I'm interested in putting an H frame subwoofer in my living room. I've had a 15" driver for several years that I would like to inexpensively utilize. Before you ask, the sub has a low qtc, and fairly high excursion(supposedly not great for H-frame) though the Fs is a nice 18.7Hz. I'm looking for accurate deep bass, but not necessarily loud. (I already have a media room setup for intense listening) This room would be more jazz/classical/low key listening WITH other humans. I've read up on Martins site, linkwitz, musicandesign, and all over the forums. I have a fair working knowledge on theory behind these designs but no real world exprience.
The room is about 15feet square by 9feet tall. Hardwood
There is a 4' diam circular marble coffee table on a wooden pedestal in the center of the room...Stone is almost 100lbs.
My thought is to remove the pedestal (not attached whatsoever) and build an H frame that will become the new base
The stone on top should greatly remove vibration and baffle movement
Here goes...
Should I expect any extension in lower frequencies compared to an h frame without the stone top?...Will the 4' diam circle on top act as a large baffle allowing for longer wavelengths to "come through" and be represented in the room?
Since I have a bit of room to work with under the table top, will increasing the size of the "H" part of the baffle help longer wavelengths?
I know dipoles are best taken advantage of when removed from walls. I've seen tests online that seem to say the closer to the middle of the room the better. This table would be just about dead center. I'm hoping this will help disperse to the whole room as opposed to sweet spot? or will the dipole action create lobes where sound can and can't be heard cohesively?
There is seating all around this table (as is usual with a coffee table😉) Any ideas of the SQ while sitting that close to the system? I know H frames tend to drop fairly quickly as you move farther, but is this too close?
Please also see the attached photo...I've looked all over for flared or tapered h frames (as Martin briefly mentions at the end of his article) and what effects, if any this has on SQ, and have not found anything very solid. In the photo I've flared the wing baffles and truncated them respectively. You can see a representation of the marble table top as well. The baffle/wing sizes are simply free handed mock ups and are just a guide to express the idea. I've seen it mentioned that the taper/flare doesn't make much of an impact. (probably not as much as my high Q and excursion)
I know I've brought up a lot here but am just trying to get a taste of others' experience with ANY of these issues that go beyond the standard "H".
Since they're relatively simple to construct I'll eventually just build the darn thing and let you all know MY findings.
Thanks All!
The room is about 15feet square by 9feet tall. Hardwood
There is a 4' diam circular marble coffee table on a wooden pedestal in the center of the room...Stone is almost 100lbs.
My thought is to remove the pedestal (not attached whatsoever) and build an H frame that will become the new base
The stone on top should greatly remove vibration and baffle movement
Here goes...
Should I expect any extension in lower frequencies compared to an h frame without the stone top?...Will the 4' diam circle on top act as a large baffle allowing for longer wavelengths to "come through" and be represented in the room?
Since I have a bit of room to work with under the table top, will increasing the size of the "H" part of the baffle help longer wavelengths?
I know dipoles are best taken advantage of when removed from walls. I've seen tests online that seem to say the closer to the middle of the room the better. This table would be just about dead center. I'm hoping this will help disperse to the whole room as opposed to sweet spot? or will the dipole action create lobes where sound can and can't be heard cohesively?
There is seating all around this table (as is usual with a coffee table😉) Any ideas of the SQ while sitting that close to the system? I know H frames tend to drop fairly quickly as you move farther, but is this too close?
Please also see the attached photo...I've looked all over for flared or tapered h frames (as Martin briefly mentions at the end of his article) and what effects, if any this has on SQ, and have not found anything very solid. In the photo I've flared the wing baffles and truncated them respectively. You can see a representation of the marble table top as well. The baffle/wing sizes are simply free handed mock ups and are just a guide to express the idea. I've seen it mentioned that the taper/flare doesn't make much of an impact. (probably not as much as my high Q and excursion)
I know I've brought up a lot here but am just trying to get a taste of others' experience with ANY of these issues that go beyond the standard "H".
Since they're relatively simple to construct I'll eventually just build the darn thing and let you all know MY findings.
Thanks All!
Attachments
The tabletop will certainly change things (everything you do will have an effect) but as far as the lowest frequencies go, the distance from one side of the cone to the other is what matters, and that doesn't change by adding a tabletop, the distance around the sides is still the same.
Absolutely the longer wavelengths will benefit from and extended H, but the problem with this is that the H frame's upper usable bandwidth is defined by the H frame's first resonance. You might have noticed that MJK's H frame paper shows a fairly shallow cavity on both sides of the driver. Make that cavity bigger (deeper) and you will have big problems with resonances inside your desired passband.
This is all simulated easily with available software.
Dipoles have a figure 8 dispersion pattern. You've probably seen pictures of this radiation pattern in your studies. What this means is that the seats on either end of the H frame (the seats where you can look right at the driver and see it) will have nice bass. The seats to the sides (the ones where the driver is not visible due to being obstructed by the side wall) will have little or no bass. At 90 degrees off axis a dipole is useless, this is the cancellation zone. If you want to be able to seat people 360 degrees around the speaker and have all guests enjoy the same presentation, dipole is not what you want.
The effect of flares is easily simulated. You will find that a flare will change the frequency of your first resonance peak. This can be very useful if you want to make a deep(er) H frame.
I forgot to comment on this part. If you have dsp capability this driver will do fine. If you don't this driver will not work, you'll never get anything resembling flat response out of it.
MJK's papers reveal how you can use driver choice as a means of creating a very simple system with inexpensive passive components. The alternative is a much more complex dsp based system. Either option is fine as long as you know what you are getting into.
Thank you, Just a guy.
