Hi all,
I was posting a little about this in my thread in PA systems, but thought it might deserve its own write-up.
It started off with a pair of 6.5" PA speakers, and I thought it'd be good to have some low-profile subs to go with them - something to act as a base for a speaker pole to go between. It doesn't need to be anything huge, since the main speakers are small and fairly light.
Ideally it'd go as low as possible while still retaining good output levels for smaller spaces, or venues where keeping the PA system near-invisible is important.
After looking around online at a variety of 6-8" drivers, I concluded that it would be a difficult line to follow - most 8" PA drivers have plenty of thermal power handling, but very little Xmax, while home theatre/car subs had plenty of Xmax, but low sensitivity and relatively little power handling. Something in-between was needed. There were a few drivers from the likes of Faital Pro that would do the job, often with a half-roll rubber surround, but they were usually expensive if you wanted useful Xmax, power handling, and T/S parameters that lend themselves to smallish subwoofers.
After even more looking around, I found some cabinets for sale that had drivers in that looked very similar to the B&C 6PS38, so I bought a couple and tested them. They weren't the 6PS38, but a custom version built for a speaker manufacturer.
The versions I have are 16ohm, and some of the T/S parameters are a little different. With a bit of tweaking, though, I expect the cabinet here would work with a variety of drivers.
Initially, I had wanted something like a tapped horn, but folding into a nice form-factor got very difficult. So, I set off with a ported box, like this one:
Which simulated pretty well - around 115dB from 40Hz upwards, with 30v into all three drivers in parallel. I liked the square form-factor etc. However, I wasn't so keen on having the port on one side away from the drivers. Output on two faces makes positioning more difficult when they might be used in any venue you can imagine.
The current iteration looks like this:
This time, I've slot-loaded the drivers. This improves matters in a couple of ways:
1 - actually a little more efficiency around 60Hz
2 - all the output from the cabinet is from one face
3 - centre of gravity is now more central - should be more stable for the main speakers.
4 - drivers are hidden from view - attaching grilles etc is much easier
Downside is I've now got to include an access panel, probably the back panel of the cabinet. That's not a big deal, though - Doubled up plywood and plenty of T-nuts will do it just fine.
Here's what Hornresp thinks they'll do:
That's 30v input, and touching Xvar (B&Cs version of Xmax) at 65Hz. Not a bad little box. Some EQ will be needed to level out the response, since there's 8dB of difference from the top to the bottom of the useful range. That's all easily done, though, since there'll be some form of digital processing available.
I'm hoping the rest of the crossover components will arrive in the next couple of days so I can finish off the main speakers and get to work on these subs. It's going to be a cute little PA system that will (hopefully) put some larger cabinets to shame. 15" 2-way plastic boxes, I'm looking at you.
Cheers,
Chris
I was posting a little about this in my thread in PA systems, but thought it might deserve its own write-up.
It started off with a pair of 6.5" PA speakers, and I thought it'd be good to have some low-profile subs to go with them - something to act as a base for a speaker pole to go between. It doesn't need to be anything huge, since the main speakers are small and fairly light.
Ideally it'd go as low as possible while still retaining good output levels for smaller spaces, or venues where keeping the PA system near-invisible is important.
After looking around online at a variety of 6-8" drivers, I concluded that it would be a difficult line to follow - most 8" PA drivers have plenty of thermal power handling, but very little Xmax, while home theatre/car subs had plenty of Xmax, but low sensitivity and relatively little power handling. Something in-between was needed. There were a few drivers from the likes of Faital Pro that would do the job, often with a half-roll rubber surround, but they were usually expensive if you wanted useful Xmax, power handling, and T/S parameters that lend themselves to smallish subwoofers.
After even more looking around, I found some cabinets for sale that had drivers in that looked very similar to the B&C 6PS38, so I bought a couple and tested them. They weren't the 6PS38, but a custom version built for a speaker manufacturer.
The versions I have are 16ohm, and some of the T/S parameters are a little different. With a bit of tweaking, though, I expect the cabinet here would work with a variety of drivers.
