For amps and DSP processor on this project, I am leaning towards the MiniDSP PWR-ICE 250 amps.
It is 2x250watts @ 4ohms. It has dual analog inputs, and it allows a stereo configuration where input 1 goes to channel 1, input 2 goes to channel 2. And then I can EQ both channels independently. Instead of having a right and a left on each, I will have 2 rights on one and 2 lefts on the other.
But for that to work, I need 2 right channels and 2 left channels for these new speakers. And then I need a 3rd right and left for my main speakers. So need a line level splitter. I haven't found that yet.
Originally I was thinking I would use the MiniDSP 4x10 unit, but I'm finding it has been out of stock for a long time, and there is no estimate of when it will be back in stock.
Here's an amp and driver integration question: Is it better to use a lower sensitivity driver that will just reach it's max excursion at the amp's rated power? Or use a driver with higher sensitivity, that is excursion limited well below the amp's rated power? Both reach the same SPL, and have nearly the same low frequency extension. And they work in roughly the same enclosure.
I'm curious what you guys think! Will both provide the same sound quality, or will there be audible differences? Should I worry about overdriving woofers when I'm using DSP? (It should have some type of limiter, but I'm not sure how adjustable it is.)
It is 2x250watts @ 4ohms. It has dual analog inputs, and it allows a stereo configuration where input 1 goes to channel 1, input 2 goes to channel 2. And then I can EQ both channels independently. Instead of having a right and a left on each, I will have 2 rights on one and 2 lefts on the other.
But for that to work, I need 2 right channels and 2 left channels for these new speakers. And then I need a 3rd right and left for my main speakers. So need a line level splitter. I haven't found that yet.
Originally I was thinking I would use the MiniDSP 4x10 unit, but I'm finding it has been out of stock for a long time, and there is no estimate of when it will be back in stock.
Here's an amp and driver integration question: Is it better to use a lower sensitivity driver that will just reach it's max excursion at the amp's rated power? Or use a driver with higher sensitivity, that is excursion limited well below the amp's rated power? Both reach the same SPL, and have nearly the same low frequency extension. And they work in roughly the same enclosure.
I'm curious what you guys think! Will both provide the same sound quality, or will there be audible differences? Should I worry about overdriving woofers when I'm using DSP? (It should have some type of limiter, but I'm not sure how adjustable it is.)
As a speaker motor heats up, DC Re increases, and motor strength falls off some. When the driver is warm, it prefers a larger box. Suspension when warm also has a temporary stiffness change, lowering the Fs.
There are many factors that affect the final outcome, but the moral of the story is to make the box a bit towards the larger side, since it will allow you to use a larger eqivalent diameter port ( or slot ) with less overall length than a smaller volume box.
Line level splitter = Y cables, either RCA or XLR depending on the rest of your system.
Spend some time perusing the Data-Bass tests before you commit to a design and spend money.
There are many factors that affect the final outcome, but the moral of the story is to make the box a bit towards the larger side, since it will allow you to use a larger eqivalent diameter port ( or slot ) with less overall length than a smaller volume box.
Line level splitter = Y cables, either RCA or XLR depending on the rest of your system.
Spend some time perusing the Data-Bass tests before you commit to a design and spend money.
Here is the updated design. It is bigger than before, as expected with the 15" low frequency subwoofer.
The subwoofer is flat down to 20hz in this simulation, with an F3 of 17hz, which is also the tuning frequency.
In the simulation at peak power, the low frequency driver hits it's Xmax @ 17hz (250 watts in 4 ohms, 108.5db), and the port velocity is ~35m/s @ 23hz @111 db. It drops to 8.6 m/s @ 30hz at 111db. Average SPL is 111db at 250 watts.
As Diyuser2010 mentioned, the actuall motor force will decrease due to heat build up, so I expect these numbers to be somewhat lower.
If you guys think that 35m/s figure is too high, keep in mind it is at theoretical max power of 250 watts, and only at 23hz. There isn't much content at 23hz, and it'll be so loud at max power I will never crank it that high! Keep in mind there are 4 15" woofers, each playing at 111db. That must be a +12db gain for 123db! (I feel like I should make the port smaller and shorter, designing it for actual use case, not the max p ower. It might be good if it starts making noise at the max power, because that's a warning that you're pushing it too hard?)