---It makes sense that the largest wavelength represented will be the smallest distance from driver-driver. The table will help with vibration but wont act as a larger baffle.
---I do remember MJK's very shallow (under 8" I believe) wings defining his crossover point. Mine would be roughly the same depth based on 1/4 wavelength. This sub is rated to a high of 250Hz. I'm going to end up crossing somewhere between 150-200 I think. I'm building a Mono 16 driver (3") quasi line array for mids/highs to run floor to ceiling in the corner. The enclosure will run full height, but the drivers will be as close as possible to each other. I'll tune that to run down to ~120Hz.
---The 'lobing' of the sound is what I was afraid of. Perhaps this is not the correct direction for full, even dispersal of sound among a ring of people; although mingling people might work. Since it is a simple build I will try anyway and see my impressions of the atmosphere. I'm just itching to put this driver to work since I have it on hand.
---I'll model my mock-ups and see how the flare/taper might affect SQ.
---By DSP do you simply mean digital equalization? My Yamaha receiver does have equalization for all channels, but only a crossover point for the sub. I have an external equalizer but think the lowest sliders begin at 60Hz, 150Hz, 400Hz, Etc.
I appreciate your time and consideration on my post. I'm always happy coming to this community!
---It makes sense that the largest wavelength represented will be the smallest distance from driver-driver. The table will help with vibration but wont act as a larger baffle.
---I do remember MJK's very shallow (under 8" I believe) wings defining his crossover point. Mine would be roughly the same depth based on 1/4 wavelength. This sub is rated to a high of 250Hz. I'm going to end up crossing somewhere between 150-200 I think. I'm building a Mono 16 driver (3") quasi line array for mids/highs to run floor to ceiling in the corner. The enclosure will run full height, but the drivers will be as close as possible to each other. I'll tune that to run down to ~120Hz.
---The 'lobing' of the sound is what I was afraid of. Perhaps this is not the correct direction for full, even dispersal of sound among a ring of people; although mingling people might work. Since it is a simple build I will try anyway and see my impressions of the atmosphere. I'm just itching to put this driver to work since I have it on hand.
---I'll model my mock-ups and see how the flare/taper might affect SQ.
---By DSP do you simply mean digital equalization? My Yamaha receiver does have equalization for all channels, but only a crossover point for the sub. I have an external equalizer but think the lowest sliders begin at 60Hz, 150Hz, 400Hz, Etc.
I appreciate your time and consideration on my post. I'm always happy coming to this community!
Dipole - Transmission Line / Quarter Wave
I just thought of another potential.
Separating the front and back waves with a specific distance delineates what frequency we can observe.
What about a transmission line where the driver is located in the middle of the line, completely separating the front line from the rear.
Take a look at the attachment.
If both red and blue lines (including distance between the exit ports) are 7.8 feet long, the total distance that the front and rear waves have traveled when they meet will be 15.6 feet, yielding the 18Hz Fs of this particular driver. the lines are 1/2 sd of the driver to load similar to a W-Frame dipole (mostly to save space to fit under my round 4' table top.)
I did some brief searching and did not find any mention of this style. IS THIS SIMPLY A BANDPASS BOX?
I feel like its different...thoughts?
I just thought of another potential.
Separating the front and back waves with a specific distance delineates what frequency we can observe.
What about a transmission line where the driver is located in the middle of the line, completely separating the front line from the rear.
Take a look at the attachment.
If both red and blue lines (including distance between the exit ports) are 7.8 feet long, the total distance that the front and rear waves have traveled when they meet will be 15.6 feet, yielding the 18Hz Fs of this particular driver. the lines are 1/2 sd of the driver to load similar to a W-Frame dipole (mostly to save space to fit under my round 4' table top.)
I did some brief searching and did not find any mention of this style. IS THIS SIMPLY A BANDPASS BOX?
I feel like its different...thoughts?
Attachments
Let's back up a bit and start with your goals. What are you trying to achieve? It seems that you want to try dipole and you want to use the driver you already have.
If these two goals are accurate, this new design probably isn't a good idea. People usually want to use dipole to bypass box👍 effects, aka resonances. This design will be full of resonances inside the passband and will need a lot of stuffing to get anything resembling flat response.
If you provide a list of goals including desired bandwidth, size, spl requirements, etc, it might be easier to guide you into something that matches these goals. There are all kinds of possibilities.
FWIW, this new design can be easily simulated too, if you post up the t/s parameters I might do a couple of sims for you if I can find the time, although it is always best if you do them yourself so you can learn what happens when you change things.
If these two goals are accurate, this new design probably isn't a good idea. People usually want to use dipole to bypass box👍 effects, aka resonances. This design will be full of resonances inside the passband and will need a lot of stuffing to get anything resembling flat response.
If you provide a list of goals including desired bandwidth, size, spl requirements, etc, it might be easier to guide you into something that matches these goals. There are all kinds of possibilities.
FWIW, this new design can be easily simulated too, if you post up the t/s parameters I might do a couple of sims for you if I can find the time, although it is always best if you do them yourself so you can learn what happens when you change things.
It might be tough to force the first resonance above 250 hz even with a shallow cavity.
With a sub very close to the listening position and the line array very far away you might need a bit of delay on the sub to match properly.
There are ways to deal with that. For example, putting stuffing in one side of the H makes the design cardiod instead of dipole, this will help with the off axis cancellation. But uneven dispersion in a small room is a given anyway due to room modes and such.
By dsp I mean digital signal processing. This is more than just eq, although it does include eq. For example, a good dsp will include delay, which you will probably need if your sub and line array are different distances from the listening position. Additionally the dsp will allow for filters that can correct your response curve in ways that a graphic eq can't.
- Status
- Not open for further replies.