Hi all,
Forgive me if some of this is not correct, I'm new to designing loudspeakers and subs but I'm really eager to learn more. I've been dabbling with modeling software for about a week now and playing around with various drivers and enclosures to get a better idea of things.
The application is home theater with my primary interest, from a subwoofer duty being 20hz to 80hz. I'm not looking at this time to hit reference at 15hz or less or anything.
Anyhow, I got really interested in the awesome VBSS design (by MTG) and then went down the rabbit hole of looking for ways to build either efficient or sensitive speakers, or both really, and then use DSP to limit the driver. So I started modeling with inexpensive drivers in various cabinets and came up with my own variation of the idea with similar hardware, but instead of using high powered amps, I'm looking at it from the perspective of... could this run off 50 watts at 8ohms? This makes amplification a breeze. To build SPL, it would be done with several units. This may be flawed thinking, so please, let me know where I'm falling off the cliff if I am!
Here's what I'm currently thinking about and modeling:
Slot vented MDF based enclosure, 24" x 22" x 18.45" with a 20hz tuning frequency. Net internal volume is 4 cubic feet (taking into account the driver displacement, etc, maybe a little more for bracing). I originally had a larger cabinet but the driver excursion was getting out of hand. 20hz is chosen to get that high SPL output with minimal excursion from the driver, so that the woofer is only really having to stress around 30~35hz with its excursion (if I modeled it correctly).
I'm modeling with the GRS 18PT-8 18" Pro Driver (8ohm). It's only $70.
Here's the driver excursion relative to xmax:
I applied a few filters to control the driver. A highpass filter at 17hz. A lowpass filter at 200hz, but this could be at 110~120hz, it won't matter much as I will crossover. PEQ +15db at 20hz with Q of 0.75 to boost the near port output without asking the driver to have high excursion while doing it (I think?). PEQ -8db centered on 55hz with Q of 1.4 to smooth out the peak that was there and further drop excursion. The resulting excursion model in this box at the intended signal input is graphed below, so that the driver doesn't exceed xmax until well under port tuning frequency which I will cut off.
For the slot vent, it's 1.0" x 20.74" x 29.74" in this enclosure to produce the 20hz tuning frequency, but also to limit air velocity from the slot at the intended input signal so that it doesn't exceed 18, conventional wisdom being this will eliminate the chuffing sound.
Those models and filters in place, the following is the SPL graph result.
This is where things get sketchy maybe. I'm modeling this based on a 10 foot distance from the driver face to the listening position and an input of only 50 watts at 8 ohms. I know this is nothing outstanding, but for me this is interesting because for just 50 watts, I'm looking at a nice curve to start working with from just one sub and only 50 watts of power that will be very stable to run. The goal will be to build several of these, 4 of them to start. So the idea is that if I'm getting potentially 96db at 20hz and up, when I add the 2nd sub, I could expect to be close to 100db at 20hz, and add two more and get closer to 105db potentially at 20hz, before room gain (sealed room).
I already have a miniDSP HD 2x4 that I use, so I would use that to align the subs and produce the summation virtual sub and then EQ that for a gentle house curve, based on whatever the 20hz performance ends up at.
I already have an AudioSource AMP100 amp that does 50 watts into 8ohms, two channels, with a nice big power supply. These are inexpensive and pretty solid. I could add another for the 2nd pair of subs. I probably could get way cheaper amplifiers for this, but I'm not certain of quality in tiny class D amps around 50 watts into 8ohms out there with good quality output down to 20hz. I'm totally open to suggestions if you have them!
So end result would be 4 of these enclosures, each running at 50 watts max, for a very low total power draw running full tilt at all times if wanted, stable. Again, from the standpoint of 10 feet distance to the subs (they would be in a sealed room 10 feet from the listening position area) in the home theater room.
If this is modeled properly, and will work (I have my doubts, which is why I'm asking!) the break down is:
Driver x 4 at $70 each, so $280
MDF, glue, screws, terminal binding posts, typical wire, x 4 - $200
Amp x 2 at $150 each, $300
About $780 for this project, unpainted. I would probably just black paint it matte or duratex type look, utilitarian.
Do you all think this is plausible?
Are the models ok or is something horribly wrong?
I do realize there are limits here, and that asking this system to do more than its max is not going to go anywhere. I modeled it for its max use and distance in the room for its purpose.
Any input appreciated.
Very best,
Forgive me if some of this is not correct, I'm new to designing loudspeakers and subs but I'm really eager to learn more. I've been dabbling with modeling software for about a week now and playing around with various drivers and enclosures to get a better idea of things.
The application is home theater with my primary interest, from a subwoofer duty being 20hz to 80hz. I'm not looking at this time to hit reference at 15hz or less or anything.