The woofer also looks pretty good to me. It ends up with an F3 of 33hz in this volume, but I am planning a high pass filter @50hz because it doesn't have enough excursion to go much lower. With the high pass filter, it'll be flat down to 50hz, and play at an average SPL of 112db @ 250 watts. Again, that's louder than I need, but I feel like the two drivers match up pretty well as far as power handling and SPL in this configuration.
The subwoofer is flat down to 20hz in this simulation, with an F3 of 17hz, which is also the tuning frequency.
In the simulation at peak power, the low frequency driver hits it's Xmax @ 17hz (250 watts in 4 ohms, 108.5db), and the port velocity is ~35m/s @ 23hz @111 db. It drops to 8.6 m/s @ 30hz at 111db. Average SPL is 111db at 250 watts.
As Diyuser2010 mentioned, the actuall motor force will decrease due to heat build up, so I expect these numbers to be somewhat lower.
If you guys think that 35m/s figure is too high, keep in mind it is at theoretical max power of 250 watts, and only at 23hz. There isn't much content at 23hz, and it'll be so loud at max power I will never crank it that high! Keep in mind there are 4 15" woofers, each playing at 111db. That must be a +12db gain for 123db! (I feel like I should make the port smaller and shorter, designing it for actual use case, not the max p ower. It might be good if it starts making noise at the max power, because that's a warning that you're pushing it too hard?)
The woofer also looks pretty good to me. It ends up with an F3 of 33hz in this volume, but I am planning a high pass filter @50hz because it doesn't have enough excursion to go much lower. With the high pass filter, it'll be flat down to 50hz, and play at an average SPL of 112db @ 250 watts. Again, that's louder than I need, but I feel like the two drivers match up pretty well as far as power handling and SPL in this configuration.
Just a quick note that, with typical music program and peak power inputs of approx 200w, heat buildup is unlikely to be an issue.
Chris
Chris
Generally, the level of the music will be so far above the port wind noise it will not be heard, but at higher velocities the port low frequency output will be reduced. You won't notice the noise, but you may notice less relative peak low frequency output.
Alright, so I figured out that everything I've simulated is wrong! "Isobaric" was selected, which changes everything. LOL! I was hoping you guys looking on might notice if I was doing something dumb, but I guess I snuck that one past y'all!
An updated design is forthcoming.
I went back through all the drivers (mainly Dayton Audio), and I came up with a completely different one this time. The only one that appears to go down to 20hz in a workable size cabinet is the 12" Dayton UM12-22.
It looks like the total volume is going to be ~8ft^3 for the entire cabinet.
An updated design is forthcoming.
I went back through all the drivers (mainly Dayton Audio), and I came up with a completely different one this time. The only one that appears to go down to 20hz in a workable size cabinet is the 12" Dayton UM12-22.
It looks like the total volume is going to be ~8ft^3 for the entire cabinet.
How could we? You didn't attach your sim(s) data for us to install whatever program used or even posted any screen shots.
Also, with only one driver, surprised one can even choose isobaric without it returning an error code/whatever.
Also, with only one driver, surprised one can even choose isobaric without it returning an error code/whatever.
Now that I am aware of my simulation error, my new simulations show that a 15" woofer + 12" subwoofer was going to end up closer to 8ft^3, which makes it similar sized to the Genelec. But I think that is too big.
So I started playing with 10" subwoofers. I originally wanted to go bigger than 10's, because my current subs are 10". But they are cheap 10" subs, so I expect I can improve on them quite a bit while using the same size driver.
The 10" subwoofer got me back down to ~100L/ 3.5ft^3 for the subwoofer portion, while still reaching 20hz. F3 is 19.4hz as modeled currently.
To further reduce the size of the cabinet, I started looking at 12" woofers. But a 12" woofer has trouble reaching down far enough in a reasonably sized sealed cabinet. That lead me to try a 12" woofer with matching 12" passive radiator in a 45L volume. That got me back to where I was with the single 15" woofer, with an F3 of ~45hz. In comparison, the 15" woofer needed about 100L to get a similar response.