Initially, I had wanted something like a tapped horn, but folding into a nice form-factor got very difficult. So, I set off with a ported box, like this one:
Which simulated pretty well - around 115dB from 40Hz upwards, with 30v into all three drivers in parallel. I liked the square form-factor etc. However, I wasn't so keen on having the port on one side away from the drivers. Output on two faces makes positioning more difficult when they might be used in any venue you can imagine.
The current iteration looks like this:
This time, I've slot-loaded the drivers. This improves matters in a couple of ways:
1 - actually a little more efficiency around 60Hz
2 - all the output from the cabinet is from one face
3 - centre of gravity is now more central - should be more stable for the main speakers.
4 - drivers are hidden from view - attaching grilles etc is much easier
Downside is I've now got to include an access panel, probably the back panel of the cabinet. That's not a big deal, though - Doubled up plywood and plenty of T-nuts will do it just fine.
Here's what Hornresp thinks they'll do:
That's 30v input, and touching Xvar (B&Cs version of Xmax) at 65Hz. Not a bad little box. Some EQ will be needed to level out the response, since there's 8dB of difference from the top to the bottom of the useful range. That's all easily done, though, since there'll be some form of digital processing available.
I'm hoping the rest of the crossover components will arrive in the next couple of days so I can finish off the main speakers and get to work on these subs. It's going to be a cute little PA system that will (hopefully) put some larger cabinets to shame. 15" 2-way plastic boxes, I'm looking at you.
Cheers,
Chris
Only 6db boost "down there" will at least quadruple the required amp power.
Not to mention the excursion of the drivers.
Power handling and xmax...
...and then max spl...
PA 6.5" are not really subwoofer drivers.
So you "touch" xmax and still want to boost using eq?
I don`t really know if that's a good approach.
Not to mention the excursion of the drivers.
Power handling and xmax...
...and then max spl...
PA 6.5" are not really subwoofer drivers.
So you "touch" xmax and still want to boost using eq?
I don`t really know if that's a good approach.
I'm not sure you understand.
The simulated graph was to show roughly how loud the sub will go before sounding stressed. You can EQ it any way you like, but the conclusion is that it'll do 115dB at any frequency in it's operating band. Some frequencies, it will go louder than that.
It's how far I push the volume that determines how much power gets sent to the subwoofer - the EQ can be anything at all.
If I cut 150Hz by 8dB, will that make you happy?
Chris
The simulated graph was to show roughly how loud the sub will go before sounding stressed. You can EQ it any way you like, but the conclusion is that it'll do 115dB at any frequency in it's operating band. Some frequencies, it will go louder than that.
It's how far I push the volume that determines how much power gets sent to the subwoofer - the EQ can be anything at all.
If I cut 150Hz by 8dB, will that make you happy?
Chris
You were right about PA 6.5" drivers not being subwoofer drivers - they're definitely not.
If you give them lots of friends, though, between them they can put out some bass. Remember here there's around 400cm^2 of cone area, which is more than a 10" but less than a 12". Add in a good Xmax and reasonable power handling, and you can move some air.
I would've liked to have used the Tang Band 6.5" mini-subwoofer, but I found the response was very peaky when tuned around 40Hz, and the 100w power handling means you cannot use all of their excursion without risking thermal failure.
Chris
If you give them lots of friends, though, between them they can put out some bass. Remember here there's around 400cm^2 of cone area, which is more than a 10" but less than a 12". Add in a good Xmax and reasonable power handling, and you can move some air.
I would've liked to have used the Tang Band 6.5" mini-subwoofer, but I found the response was very peaky when tuned around 40Hz, and the 100w power handling means you cannot use all of their excursion without risking thermal failure.
Chris
Some thoughts:
1. At 16 ohms per driver, why three drivers, instead of say four?
2. How much power will each driver be seeing @30V?
3. What are the t/s parameters of the drivers? The B&C driver has an Fs of 75 Hz, suggesting that going much below that is going to result in high GD.
