GM,
I sincerely apologize for the wording. Nobody is "supposed" to do anything here to help others, and I greatly appreciate any and all help, guidance, and opinions from everyone, especially those who are highly experienced. For everyone's benefit, not just mine, please do take the time to post your insights regarding this project.
At the time I was happy that I caught the mistake and sad at the same time because I needed to redo the design. I wanted to post something quickly, so I wouldn't be one to spread incorrect info online. BTW if anyone has contact info for the designer of the Big Fun Horn, please thank him for posting all that great info about the design and that he should correct the error in horn length vs the upper frequency cutoff.
I sincerely apologize for the wording. Nobody is "supposed" to do anything here to help others, and I greatly appreciate any and all help, guidance, and opinions from everyone, especially those who are highly experienced. For everyone's benefit, not just mine, please do take the time to post your insights regarding this project.
At the time I was happy that I caught the mistake and sad at the same time because I needed to redo the design. I wanted to post something quickly, so I wouldn't be one to spread incorrect info online. BTW if anyone has contact info for the designer of the Big Fun Horn, please thank him for posting all that great info about the design and that he should correct the error in horn length vs the upper frequency cutoff.
Re: Re: Back to the drawing board.
GM,
When I had read that comment by JohninCR it came across clearly to me that John was making it in playful jest.
Taking the time as you do to make calculations and figure things out to help other members here is I am sure appreciated very much by all and is an example of the core value of these forums in how they will inspire participation such as yous. Please do not be upset.
GM said:
GM,
When I had read that comment by JohninCR it came across clearly to me that John was making it in playful jest.
Taking the time as you do to make calculations and figure things out to help other members here is I am sure appreciated very much by all and is an example of the core value of these forums in how they will inspire participation such as yous. Please do not be upset.
Paul is going to be kind enough to post my 2 newer versions.
The 120hz upper cutoff version is substantially smaller than the original. If you painted it the same color as the wall, you almost wouldn't notice it in the corner. The 240hz version is bigger than the original but has a tiny chamber, only 7.5L for 16 drivers.
120hz seems like the right cutoff for NSB's in a small sealed box but 240hz seems to match the open air rolloff. So which is better?
Another concern I have is that both chambers are quite small. What kind of affect will that have on the sound coming from the front of the drivers? HornRESP only tells me what the horn is supposed to do. I have no clue how the front radiated sound is affected.
The 120hz upper cutoff version is substantially smaller than the original. If you painted it the same color as the wall, you almost wouldn't notice it in the corner. The 240hz version is bigger than the original but has a tiny chamber, only 7.5L for 16 drivers.
120hz seems like the right cutoff for NSB's in a small sealed box but 240hz seems to match the open air rolloff. So which is better?
Another concern I have is that both chambers are quite small. What kind of affect will that have on the sound coming from the front of the drivers? HornRESP only tells me what the horn is supposed to do. I have no clue how the front radiated sound is affected.
Mike,
Your point is why I haven't built 1 yet. I'm still hoping that someone with rear horn design experience steps in to offer some advice.
While the much shorter horn isn't as smooth on the low end, it's much more preferrable in terms of space and construction materials. I'm still uncertain which should be better. Both show really strong output according to HornRESP.
The 1/2 L per driver chamber size of the long horn worries me in terms of front radiation sound quality, but the horn output and front output are only directly out of phase at 160hz, where the horn output should be much greater.
The 1 L + chamber size for the short array seems better to me but the 2 waves are directly out of phase at 240hz which HornRESP show there is still 100db of output from the horn. Also,
the larger chamber give me more flexibility to reduce the chamber volume for fine tuning purposes.
With either, maybe the answer is to EQ down the horn output below it's upper cutoff frequency, then these out of phase points won't be significant because one output would be far greater than the other at these critical points.
Your point is why I haven't built 1 yet. I'm still hoping that someone with rear horn design experience steps in to offer some advice.