The 10" subwoofer is the Dayton RSS265HF-4. The 12" is Dayton DSA315-8. And the passive radiator is a 12" DSA315PR.
And I think I can get away with the smaller PWR-ICE125 plate amps--they produce 2x125watts into 4 ohms, as opposed to 2x250 watts.
I realize the 12" driver is 8ohm while the subwoofer is 4 ohm. But the simulated output of the 12" @ 40 watts is about the same as the 10" subwoofer @125 watts (105db). So I think I can make that work.
What do you think? Do you think it a mistake to go away from the single 15" for the 12" + PR? I like the idea; who doesn't want an array of big drivers staring them down? I think this is a more practical form factor. It isn't small, but it's about half the size of the original idea.
So I started playing with 10" subwoofers. I originally wanted to go bigger than 10's, because my current subs are 10". But they are cheap 10" subs, so I expect I can improve on them quite a bit while using the same size driver.
The 10" subwoofer got me back down to ~100L/ 3.5ft^3 for the subwoofer portion, while still reaching 20hz. F3 is 19.4hz as modeled currently.
To further reduce the size of the cabinet, I started looking at 12" woofers. But a 12" woofer has trouble reaching down far enough in a reasonably sized sealed cabinet. That lead me to try a 12" woofer with matching 12" passive radiator in a 45L volume. That got me back to where I was with the single 15" woofer, with an F3 of ~45hz. In comparison, the 15" woofer needed about 100L to get a similar response.
The 10" subwoofer is the Dayton RSS265HF-4. The 12" is Dayton DSA315-8. And the passive radiator is a 12" DSA315PR.
And I think I can get away with the smaller PWR-ICE125 plate amps--they produce 2x125watts into 4 ohms, as opposed to 2x250 watts.
I realize the 12" driver is 8ohm while the subwoofer is 4 ohm. But the simulated output of the 12" @ 40 watts is about the same as the 10" subwoofer @125 watts (105db). So I think I can make that work.
What do you think? Do you think it a mistake to go away from the single 15" for the 12" + PR? I like the idea; who doesn't want an array of big drivers staring them down? I think this is a more practical form factor. It isn't small, but it's about half the size of the original idea.
Alright, so I have some progress to report. I have a prototype built, drivers installed, and I played with the DSP a little. Subjectively it is performing much better than the old subwoofers. But that is a low bar!
I have some preliminary results to share, and I am interested to get some other people's perspective. I haven't done much subwoofer measurement, so any input is appreciated.
For configuration, I am putting a delay on the front 12" woofer and on the main speaker, so the rear fires first, and the sound waves wrap around and should be in phase with the front 12" woofer. It reinforces the radiation from the front driver, and also cancels some of the front radiation that would wrap around behind.
(The other option is to delay the rear driver and reverse the phase. I tried this first, but I had nothing but issues. And then I watched a Dave Rat video where he explained that he doesn't use that configuration because of "sonic issues". So I abandoned it.)
I am playing both front and rear drivers up to 200hz, and the main speaker is turned off.
I spent a couple days naively playing around and getting results like this:
This does have a touch of DSP, but only a few filters. It looks very good to me, but the caveat is I am measuring from a low height, the average of the 2 drivers.
I was going to claim victory, but then I took more measurements and I started getting some "different" results. This next screen shot is a series of measurements. The blue lines are 3 actual seating positions. The red lines are ~3' in front, and the green lines are ~6' in front. My nice results I had been looking at completely disappeared:
But the funny thing is that if I place the microphone low, the results flatten out again. Here is a measurement I took immediately after the ones above, but 1' off the floor (no EQ applied):
So it seems that it is working great, if you want to lay on the floor!
So the big question, is directivity the answer? It does not appear to work in a small, untreated room. But does directivity work in a larger, well treated room, in a mid-field configuration, where the listener is well away from boundaries? That may work....but I don't have a room remotely like that!
If directivity isn't the answer, then the next step is to build a second prototype and try multiple subwoofer EQ to see if that looks more promising.