4. Why the choice of 40 Hz cutoff? You might get a bit more out of these drivers by shooting for 48 Hz.
1. At 16 ohms per driver, why three drivers, instead of say four?
2. How much power will each driver be seeing @30V?
3. What are the t/s parameters of the drivers? The B&C driver has an Fs of 75 Hz, suggesting that going much below that is going to result in high GD.
4. Why the choice of 40 Hz cutoff? You might get a bit more out of these drivers by shooting for 48 Hz.
You could give that a try.
Car drivers are inefficient because they go low in small boxes ( Iron Law). Use serval of them in a high order or 1/4 wave resonator type enclosure to have an efficient enclosure like two alpine 8's in a TPCH or Paraflex.
They can move as much air as a 12 with decent power handling while still working in a very compact package. This is a rough sim with the alpines wired with coils in series and the two drivers in parallel (2ohm load), 300W per driver at impedance minima ( If music has good crest factor it is not a problem ). Can be eq'd flat and you get approx 122db spl 40-200Hz at 1m, decent output for a 120L box with two 8's. I think it may work well. I will try to post better sims (More sensitivity and less sag in fr).
Car drivers are inefficient because they go low in small boxes ( Iron Law). Use serval of them in a high order or 1/4 wave resonator type enclosure to have an efficient enclosure like two alpine 8's in a TPCH or Paraflex.
They can move as much air as a 12 with decent power handling while still working in a very compact package. This is a rough sim with the alpines wired with coils in series and the two drivers in parallel (2ohm load), 300W per driver at impedance minima ( If music has good crest factor it is not a problem ). Can be eq'd flat and you get approx 122db spl 40-200Hz at 1m, decent output for a 120L box with two 8's. I think it may work well. I will try to post better sims (More sensitivity and less sag in fr).
Attachments
Hey Brian, thanks for replying.
To answer your questions:
1 - Good question. I only have 6 drivers in total, and can't get hold of any more. If I blow any of them, it's game over. That's okay, though, because I got these drivers for a great price. Three drivers per cabinet also fits quite nicely into the sort of form-factor I'm looking for, but that's a happy coincidence - I was looking at using just two drivers per sub, until the extra became available to me.
2 - With their ~10ohm DCR, it comes in around 100w where impedance is low (port tuning and >100Hz.
3 - Hornresp numbers are:
Sd: 132
Cms: 2.65e-04
Mmd: 13.77
Re: 9.24
BL: 11.17
Rms: 1.60
Le: 0.26
For the most recent simulation, I've got:
Offset Driver, 3 in parallel
S1=S2=S3=200
L12=L23=26
Vrc=36
Lrc=50
Ap=200
Lpt=62
4 - I'm happy with the output level obtainable here. The tuning frequency actually comes out at 45Hz, but the slightly-high Qts means there's output at 40Hz as the cones start to unload. I'd set a highpass filter there. If you go to a 60Hz tuning (set Lpt=37), the frequency response levels out and you can go a bit louder, but the subs already outrun the tops by a good margin - I don't think the extra output would be used, while the LF extension is always nice to have.
I will have a think about this, though - would I rather have some little boxes that go low, or ridiculously loud?
The plan is a 4-channel amplifier:
Ch1 & 2 for the main PA speakers
Ch3 for subs
Ch4 for monitors/spare.
The amplifier (NU4-6000) is fine driving low impedances, and will put out up to 50v RMS. Nice little setup for live sound and discrete setups.
Chris
To answer your questions:
1 - Good question. I only have 6 drivers in total, and can't get hold of any more. If I blow any of them, it's game over. That's okay, though, because I got these drivers for a great price. Three drivers per cabinet also fits quite nicely into the sort of form-factor I'm looking for, but that's a happy coincidence - I was looking at using just two drivers per sub, until the extra became available to me.
2 - With their ~10ohm DCR, it comes in around 100w where impedance is low (port tuning and >100Hz.