While the much shorter horn isn't as smooth on the low end, it's much more preferrable in terms of space and construction materials. I'm still uncertain which should be better. Both show really strong output according to HornRESP.
The 1/2 L per driver chamber size of the long horn worries me in terms of front radiation sound quality, but the horn output and front output are only directly out of phase at 160hz, where the horn output should be much greater.
The 1 L + chamber size for the short array seems better to me but the 2 waves are directly out of phase at 240hz which HornRESP show there is still 100db of output from the horn. Also,
the larger chamber give me more flexibility to reduce the chamber volume for fine tuning purposes.
With either, maybe the answer is to EQ down the horn output below it's upper cutoff frequency, then these out of phase points won't be significant because one output would be far greater than the other at these critical points.
Horn
Hey John,
Getting in on this thread late...I think I see what you are trying to do but I'm not sure I'd do it that way. I doubt you will be satisfied with the performance, running the driver that low below Fs. My immediate reaction is I doubt that you will be able to effectively load the horn much below 80 Hz or so(guessing). You would need to keep impedance peak issues controlled as the driver is no longer loaded properly by the horn at some point. This can be measured but you then need to alter the design as you go to compensate.
If you want to build a trial horn, use foam core laminated to sheet goods, and some duct tape. This is sold as 1.5" thick insulation (pink) panels in 4 x 8 sheets around here. Not terribly expensive either, and will provide the anti-resonant behavior you need, plus it's easy to cut and use for the interior dividers. A small jig saw works great on the hard board-backed type, or electric carving knife on just the rigid foam sheets. The driver mounting baffle, of course, would need to be solid.
I'm not wild about rear horns with a long pathway developing higher frequencies(above 150 Hz) either. They usually sound bad as they resonate the enclosure more than low bass frequencies(yeah, hard to believe), and they tend to roll off HF due to the folding. Rear horns can sound colored as well unless you damp out some of the reflective splash directly behind the driver.
I'd just build a straight front horn (along with a rear chamber) to the longest dimension you can live with, and go from there. Say, start with modeling at 100 - 150Hz cut off to keep the design small as possible, yet still give useable output down low enough to tie in with a sub plate amp. My guess is ~30" deep. Conical won't load as well low, but prove to be very good sounding in the mids do to the fast flare rate. Try an exponential flare if loading lower is an important design consideration.
Maybe use a couple of 15"s/18"s each side in a w baffle or the equivalent for dipole bass duties.
Sorry, probably not what you wanted to hear, but I just have my doubts... Can't hurt to try though as a trial unit so long as materials are kept cheap, and by building just one side.
Tim
Hey John,
Getting in on this thread late...I think I see what you are trying to do but I'm not sure I'd do it that way. I doubt you will be satisfied with the performance, running the driver that low below Fs. My immediate reaction is I doubt that you will be able to effectively load the horn much below 80 Hz or so(guessing). You would need to keep impedance peak issues controlled as the driver is no longer loaded properly by the horn at some point. This can be measured but you then need to alter the design as you go to compensate.
If you want to build a trial horn, use foam core laminated to sheet goods, and some duct tape. This is sold as 1.5" thick insulation (pink) panels in 4 x 8 sheets around here. Not terribly expensive either, and will provide the anti-resonant behavior you need, plus it's easy to cut and use for the interior dividers. A small jig saw works great on the hard board-backed type, or electric carving knife on just the rigid foam sheets. The driver mounting baffle, of course, would need to be solid.
I'm not wild about rear horns with a long pathway developing higher frequencies(above 150 Hz) either. They usually sound bad as they resonate the enclosure more than low bass frequencies(yeah, hard to believe), and they tend to roll off HF due to the folding. Rear horns can sound colored as well unless you damp out some of the reflective splash directly behind the driver.
I'd just build a straight front horn (along with a rear chamber) to the longest dimension you can live with, and go from there. Say, start with modeling at 100 - 150Hz cut off to keep the design small as possible, yet still give useable output down low enough to tie in with a sub plate amp. My guess is ~30" deep. Conical won't load as well low, but prove to be very good sounding in the mids do to the fast flare rate. Try an exponential flare if loading lower is an important design consideration.
