Hi,
I'm mostly done with the construction phase of My OBLA-33 build but I've got a few questions I'm hoping to get some help with.
This is my first speaker build and my technical knowledge falls far below that of most of the members here so please don't be too harsh.
What I'm building in an open baffle line array in an MTM configuration (thanks for the suggestion Dave Smith). The mid-woofers will of course be dipole but the Dayton PT2C tweeters are mono-pole. Nine tweeters flanked by 12 mid-woofers on each side. I've mounted the drivers in a 1 1/2" thick Maple work surface that measures almost 8' tall.
Here are the drivers:
Tweeters-Dayton Audio PT2C-8 Planar Tweeter
Mid-woofers-6-1/2" Poly Cone Midbass Woofer 4 Ohm
Woofers-Dayton Audio PA380-8 15" Pro Woofer
I know, I should be hearing some serious combing but I don't. And I won't get away from the lobing either but at least it's symmetrical lobing in MTM.
These will be purely actively crossed and controlled by my MiniDSP 4x10 which makes my job a whole lot easier.
I do have a couple questions about the crossover points. Right now I've got everything set to 3rd order at 75Hz and 3200Hz and I'm using a pair of woofers to bring up the bottom end. In my proto-types that have just a single bank of mid-woofers, this sounds pretty good by ear. However, I'm thinking there could be fundamental phase issues as I don't have the knowledge to determine what's happening with phase at any given frequencies.
What are your thoughts on any phase issues regarding my crossover points/slopes?
Here's a couple pics of where I'm at with the build so far with basically only the bases remaining.
I searched the forum rules but couldn't find anything regarding a link to another site so here is a link to my build thread if anyone is interested.
Rob's OBLA-33 (Open Baffle Line Array) project. - AudioKarma.org Home Audio Stereo Discussion Forums
Moderators, if providing this link is in any way inappropriate, please delete everything after the images. Thanks
I'm mostly done with the construction phase of My OBLA-33 build but I've got a few questions I'm hoping to get some help with.
This is my first speaker build and my technical knowledge falls far below that of most of the members here so please don't be too harsh.
What I'm building in an open baffle line array in an MTM configuration (thanks for the suggestion Dave Smith). The mid-woofers will of course be dipole but the Dayton PT2C tweeters are mono-pole. Nine tweeters flanked by 12 mid-woofers on each side. I've mounted the drivers in a 1 1/2" thick Maple work surface that measures almost 8' tall.
Here are the drivers:
Tweeters-Dayton Audio PT2C-8 Planar Tweeter
Mid-woofers-6-1/2" Poly Cone Midbass Woofer 4 Ohm
Woofers-Dayton Audio PA380-8 15" Pro Woofer
I know, I should be hearing some serious combing but I don't. And I won't get away from the lobing either but at least it's symmetrical lobing in MTM.
These will be purely actively crossed and controlled by my MiniDSP 4x10 which makes my job a whole lot easier.
I do have a couple questions about the crossover points. Right now I've got everything set to 3rd order at 75Hz and 3200Hz and I'm using a pair of woofers to bring up the bottom end. In my proto-types that have just a single bank of mid-woofers, this sounds pretty good by ear. However, I'm thinking there could be fundamental phase issues as I don't have the knowledge to determine what's happening with phase at any given frequencies.
What are your thoughts on any phase issues regarding my crossover points/slopes?
Here's a couple pics of where I'm at with the build so far with basically only the bases remaining.
I searched the forum rules but couldn't find anything regarding a link to another site so here is a link to my build thread if anyone is interested.
Rob's OBLA-33 (Open Baffle Line Array) project. - AudioKarma.org Home Audio Stereo Discussion Forums
Moderators, if providing this link is in any way inappropriate, please delete everything after the images. Thanks
The Vertical center to center spacing of the 6.5" drivers should be no more than 1 wavelength at the X-over freq, and or, no more than a diameter of the driver. Speed of sound/6.5" = 2084Hz. It will be also mentioned that if those mids are side by side ctc more than a 1/2 wavelength they will be excessively lobing. Some will say 1/4 wavelength 😀
Yeah, with the 6 1/2" drivers I've introduced combing and lobing at the frequencies I'm crossing them at. Should sound bad but in the real world it doesn't I don't hear any of it on the prototypes. And with adding a second row for MTM the lobing will at least be symmetrical.
I see where you tried 2200Hz x-over. 
Many of the design areas conflict or don't fully complement others. The trick is to blend all the ideas and compromises together the best way.
Since you have DSP, have you thought of time aligning Ts to Ms?
Many of the design areas conflict or don't fully complement others. The trick is to blend all the ideas and compromises together the best way.Since you have DSP, have you thought of time aligning Ts to Ms?
Yes, and even by ear it is very much improved at 2200Hz. I'm very new with the DSP. When you say time aligning you're not talking about the different depth between the tweeters and the woofers are you? I'm a noob but if you would give me even a brief description I think I'll be able to do that. 😀
Not so much. The tweeters I'm using are the PT2C Planar tweeters that they recommend crossing at 3k or higher unless a 3rd order crossover is used then you can cross them at 2500Hz.
In my case, I'm crossing them at 2200Hz with a 4th order slope and because I use 9 per side I think they can handle it. I did crank it up for a while today but usually listen at much lower levels.
Dayton Audio PT2C-8 Planar Tweeter
The first speaker is set up and the second will be done today or tomorrow.
In my case, I'm crossing them at 2200Hz with a 4th order slope and because I use 9 per side I think they can handle it. I did crank it up for a while today but usually listen at much lower levels.
Dayton Audio PT2C-8 Planar Tweeter
The first speaker is set up and the second will be done today or tomorrow.
Last edited:
Yes, the acoustic origin or center should line up with both drivers. This point is not always the diaphragm or the voice coil. The best way to determine it is to measure them with a stationary mic at the listening position. DSP delay could then be used to set the tweeters back approximately 74uSec/inch or whatever. 😀
Good, I have an idea what that is. Which measurements should I take to determine arrival times? I only have REW if one of their measurements will work.
Manufacturer's crossover recommendations are usually based on protecting a driver from damage due to excess excursion. Using an array divides the excursion and allows a lower crossover point for a given SPL. Alternately, you can keep the same crossover point and use the array at higher volume safely.
Doubling the number of drivers allows you to either lower the crossover point by a half-octave or increase the maximum SPL by 6dB. With 4 drivers you can lower the crossover by a full octave or go up by 12dB. In actual use that will not be exactly correct, since multiple sources will produce an interference pattern, with cancellation know as comb filtering.
You are using 9 tweeters in your array, which in theory would allow you to go down 2 octaves (if the driver has usable output there). But the lowest practical crossover point will be near the fs of the driver, where the driver response is rolling off. Those Dayton planars have an fs near 1650 Hz, at a high Q. The effective response goes down to about 1.5kHz, with a steep rolloff below that. You can stay with a 2nd order crossover, which will combine with the 2nd order rolloff of the tweeter to produce a 4th order slope. That will certainly be more than enough to protect the tweeters from over-excursion.
The trick is to make sure everything blends well with the other drivers after you change the crossover point. Just because you could use the tweeters down to 1.5kHz doesn't mean that will sound best in your application.
Doubling the number of drivers allows you to either lower the crossover point by a half-octave or increase the maximum SPL by 6dB. With 4 drivers you can lower the crossover by a full octave or go up by 12dB. In actual use that will not be exactly correct, since multiple sources will produce an interference pattern, with cancellation know as comb filtering.
You are using 9 tweeters in your array, which in theory would allow you to go down 2 octaves (if the driver has usable output there). But the lowest practical crossover point will be near the fs of the driver, where the driver response is rolling off. Those Dayton planars have an fs near 1650 Hz, at a high Q. The effective response goes down to about 1.5kHz, with a steep rolloff below that. You can stay with a 2nd order crossover, which will combine with the 2nd order rolloff of the tweeter to produce a 4th order slope. That will certainly be more than enough to protect the tweeters from over-excursion.