3 - Hornresp numbers are:
Sd: 132
Cms: 2.65e-04
Mmd: 13.77
Re: 9.24
BL: 11.17
Rms: 1.60
Le: 0.26
For the most recent simulation, I've got:
Offset Driver, 3 in parallel
S1=S2=S3=200
L12=L23=26
Vrc=36
Lrc=50
Ap=200
Lpt=62
4 - I'm happy with the output level obtainable here. The tuning frequency actually comes out at 45Hz, but the slightly-high Qts means there's output at 40Hz as the cones start to unload. I'd set a highpass filter there. If you go to a 60Hz tuning (set Lpt=37), the frequency response levels out and you can go a bit louder, but the subs already outrun the tops by a good margin - I don't think the extra output would be used, while the LF extension is always nice to have.
I will have a think about this, though - would I rather have some little boxes that go low, or ridiculously loud?
The plan is a 4-channel amplifier:
Ch1 & 2 for the main PA speakers
Ch3 for subs
Ch4 for monitors/spare.
The amplifier (NU4-6000) is fine driving low impedances, and will put out up to 50v RMS. Nice little setup for live sound and discrete setups.
Chris
Chris,
Looks great!
Only suggestion would be to mount the drivers with the magnets in the plenum (slot) for heat reduction, which could also make mounting easier with the access panel on the plenum side. Reversed drivers would also increase interior volume slightly. Only down side is the center of gravity would be more off center, but you could mount weights in the port to compensate ;^).
Cheers,
Art
Last edited:
Well, these drivers are actually woofers, not subwoofers. Neither B&C nor Faital got any 6" drivers that are capable of the subwoofer range. And you should measure the TSP beforehand and then simulate it. You'll realize they aren't usable for a tuning around 40Hz.
The port works from ~40-60Hz. At that wavelength and that radiating surface the sound is radiated in a spherical pattern which means it does not matter on which side the port is as long as it's not blocked off.
Hornresp isn't very relyable in terms of bassreflex. Use WinISD or BassCADe or similar programs. I doubt you'll get a usable, half-way clean tuning or that spl.
A 8dB difference is HUGE!
It's not the question if it can be done (that's easy), it's the question if that is actually usable (hardly ever, honestly). I'm not talking about the settings, I talk about compression, thd and power-compression.Uhm, please, please simulate them first. (not in hornresp)
I'm afraid you'd be disappointed.
Last edited:
I've calculated it, at 3x 16 Ohm in parallel, you'd get 5,33 Ohm. At 30V that's less than 170W peak for 115dB (168,85 to be precise) or 84W/112dB program (42W rms). A single ready-bought commercial 12" sub for 250€ will make ~10dB more! To be honest, 112dB program will be a huge disappointment, not even close to be enough for any venue.
Experience.
It works as long as you are not using too extreme boxes and as long as you have precise parameters. Both isn't given for this project, a 16 Ohm driver with just 0,26 mH? Similar drivers with 16 Ohm are closer to 0,7-0,8 mH. So yes, hornresp can do that but not with the wrong parameters. Other simulation software solutions calculate it differently and get better, much closer results in such cases.
While that's ofcourse true, BassCADe & co are still closer and hornresp might create the illusion of having something that would work.
Hmm.. if a simulation generated by one of those programs is "closer" with the use of incorrect values entered for the driver's parameters, I probably won't be using it for long
Hornresp has been accurate enough for my needs. Pictured is the sim'd vs. measured response for my POC6 build, for example. The measured response has some noise in it due to the conditions under which the measurement was performed, but the results are close enough for my needs.
I've only come across one example where measured results don't match the sim. However in that case the sim was only an approximation of the build.
Attachments
Wouldn't "program" be higher than linear output? The "115dB" is based on the point that the drivers will exceed rated Xmax within the passband, but that's with a continuous signal at the frequency of maximum excursion. For music playback, I'll bet it can go a bit louder.
