a little lower with the crossover. That main peak at 2.5kHz has been present in all setups, even when you get 60 degrees off axis, so it's definitely the driver and not anything else.
Ang:
I thought about what you mentioned when I was drawing the box in CAD, but I figured if I used a large enough radius on the edges, then it wouldn't be much of a problem. On the plus side, the tweeter diffraction will be much better as a result of the round dips on either side. I'll report back on how it turns out once the boxes are done.
LineSource:
That box is very interesting. Like John said, I'd want to see it compared side by side with a normal sealed box to make a real call on it. I have the option of making a half sphere at the closed end of my cylindrical sealed box, so if that would be better than just a flat back wall, then maybe it could be a compromise.
John and Nick:
How does that measurement compare to what you guys have measured in your setups? Any comments/ideas on the dip between 200Hz and 700Hz?
Cheers,
Owen
opc said:
Owen,
The rule of thumb for baffle step compensation is
4560 / inches baffle width
4560 / 17 = 268 Hz for the start of 2-pi to 4-pi baffle step compensation.
You may want to download the free "EDGE" simulation program to see what bumps your box shape adds to your measurements.
http://www.tolvan.com/edge/
I would not have expected the 2k Hz big bump...6.75" wavelength
Stuff the cabinet with fiberglass if you have not yet done this measurement.
Yea...have your ears compared the sealed to TL with your favorite music?
Hi Owen,
Can you post a detailed impedance curve of the driver in the sealed cabinet as well? That will give some quick clues as to what is going on.
John
Can you post a detailed impedance curve of the driver in the sealed cabinet as well? That will give some quick clues as to what is going on.
John
- Status
- Not open for further replies.