If multiple subwoofers is the answer, then I have to figure out if 4 subwoofers combined into 2 cabinets will work? That's what I originally thought they were doing, but that is just a guess!
I have some preliminary results to share, and I am interested to get some other people's perspective. I haven't done much subwoofer measurement, so any input is appreciated.
For configuration, I am putting a delay on the front 12" woofer and on the main speaker, so the rear fires first, and the sound waves wrap around and should be in phase with the front 12" woofer. It reinforces the radiation from the front driver, and also cancels some of the front radiation that would wrap around behind.
(The other option is to delay the rear driver and reverse the phase. I tried this first, but I had nothing but issues. And then I watched a Dave Rat video where he explained that he doesn't use that configuration because of "sonic issues". So I abandoned it.)
I am playing both front and rear drivers up to 200hz, and the main speaker is turned off.
I spent a couple days naively playing around and getting results like this:
This does have a touch of DSP, but only a few filters. It looks very good to me, but the caveat is I am measuring from a low height, the average of the 2 drivers.
I was going to claim victory, but then I took more measurements and I started getting some "different" results. This next screen shot is a series of measurements. The blue lines are 3 actual seating positions. The red lines are ~3' in front, and the green lines are ~6' in front. My nice results I had been looking at completely disappeared:
But the funny thing is that if I place the microphone low, the results flatten out again. Here is a measurement I took immediately after the ones above, but 1' off the floor (no EQ applied):
So it seems that it is working great, if you want to lay on the floor!
So the big question, is directivity the answer? It does not appear to work in a small, untreated room. But does directivity work in a larger, well treated room, in a mid-field configuration, where the listener is well away from boundaries? That may work....but I don't have a room remotely like that!
If directivity isn't the answer, then the next step is to build a second prototype and try multiple subwoofer EQ to see if that looks more promising.
If multiple subwoofers is the answer, then I have to figure out if 4 subwoofers combined into 2 cabinets will work? That's what I originally thought they were doing, but that is just a guess!
Last edited:
While testing I tried laying the cabinet on it's side, and the results at the seating positions looked better in the seating positions.
It seems the drivers of the W371 being offset create an upward directivity pattern? I haven't taken my vertical cabinet outside to do thorough testing, but holding the mic near the ceiling in line with the 2 drivers gave quite flat measurements, same as holding the mic ~3ft in front. So quite possibly the W371As create a "bass bubble" at the mixing desk, not necessarily for the entire room.
So I built another box, making it more square so the driver centers are roughly centered from front to back. Keep in mind this is function over form right now!
Measurements from both are better. The deep peaks are gone. It is quite good.
As a caveat, my room is so bad that I previously believed it was impossible to get anything close to good bass. I knew my old subs are sub-par, but I figured the house was more of a limiting factor than the equipment. I think that's still partly true still. I'd like to see subs like this tested in a better room.
Some graphs. The 3 seating positions x 2 sub cabinets:
Middle seating position x 2 sub cabinets:
What does it sound like? I can hear every bass note. I hear the tone/texture of the bass guitar, the thump of the kick drum. It's really the best I've ever experienced. I'm definitely hearing bass notes and nuance I've never heard before. I watched Men In Black 3, and it sounded better than a movie theater. Granted it doesn't make any difference on music that doesn't have a lot of low frequency content, like 80's rock. But I'm kind of a bass head, and I'm listening to music that has a prominant bass line.
The cabinet is ugly right now, but it'd be fine with some chamfers, in laquor black wood grain finish, with a grill. But I'm still thinking about the design, the driver selection, and the shape and proportions. The 12" driver + 12" PR is playing flat down to ~30hz, which works great. But I'm noticing the baffle would allow a single, much larger driver. That's tempting!
It seems the drivers of the W371 being offset create an upward directivity pattern? I haven't taken my vertical cabinet outside to do thorough testing, but holding the mic near the ceiling in line with the 2 drivers gave quite flat measurements, same as holding the mic ~3ft in front. So quite possibly the W371As create a "bass bubble" at the mixing desk, not necessarily for the entire room.