Anyhow, I got really interested in the awesome VBSS design (by MTG) and then went down the rabbit hole of looking for ways to build either efficient or sensitive speakers, or both really, and then use DSP to limit the driver. So I started modeling with inexpensive drivers in various cabinets and came up with my own variation of the idea with similar hardware, but instead of using high powered amps, I'm looking at it from the perspective of... could this run off 50 watts at 8ohms? This makes amplification a breeze. To build SPL, it would be done with several units. This may be flawed thinking, so please, let me know where I'm falling off the cliff if I am!
Here's what I'm currently thinking about and modeling:
Slot vented MDF based enclosure, 24" x 22" x 18.45" with a 20hz tuning frequency. Net internal volume is 4 cubic feet (taking into account the driver displacement, etc, maybe a little more for bracing). I originally had a larger cabinet but the driver excursion was getting out of hand. 20hz is chosen to get that high SPL output with minimal excursion from the driver, so that the woofer is only really having to stress around 30~35hz with its excursion (if I modeled it correctly).
I'm modeling with the GRS 18PT-8 18" Pro Driver (8ohm). It's only $70.
Here's the driver excursion relative to xmax:
I applied a few filters to control the driver. A highpass filter at 17hz. A lowpass filter at 200hz, but this could be at 110~120hz, it won't matter much as I will crossover. PEQ +15db at 20hz with Q of 0.75 to boost the near port output without asking the driver to have high excursion while doing it (I think?). PEQ -8db centered on 55hz with Q of 1.4 to smooth out the peak that was there and further drop excursion. The resulting excursion model in this box at the intended signal input is graphed below, so that the driver doesn't exceed xmax until well under port tuning frequency which I will cut off.
For the slot vent, it's 1.0" x 20.74" x 29.74" in this enclosure to produce the 20hz tuning frequency, but also to limit air velocity from the slot at the intended input signal so that it doesn't exceed 18, conventional wisdom being this will eliminate the chuffing sound.
Those models and filters in place, the following is the SPL graph result.
This is where things get sketchy maybe. I'm modeling this based on a 10 foot distance from the driver face to the listening position and an input of only 50 watts at 8 ohms. I know this is nothing outstanding, but for me this is interesting because for just 50 watts, I'm looking at a nice curve to start working with from just one sub and only 50 watts of power that will be very stable to run. The goal will be to build several of these, 4 of them to start. So the idea is that if I'm getting potentially 96db at 20hz and up, when I add the 2nd sub, I could expect to be close to 100db at 20hz, and add two more and get closer to 105db potentially at 20hz, before room gain (sealed room).
I already have a miniDSP HD 2x4 that I use, so I would use that to align the subs and produce the summation virtual sub and then EQ that for a gentle house curve, based on whatever the 20hz performance ends up at.
I already have an AudioSource AMP100 amp that does 50 watts into 8ohms, two channels, with a nice big power supply. These are inexpensive and pretty solid. I could add another for the 2nd pair of subs. I probably could get way cheaper amplifiers for this, but I'm not certain of quality in tiny class D amps around 50 watts into 8ohms out there with good quality output down to 20hz. I'm totally open to suggestions if you have them!
So end result would be 4 of these enclosures, each running at 50 watts max, for a very low total power draw running full tilt at all times if wanted, stable. Again, from the standpoint of 10 feet distance to the subs (they would be in a sealed room 10 feet from the listening position area) in the home theater room.
If this is modeled properly, and will work (I have my doubts, which is why I'm asking!) the break down is:
Driver x 4 at $70 each, so $280
MDF, glue, screws, terminal binding posts, typical wire, x 4 - $200
Amp x 2 at $150 each, $300
About $780 for this project, unpainted. I would probably just black paint it matte or duratex type look, utilitarian.
Do you all think this is plausible?
Are the models ok or is something horribly wrong?
I do realize there are limits here, and that asking this system to do more than its max is not going to go anywhere. I modeled it for its max use and distance in the room for its purpose.
Any input appreciated.
Very best,
Yea, sorry, not sure how that answers any of the questions above. But thanks for sharing the interesting build. Looks like it will generate high SPL for lower power than less efficient builds, I suppose from being a horn design?
Ultimately I'm looking to do 4 subs to produce higher SPL but also more specifically for even room response at the seating (home theater) to handle the room nulls.
Very best,
Ultimately I'm looking to do 4 subs to produce higher SPL but also more specifically for even room response at the seating (home theater) to handle the room nulls.
Very best,
My design cures your "PEQ -8db centered on 55hz with Q of 1.4 to smooth out the peak that was there and further drop excursion."
Ok thanks, I'll study this further, it's beyond me at the moment as I'm new to this. Thanks again for the build.
Very best,
Doh, so I just learned the PEQ filters applied are not accurate when looking at xmax, SPL and velocity, so the above graphs are not going to be real and so that kills this potential build. Alas, good to learn.
Very best,
Very best,
Your design looks quite doable to me. Check the Monoprice Unity amps if you're looking for something class D, rock-solid and small and fairly cheap.
I don't really want to take up the OP's thread with discussion about another design, but the upper cutoff of your design is wayyyyy too sharp. I've experimented with a few designs with extremely high cutoff slopes in the past, and they just didn't sound right to me, when combined into a full-range system. Even the 30dB/octave cutoff of the subwoofer system in my car is a little too steep for my liking, but unfortunately I can't reduce it because it's the best result I can get given the car's acoustics.
Hi,
It turns out that after I remove the PEQ filters so that the SPL, excursion and velocity graphs were correct, the output of course went away with the mere 50 watts. That was the problem, leaving those filters on. To get the SPL back takes more power of course, but then excursion goes way past xmax (distortion). Using any more power than 50 watts puts me on different tier amps and so the value of this just went down big time. I just didn't realize the issue of applying PEQ filters and that the SPL/Excursion/Velocity graphs wouldn't calc the physical components first, then add the filters on top, instead it just puts it all together at the same time which doesn't work. Live and learn!
Very best,
So I moved on to another low power near field build.
Using the GRS 12SW-4HE 12" 4ohm driver, it's dirt cheap and has high xmax and is ok sensitivity and so it responds really well to a ported box with a low frequency tuning with no filters applied. It can take 250 watts RMS, but I'm going to be modeling it for my workstation, so it will be near field and just 2.5 to 3 feet away from my ears under my desk and it will share signal with my Philharmonics that I built that I use for monitors on stands at my workstation; and I think I will model it on 100 watts at 4ohms at this range.
Here's the enclosure:
24" x 20" x 19.63" (3.7 cubic feet net internal volume) with a 20hz tuning port frequency (to limit excursion needs down at 20hz and increase output at 20hz):
Here's the excursion graph relative to xmax, the driver does not go past xmax until well under 20hz, which I will cut off anyways.
The slot vent is 1.25" x 18.74" x 36.66" and so air velocity doesn't exceed 18, hoping that will keep no port noise.
Here's the final SPL graph on this. The volume of the enclosure and the port tuning of 20hz results in a nice 20hz to 80hz graph without any EQ applied, with this cheap driver. Keep in mind this is relative at 2.5~3 foot distance (near filed) and 100 watts of power max.
This one will cost $60 for the driver, $50 for a basic little class-D 100 watt at 4ohm amplifier with an active crossover to limit it down to 80~100hz or so and about $50~60 in materials for the enclosure and binding posts. So I should be able to build this potentially for about $160 give or take and have a pretty solid reference level 20hz to 80hz near field sub that doesn't take much power.
Any thoughts on this one?
Very best,
Using the GRS 12SW-4HE 12" 4ohm driver, it's dirt cheap and has high xmax and is ok sensitivity and so it responds really well to a ported box with a low frequency tuning with no filters applied. It can take 250 watts RMS, but I'm going to be modeling it for my workstation, so it will be near field and just 2.5 to 3 feet away from my ears under my desk and it will share signal with my Philharmonics that I built that I use for monitors on stands at my workstation; and I think I will model it on 100 watts at 4ohms at this range.
Here's the enclosure:
24" x 20" x 19.63" (3.7 cubic feet net internal volume) with a 20hz tuning port frequency (to limit excursion needs down at 20hz and increase output at 20hz):
Here's the excursion graph relative to xmax, the driver does not go past xmax until well under 20hz, which I will cut off anyways.
The slot vent is 1.25" x 18.74" x 36.66" and so air velocity doesn't exceed 18, hoping that will keep no port noise.
Here's the final SPL graph on this. The volume of the enclosure and the port tuning of 20hz results in a nice 20hz to 80hz graph without any EQ applied, with this cheap driver. Keep in mind this is relative at 2.5~3 foot distance (near filed) and 100 watts of power max.
This one will cost $60 for the driver, $50 for a basic little class-D 100 watt at 4ohm amplifier with an active crossover to limit it down to 80~100hz or so and about $50~60 in materials for the enclosure and binding posts. So I should be able to build this potentially for about $160 give or take and have a pretty solid reference level 20hz to 80hz near field sub that doesn't take much power.
Any thoughts on this one?
Very best,
Really nice, thanks for sharing! It's beyond me at this stage, but something I will enjoy learning!
Very best,
Hey Brian, you stated that it was nice going back to a simple BR enclosure. Do you think that 8" BR sound better than the TH's you built?
The THs were with 12" and above drivers, so it really isn't fair to compare them with an 8" BR, LOL.
I did think though that the 12" BR (POC6) integrated better with the tops than the 12" TH that I built using the same driver. It did not go down as low however (50 Hz vs 40 Hz), and that was also pretty noticeable.