It's pretty interesting that Chris seems have come up with a design that gets 115dB (theoretically) down to 40 Hz. My POC3 build required a 12" driver (5mm Xmax though) in a box that's just over 60% larger to achieve that, though it can achieve it with just 22V. I've used that build for the occasional office get-together and beach gathering. It's not going to break down any doors, but it does fill out the bottom end quite nicely. Not everyone's trying to get SPL levels that require people to scream at each other to be heard
.
Last edited:
A down firing 12 or two 10's would perform far better.
A lab 12 in a wide and flat enclosure (which will actually be heavy enough to add meaningful support) down firing would absolutely destroy these boxes.
115db maximum is rather low for anything but background music in my opinion and experience. You always want MORE bass than top end for live music playback, so 115db maximum output at LF would mean 105 or so for mid/hf at 1m. That's barely 95-96db PEAK at 10m for bass, 85 for midband. Nowhere near enough for anything but a solo guitar in a quiet restaurant.
10m from the speakers with background noise(people talking/yelling etc) would almost drown this out.
115db midband and 125db bass down to 40hz would be the MINIMUM for anything with 50-100 guests.
Bose tried a similar idea with the MB4 'sub' which has widely been shunned by the PA community as you need 4 or more to replace a 'real' PA sub.
They have since also released the double 10 bass module(B2) for their L1 systems due to wide criticism of their weak B1 enclosure(which uses 2 small drivers but 2 enclosures are usually used)
A lab 12 in a wide and flat enclosure (which will actually be heavy enough to add meaningful support) down firing would absolutely destroy these boxes.
115db maximum is rather low for anything but background music in my opinion and experience. You always want MORE bass than top end for live music playback, so 115db maximum output at LF would mean 105 or so for mid/hf at 1m. That's barely 95-96db PEAK at 10m for bass, 85 for midband. Nowhere near enough for anything but a solo guitar in a quiet restaurant.
10m from the speakers with background noise(people talking/yelling etc) would almost drown this out.
115db midband and 125db bass down to 40hz would be the MINIMUM for anything with 50-100 guests.
Bose tried a similar idea with the MB4 'sub' which has widely been shunned by the PA community as you need 4 or more to replace a 'real' PA sub.
They have since also released the double 10 bass module(B2) for their L1 systems due to wide criticism of their weak B1 enclosure(which uses 2 small drivers but 2 enclosures are usually used)
Last edited:
Your simulation looks very close. The differences could even be just measurement tolerances. Well, if I simulate the 'close' B&C 6PS38, however, I come to this:
Do you see that bump at the end? That will get you, that sounds really bad and it becomes much worse with the power compression. And the excursion and the hole in the response/spl above it is no laughing matter either.
I've simulated a whole bunch of variants and none looked even remotely like the hornresp result, none are clean or usable for more than moderate party level. And none of the other PA drivers of that size will get you much better results, they aren't designed to play that deep, they are intended for tops/sat and small enclosures.
You'll get better results with hifi or even car audio drivers and while I'm not a fan of the latter, they will actually go deep enough and need a much smaller enclosure, 70x53,6x20 isn't exactly 'invisible'. I'm ofcourse not any kind of reference but that size and that low max-spl are exactly the opposite of the listed design targets. Just to be clear: I would not use hifi- or car-speakers either, I'd use actual PA subwoofers.
Do you see that bump at the end? That will get you, that sounds really bad and it becomes much worse with the power compression. And the excursion and the hole in the response/spl above it is no laughing matter either.
I've simulated a whole bunch of variants and none looked even remotely like the hornresp result, none are clean or usable for more than moderate party level. And none of the other PA drivers of that size will get you much better results, they aren't designed to play that deep, they are intended for tops/sat and small enclosures.
You'll get better results with hifi or even car audio drivers and while I'm not a fan of the latter, they will actually go deep enough and need a much smaller enclosure, 70x53,6x20 isn't exactly 'invisible'. I'm ofcourse not any kind of reference but that size and that low max-spl are exactly the opposite of the listed design targets. Just to be clear: I would not use hifi- or car-speakers either, I'd use actual PA subwoofers.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.