So I built another box, making it more square so the driver centers are roughly centered from front to back. Keep in mind this is function over form right now!
Measurements from both are better. The deep peaks are gone. It is quite good.
As a caveat, my room is so bad that I previously believed it was impossible to get anything close to good bass. I knew my old subs are sub-par, but I figured the house was more of a limiting factor than the equipment. I think that's still partly true still. I'd like to see subs like this tested in a better room.
Some graphs. The 3 seating positions x 2 sub cabinets:
Middle seating position x 2 sub cabinets:
What does it sound like? I can hear every bass note. I hear the tone/texture of the bass guitar, the thump of the kick drum. It's really the best I've ever experienced. I'm definitely hearing bass notes and nuance I've never heard before. I watched Men In Black 3, and it sounded better than a movie theater. Granted it doesn't make any difference on music that doesn't have a lot of low frequency content, like 80's rock. But I'm kind of a bass head, and I'm listening to music that has a prominant bass line.
The cabinet is ugly right now, but it'd be fine with some chamfers, in laquor black wood grain finish, with a grill. But I'm still thinking about the design, the driver selection, and the shape and proportions. The 12" driver + 12" PR is playing flat down to ~30hz, which works great. But I'm noticing the baffle would allow a single, much larger driver. That's tempting!
I found the answer to a question I was asking before. Should I sum the low frequencies and play them through both cabinets?
Now I believe the answer is a definitive "No".
The answer came from a Dave Rat video. He explained that comb filtering is only caused when 2 identical signals are played from multiple sources at the same time. He says 2 speakers side by side do not interfere with each other until they play the same signal.
His approach for live sound is to use 2 different microphones for everything, because each microphone outputs a slightly different signal. This gives him 2 different versions of the same sound. And when he plays one version on the left and the other on the right, they two signals are different enough that they do not create comb filtering in the middle. Genius!
Anyway, so I realized that a summed subwoofer signal played through multiple subwoofers will create comb filtering. But with a stereo subwoofer arrangement, the left and a right subwoofer should be playing slightly different signals, reducing or eliminating comb filtering. But I'm not sure how much difference there is in low frequency content between the left and right channels in a typical stereo mix.
Now I believe the answer is a definitive "No".
The answer came from a Dave Rat video. He explained that comb filtering is only caused when 2 identical signals are played from multiple sources at the same time. He says 2 speakers side by side do not interfere with each other until they play the same signal.
His approach for live sound is to use 2 different microphones for everything, because each microphone outputs a slightly different signal. This gives him 2 different versions of the same sound. And when he plays one version on the left and the other on the right, they two signals are different enough that they do not create comb filtering in the middle. Genius!
Anyway, so I realized that a summed subwoofer signal played through multiple subwoofers will create comb filtering. But with a stereo subwoofer arrangement, the left and a right subwoofer should be playing slightly different signals, reducing or eliminating comb filtering. But I'm not sure how much difference there is in low frequency content between the left and right channels in a typical stereo mix.
I don’t think your comb filtering theory typically applies to sub frequencies. Subwoofers often play the same signal and are often stacked side by side etc. try them summed to mono and see what you think.
Maybe it isn't important, but it is a good excuse for me to not worry about it!
Don't pro subwoofer arrays use delays? They do for cardiod setups, but I think those big arrays only work with the right delays.
I feel like I have some experience quite recently. When I delayed the rear driver on my cabinet to get a cardiod pattern, I got significant peaks and nulls. This is from the energy from the rear driver reflecting and creating interferrence. measurements that i could not fix. That is essentially a comb filter, right? Different outputs from different places and different times, causing summing and cancellation?
Yesterday I spent a bunch of time trying to use a summed signal for both the rear sub drivers. And I tried to add a 3rd sub. The idea was to use the rear drivers as though they were separate subwoofers.
When I played both rear drivers at the same time, the output was not pretty! But the response was pretty good individually. And the response was pretty good when I combined the front and rear drivers on either the left or right, but when I combined left and right in any combination, the measurements because quite bad. I realized it was the interaction of the same sound from 2 different sources.