I did think though that the 12" BR (POC6) integrated better with the tops than the 12" TH that I built using the same driver. It did not go down as low however (50 Hz vs 40 Hz), and that was also pretty noticeable.
Update,
Ordered some drivers. One of them is the GRS 12SW-4HE driver mentioned above. Really fun driver for such a tiny price. Also ordered a couple of Dayton MX15-22's to do some large boxes with (but those will have much higher power and will be building 4 of those for the theater room). This little 12" GRS driver will go in a smaller box (well, still large as pointed out above!) and is for my office. Replacing an ancient commercial 12" Polk sub from over a decade ago that has been fine, but wanted to update things and refresh.
I'll do this one's enclosure first for practice, 3/4th inch MDF from local Lowes ($57 a sheet). Will be doing the cuts with my Jigsaw and may attempt to use my router for the driver hole. I will do a window brace I think, or thick dowels for bracing?
Re-did the enclosure, 22" x 16" x 22.63" at 3 cubic feet net internal, tuned to 21hz now to make it easier on the port needs. I reduced it to shave space and it will fit under my desk and out of the way. I also wanted to be able to turn it up if needed, if I chose to give it more power and sit farther away in another use. But for now it's meant for under the desk. 2.5~3 feet listening distance, about 100 watts going to it max.
For Excursion, I modeled it at 250 watts to ensure if I pushed it, it wouldn't go past xmax and add distortion. It does below 20hz, but that's ok, I will not hear that.
For air velocity at the port, 1.36" x 14.74" x 35.23" gives me no chuffing, even at 250watts, to 20hz. Below 20hz it will start to flub chub, but I will cut if off past 20hz and won't hear that.
So here's the SPL graph with no EQ, modeled at 3 feet and 100 watts (can go as far as 250 watts, the driver's RMS, but will be fine at 100 watts, maybe less even).
Very best,
Ordered some drivers. One of them is the GRS 12SW-4HE driver mentioned above. Really fun driver for such a tiny price. Also ordered a couple of Dayton MX15-22's to do some large boxes with (but those will have much higher power and will be building 4 of those for the theater room). This little 12" GRS driver will go in a smaller box (well, still large as pointed out above!) and is for my office. Replacing an ancient commercial 12" Polk sub from over a decade ago that has been fine, but wanted to update things and refresh.
I'll do this one's enclosure first for practice, 3/4th inch MDF from local Lowes ($57 a sheet). Will be doing the cuts with my Jigsaw and may attempt to use my router for the driver hole. I will do a window brace I think, or thick dowels for bracing?
Re-did the enclosure, 22" x 16" x 22.63" at 3 cubic feet net internal, tuned to 21hz now to make it easier on the port needs. I reduced it to shave space and it will fit under my desk and out of the way. I also wanted to be able to turn it up if needed, if I chose to give it more power and sit farther away in another use. But for now it's meant for under the desk. 2.5~3 feet listening distance, about 100 watts going to it max.
For Excursion, I modeled it at 250 watts to ensure if I pushed it, it wouldn't go past xmax and add distortion. It does below 20hz, but that's ok, I will not hear that.
For air velocity at the port, 1.36" x 14.74" x 35.23" gives me no chuffing, even at 250watts, to 20hz. Below 20hz it will start to flub chub, but I will cut if off past 20hz and won't hear that.
So here's the SPL graph with no EQ, modeled at 3 feet and 100 watts (can go as far as 250 watts, the driver's RMS, but will be fine at 100 watts, maybe less even).
Very best,
I think you should try out a simple sealed alignment with that 12" driver, with some EQ at the low end to flatten the response out (even better would be if the signal was processed with a DSP with dynamic EQ ability, so it could reduce the amount of boost at higher input levels). Smaller box, and with a 100W on tap, it should be more than enough bass if you're listening 3 feet away, unless you're a bass head in disguise 🙂.
I'll model it! I tried a similar model, and mostly went back to large ported simply for the 20hz SPL output. I am a bit of bass head, but not in that I want loud crazy bass, but actually just really enjoy specifically that 20~30hz range. I actually dislike heavy SPL from 40~80hz. Kind of a weird preference. This sub will be put inline with an all analog setup I use with Philharmonic monitors that I assembled and wired running from a first gen Emotiva A-100 Mini-X amplifier and using its line out to the sub's amp (which will have an active cross over). I won't have phase control or DSP or anything unless I add more equipment to this particular setup.
I do like the idea of potentially doing a sealed build too. Will model it up tonight.
Very best,
DSP can really make a difference. I use a Dayton DSP408 for my desktop system, wired between my PC's soundcard output and the input to the system's amplifier. Dayton also sells a cheaper DSP unit specifically designed for subwoofer duty.