The trick is to make sure everything blends well with the other drivers after you change the crossover point. Just because you could use the tweeters down to 1.5kHz doesn't mean that will sound best in your application.
Thanks so much for that explanation Steve. It's just what I needed to know and even with my limited knowledge I was able to understand it.
I certainly have more room to work with than what I originally thought. I could hear a noticeable improvement lowering the crossover point from 3200Hz to 2200Hz. Maybe there will be further improvement going even lower.
Thanks again!
I certainly have more room to work with than what I originally thought. I could hear a noticeable improvement lowering the crossover point from 3200Hz to 2200Hz. Maybe there will be further improvement going even lower.
Thanks again!
I built several line arrays (original Linus Arrays) with similar planar tweeters (Silver Flute YAG versions from more than 10 years ago). These designs were routinely crossed in the 2200 Hz area. If you measure the woofers in your lines, they likely are rolling off before 2000 Hz or so anyway. You can jock the crossover point between the woofer and tweeter lines to yield best results but 2000 to 2500 Hz would be a good range.
Dr. Robert E. Greene (famous UCLA professor who moonlights as an audio expert in The Absolute Sound magazine) points out the benefits of a symmetrical line array in his review of a MacIntosh XRT28 array in this posting:
http://www.regonaudio.com/McIntosh Supplement.html
Rick Craig designed his Symmetrica Array with dual woofer lines similar to your creation several years ago. His array used smaller woofers which would be highly recommended by others who plan a similar symmetrical array. It makes crossing over without creating off axis issues more achievable.
Gallery ? Selah Audio
http://www.audioroundtable.com/misc/nflawp.pdf
Dr. Robert E. Greene (famous UCLA professor who moonlights as an audio expert in The Absolute Sound magazine) points out the benefits of a symmetrical line array in his review of a MacIntosh XRT28 array in this posting:
http://www.regonaudio.com/McIntosh Supplement.html
Rick Craig designed his Symmetrica Array with dual woofer lines similar to your creation several years ago. His array used smaller woofers which would be highly recommended by others who plan a similar symmetrical array. It makes crossing over without creating off axis issues more achievable.
Gallery ? Selah Audio
http://www.audioroundtable.com/misc/nflawp.pdf
Last edited:
Thanks for your insight and links Jim. I'm still struggling to digest much of what I read but I'm learning.
Not long after beginning this project I realized smaller woofers would have been a better choice. I'll have to look for more suitable drivers on my next go around. I wanted high Qts for OB and I wanted them to go low to around 70Hz-75Hz where my subs kick in.
I don't have a lot of experience with very high end speakers but so far these seem to be working for me. Being my first speaker build I've learned a lot and feel fortunate to have so many experts such as yourself sharing your knowledge.
Thank you!
Edit: I've read your paper.....a several times and have it saved to my favorites. I'd like to think my understanding gets a little better each time I read it.
Not long after beginning this project I realized smaller woofers would have been a better choice. I'll have to look for more suitable drivers on my next go around. I wanted high Qts for OB and I wanted them to go low to around 70Hz-75Hz where my subs kick in.
I don't have a lot of experience with very high end speakers but so far these seem to be working for me. Being my first speaker build I've learned a lot and feel fortunate to have so many experts such as yourself sharing your knowledge.
Thank you!
Edit: I've read your paper.....a several times and have it saved to my favorites. I'd like to think my understanding gets a little better each time I read it.
Last edited:
The MiniDSP will be a big help, especially since you won't have enough gain for the planars to match the woofers. The ability to measure and adjust the DSP will help out quite a bit.
I've used the PT2's in quite a few arrays and the crossover point really should be in the 2.2K-2.5K area. Most of the advantage you gained by lowering the crossover from 3.2K was in the woofer section. A high order 48db slope will help for both the woofers and tweeters.