I did some reading after this, and it seems I missed the requirement that multiple subwoofer arrangements for home theater need a symmetrical, rectangular room. And then the subs need to be placed symmetrically in the room, apply DSP, etc.
My room is not really rectangular, not symmetrical, and I can't place my speakers symmetrically in the room. So it seems this technique I'm experimenting with is a good solution.
Fredy,
You are equating a live mix with individual sources sent to multiple locations in a large scale application to a small room where you have no control over the source material, two very different scenarios requiring different approaches.
In your small room, the low frequency peaks and dips are from room mode problems rather than direct path length difference. A cardioid pattern won't fix the room mode problems.
Art
I think I disagree with everything you said!
I am experimenting with a design from a top tier manufacturer! That is, Genelec.
If you understand what they're doing, and if you know it is something other than the options I have voiced, please enlighten us!
(The claim of "fix the room mode problems." was your addition. I did not claim I was fixing them.)
Can't enlighten you as to exactly what Genelec is doing with the W371A.
Their description is not of a cardioid subwoofer response, it is basically omni-directional below the crossover region.
The "Continued Directivity Mode" and "Null Steering Mode" appear to work primarily on directivity in the wide range of overlap in the default crossover range of 150-250 Hz between the W371A and a monitor that typically sits on top of it.
Have you used the W371A?
Art
Hey- not trying to be a dich here, just a heads up that Art is an OG and knows a thing or two about which he speaks. Ask questions sure but he deserves some respect.
It seems to me that Genelec are being clever, and I'm just curious about what they are doing. It seems this is a case of there is more going on than meets the eye, so I think it is a mistake to fall back on rules of thumb.
I watched an interview with a Genelec engineer who talked about using the overlap in frequency range between the different drivers to get a flatter response. That seemed pretty straight forward and do-able with a normal DSP unit. It seems too simple, really.
The question about directivity was brought to my attention by some other forum members, and it led me to dive into the world of pro audio. I know it is not exactly the same, but sound waves are sound waves. I found Dave Rat on youtube, and I feel I am better off for it.
The cabinet definitely resembles a pro cardiod subwoofer. I'm sure they use a delay, because otherwise they would have phase issues. Adding the delay makes it a cardiod sub! I agree that it probably does not make a clear cardiod pattern in a room, but it would do so in free space.
Genlec could have designed the cabinet so both drivers were front facing, which would make the delay unnecessary and make it NOT cardiod. But that isn't what they did.
I don't know how they could control horizontal directivity with vertically aligned drivers. I'm looking for new explanations, but the measurements you attached do indicate a horizontal directivity characteristic of a cardiod subwoofer. How would they achieve that otherwise?
The common rule of thumb says this is pointless in a small room, but Genelec went ahead and did it anyway. I expect is is more about active cancellation than cardiod pattern?
I'm not trying to saying someone is right and someone else is wrong. I'm still open to suggestions. But at this point it is difficult for me to rule out the cardiod sub/ low frequency directivity idea.
My background is also in pro audio sound reinforcement, Dave and I started in the same era building our own systems. He provides loads of good information in his videos and interviews.
Sound waves are sound waves, and frequencies are frequencies, and because of long wavelengths compared to room dimensions, low frequencies behave differently in small rooms than large rooms or free space.
Having a cardioid (shaped heart-like) pattern makes a subwoofer cardioid.
There is no mention of a cardioid pattern in the W371A user manual, and no free space measurements- in fact the manual states "This product is designed for indoor use only."
The Genelec W371A in combination with a top cabinet and the GLM (Genelec Loudspeaker Management) provides 3 separate functional locations to provide "Null Steering" or "Continuous Directivity Matching" within the crossover overlap region. The three different locations, and their three distances to boundaries are key elements to the GLM.
Again, the directivity characteristic of the Genelec W371A as provided in the manual is not that of a cardiod subwoofer, the patterns show "Null steering" or "Continued Directivity" in the crossover overlap region down to around 60 Hz, the "sub" region from 25-60Hz is an omnidirectional pattern not affected by processing.
Yes, I agree "active cancellation" and "null steering" are somewhat synonymous, though not the same as a cardioid pattern.
Art
Last edited: