Site Links
Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Quick Links
Categories
In this Discussion
Using drivers in "up-firing" orientation - thoughts?
Hello All! Hope you are doing well. I'm working on a design for a two-woofer system that will be used in low-slung box. It's the woofer part of the loudspeaker only. One configuration that I have been considering is placing both woofers on the top of the cabinet, which measures 60x120cm (about 2x4 feet).
It's been a very long time since I have considered using a driver in a non-horizontal orientation. Modern drivers do not have suspensions as soft as those from long ago, so sag is probably not too much of a concern. I am more worried about increased distortion in this orientation. Has anyone looked into that in DIY circles?
Would love to hear some thoughts on this topic. Thanks and happy holidays!
-Charlie
P.S. the driver is the SB Acoustics SB23MFCL45-4
Comments
If they are being used as subwoofers playing under 200hz in a sealed cabinet I doubt distortion would be affected. At higher frequencies, the horizontal orientation will be complicated to work with.
Distortion from difference in apparent motor force due to gravity?
I am hoping to use the drivers from 30Hz to 300Hz.
Yes, the effect is due to gravity acting on the moving mass. It assists down acceleration and resists up acceleration, so it is IIRC increasing even order distortion. But by how much I do not know.
Maybe this is not worth worrying about with modern drivers...
Resrc - How to Tell if a Woofer Can Be Mounted Up- or Down-Firing. In PE tech resources
Sounds like an omni woofer project. I've been playing around a little with the Dayton Audio GF180-4, and now with the buyout SIG180-4. Both firing straight up with the OB mids and tweeter directly above the woofers. Thread here:
https://diy.midwestaudio.club/discussion/2777/pwrryds-omni-ob-whatever-thread/p1
I've let other projects get in the way and haven't even taken the new measurements yet. Initial playing around with the cheap GF180, no issues playing them to 500 Hz. I have some days off work so I might jump back on this project before the new year.
@Eggguy I use Unibox for quick TS modeling and it has that built it. Comes to 0.97 e.g. 3% sag, so no problem there.
@PWRRYD This setup uses two woofer that are separated by about 12" to 16" C-to-C and operated as a dipole at "high" frequency, e.g. above 200Hz. A filter is used to manipulate the phase of one driver, which causes the radiation pattern to change from dipole to cardioid to monopole (in phase) with decreasing frequency. I have modeled this about 10 years ago, then better more recently and I finally want to build it. The advantage is that you get a dipole pattern that will match that of a dipole mid, but then the output becomes monopole at LF and you get room pressurization without dipole losses. I am just trying to decide on where to place the two drivers and two PRs on the cabinet so that they have the correct spacing and are still ground coupled. Putting them on the sides at the ground makes them a little too far apart (over 24"). Then there is the option to put them both on the top, facing up. The third option I am considering is making two separate cabs, each half the volume (since each will only have one driver and one PR), which will be like two cubes. The cabs would be placed with the drivers facing each other with about 12" of air in between. This creates a "slot" that will likely resonate but I would have the flexibility to move the cabs and drivers closer together or farther apart. I can't actually build anything until the springtime due to no heat in the work area, so I have too much time to think!
Yeah way different than what I'm playing around with.
So above 200 Hz you should expect to get some directionality based on the placement of the woofers. I'd be concerned about the room interactions, which would make the design specific to one location.
The two woofer boxes gives you the most flexibility. A typical cardioid woofer doesn't deliver the same alignment? A drawing would probably help.
Hey Charlie, are you in Germany now?
That is what his flag designation is over on DIYaudio.
InDIYana Event Website
Yes, have been over here since May 1.
The two-woofer setup will be operated as a dipole in the upper part of its passband (eg 150Hz and above). If you take two monopoles, reverse the phase of the "rear" one and move them apart in space this gives rise to a dipole radiation pattern. Despite using two monopoles, you get more and more cancellation at the on axis listening location with decreasing frequency. This is because the separation between the two sources that are 180 degrees out of phase becomes less and less of a wavelength at lower and lower frequencies. Thus you get the classic 6dB per octave fall off due to the interaction between the output of the two woofers. As you get below 100Hz you add to that the rolloff of each driver's LF alignment, e.g. sealed, vented, etc. For example if it is a sealed box, the rolloff would increase from 6dB to 18dB/oct with the addition of the CB rolloff. These LF losses can really add up. You can equalize the response within the 6dB/oct rolloff region to be flat via passive or active means, but the losses eat up Xmax and sensitivity.
It would be nice if you could avoid the LF losses and it turns out that if you can just change the radiation pattern into two monopoles you do exactly that. So I have come up with a way to use an allpass filter that provides the correct phase rotation to transition over from dipole at HF to two monopoles at LF. The monopoles are in relative phase but the front one lags by 360 degrees at the lowest frequencies, below 60Hz.
It is possible to just sum a monopole and dipole to get a cardioid. If you cross over between the two this is another way to transition from dipole to monopole. The problem is that this only applies to a dipole that has already been equalized to be flat and to have the same gain as the monopole before summing them. This puts a lot of demand on the dipole which means either a large driver or a large baffle. So I am investigating this other way with the two woofers and the AP filter. In either case you want the system to have a dipole radiation pattern where you cross over to the midrange.
By combining the multi-pattern woofer system with a 6" class cone midrange (nude) and an AMT dipole tweeter I can make a system with 3 bands. Technically it is a 4-way since each woofer has a different input signal due to the AP filter used on one of them. But this will allow me to put the two woofers at the floor, and then place the mid and tweeter above with nothing in between but air. This helps the front wall reflection return to the listener without diffracting off of some big ass baffle, which IMO ruins the imaging effect of a nice dipole system and is why I am not in favor of large planar dipole/OB systems, e.g. PAP, etc.
That's sounds pretty bitchin, Charlie. To bad your not around to demo at one of the events.....if it all pans out.
Very well elucidated as usual, Mr. Laub.
(I actually understood it).
I assume Charlie is organizing Deutschland DIY for us.