The planars will couple (gain output) to a greater degree in the 1-2K range which can be misleading, so don't attempt to cross them that low. The good thing is that you can use the DSP to cut the output in that range which will give you cleaner sound and greater headroom. You can then use the PEQ to smooth out the top octaves. The best place to measure is at your listening position and adjust the DSP from that distance.
As you noted it's easier if the woofers are smaller in diameter but since you're running dipole the high Q of the woofer is ideal. Many of the dipole / open baffle arrays I see have lower Q woofers which don't work as well. You could apply some resistive shading to the woofer lines which will help the integration.
I'm not that familiar with REW but you can adjust the high pass delay on the tweeters to optimize the phase integration. You can measure with normal and reverse polarity to check for any issues as well.
I've used the PT2's in quite a few arrays and the crossover point really should be in the 2.2K-2.5K area. Most of the advantage you gained by lowering the crossover from 3.2K was in the woofer section. A high order 48db slope will help for both the woofers and tweeters.
The planars will couple (gain output) to a greater degree in the 1-2K range which can be misleading, so don't attempt to cross them that low. The good thing is that you can use the DSP to cut the output in that range which will give you cleaner sound and greater headroom. You can then use the PEQ to smooth out the top octaves. The best place to measure is at your listening position and adjust the DSP from that distance.
As you noted it's easier if the woofers are smaller in diameter but since you're running dipole the high Q of the woofer is ideal. Many of the dipole / open baffle arrays I see have lower Q woofers which don't work as well. You could apply some resistive shading to the woofer lines which will help the integration.
I'm not that familiar with REW but you can adjust the high pass delay on the tweeters to optimize the phase integration. You can measure with normal and reverse polarity to check for any issues as well.
Hey Rob, long time no see.
I own Ricks Symmetricas and another of his arrays that use the P2s.
The Syms are used everyday I'm home with great delight, a soundstage
you have to hear to believe.
Hope you get them to where you want them, they look very cool....🙂
Perry
I own Ricks Symmetricas and another of his arrays that use the P2s.
The Syms are used everyday I'm home with great delight, a soundstage
you have to hear to believe.
Hope you get them to where you want them, they look very cool....🙂
Perry
I'll attach a sweep I did earlier today. You're right, I should measure from my listening position but this was measured in front of one speaker since the other isn't finished yet. It is crossed at 2200 but with 24db slope . It was suggested to me that the bottom roll off of the tweeters with a 2nd order slope would effectively function like a 4th order slope. I'll have to try different slopes to see where it goes.
Hey PDR, I usually hang out at another site due to so much being over my head here. 🙂
I am very pleased with the results so far. I can already see they have far exceeded my expectations on all fronts. Fortunately, I've been fortunate to have been able to receive very good advice from members here and elsewhere. It has made a huge difference in the outcome and I am thankful for that.
Anyway, here is a screenshot of the sweep after a little bit of smoothing. The 20-70Hz provided by my 15" woofers located elsewhere in the room suffered a bit since the mic was pointed in the opposite direction at the array. Plus, I now realize I need to start from scratch for the augmented, distributed bass. The 15" drivers I have now aren't going to reach as low as I'd like.
My apologies, being a newbie with this stuff the sweep I posted in my last post is not smoothed out. Here is a plot that has been 1/3 octave smoothing. There is more adjusting to be done.
What I meant is that you can measure one array on-axis but be sure to do it at a distance similar to your listening distance. If you measure too close it will skew the results due to proximity effect. Once you get it dialed in then you can fine tune at the listening position.
I would avoid a simple 12db slope on the planars because it will limit your headroom. A simple 2nd/3rd harmonic distortion test will help if you can do that. I would also do driver correction with EQ applied while measuring the separate lines before you do any auto EQ with REW.
I would avoid a simple 12db slope on the planars because it will limit your headroom. A simple 2nd/3rd harmonic distortion test will help if you can do that. I would also do driver correction with EQ applied while measuring the separate lines before you do any auto EQ with REW.
- Status
- Not open for further replies.