Maybe use a couple of 15"s/18"s each side in a w baffle or the equivalent for dipole bass duties.
Sorry, probably not what you wanted to hear, but I just have my doubts... Can't hurt to try though as a trial unit so long as materials are kept cheap, and by building just one side.
Tim
Tim,
Thanks for stopping in. You've pushed me over the edge to try the short version (120hz upper cutoff). Regarding the impedance
issue, HornRESP is showing a 30ohm spike at 40hz and a few ohm rise at 120hz.
I realize that one or even a few of these little things won't load a horn properly. In fact HornRESP show only a spike at 80hz with just a few. Essentially, I'm talking about stacking 16 horn mouths together in a 2 meter stack. 16 tiny voice coils with a combined BL of 18 working together to move a cone area the same size as a 15", with 98db sensitivity combined. An added advantage of being tall is the tiny amount of corner space needed.
I think I can make one with 1 sheet of 1/2" ply and some scraps all of which I already have. I already have a 12 driver baffle and the 4"ers I have model a little better than the NSB's due to their slightly lower Fs of 95. The hard part is done, so I might be able to do it all in a day, just have to get off my butt and go for it. I just hope HornRESP is right.
Thanks for stopping in. You've pushed me over the edge to try the short version (120hz upper cutoff). Regarding the impedance
issue, HornRESP is showing a 30ohm spike at 40hz and a few ohm rise at 120hz.
I realize that one or even a few of these little things won't load a horn properly. In fact HornRESP show only a spike at 80hz with just a few. Essentially, I'm talking about stacking 16 horn mouths together in a 2 meter stack. 16 tiny voice coils with a combined BL of 18 working together to move a cone area the same size as a 15", with 98db sensitivity combined. An added advantage of being tall is the tiny amount of corner space needed.
I think I can make one with 1 sheet of 1/2" ply and some scraps all of which I already have. I already have a 12 driver baffle and the 4"ers I have model a little better than the NSB's due to their slightly lower Fs of 95. The hard part is done, so I might be able to do it all in a day, just have to get off my butt and go for it. I just hope HornRESP is right.
John,
There will be a delay between the direct radiator vs horn output due to the length of the pathway, but since the horn is working(now) at frequencies below the direct radiator output, you should not have the strong interaction and comb filtering problems that arise when producing the same frequencies.
In your experiments, I wouldn't hesitate to flip the baffle around, create a tall sealed rear chamber with the help of horn response, and call it a front horn. Of course you would lose the high freq output above cutoff from the direct radiator loss, but that's what numerous cheap drivers are for - just add another row in front with their own front horn flare - Ha!.
What ever it ends up being, it will be real damn loud, which is kinda fun by itself, but it may also give you a glimpse into the possibilities of the great bass available through horn loading.
Tim
There will be a delay between the direct radiator vs horn output due to the length of the pathway, but since the horn is working(now) at frequencies below the direct radiator output, you should not have the strong interaction and comb filtering problems that arise when producing the same frequencies.
In your experiments, I wouldn't hesitate to flip the baffle around, create a tall sealed rear chamber with the help of horn response, and call it a front horn. Of course you would lose the high freq output above cutoff from the direct radiator loss, but that's what numerous cheap drivers are for - just add another row in front with their own front horn flare - Ha!.
What ever it ends up being, it will be real damn loud, which is kinda fun by itself, but it may also give you a glimpse into the possibilities of the great bass available through horn loading.
Tim
Tim Moorman said:
I've built arrays, so I know it will be loud. Extension is what I'm looking for. Also, I got an incorrect formula for rear horns online and the first design put the horn length exactly at 1 wavelength of the acoustical crossover frequency, so it was back to the drawing board. If I do end up needing to go front horn plus another array for the front radiation, I'll go back to the original dimensions and wire the horn out of phase.
Isobarik loading may help too, with double the motor strength to move and half the Vas to move the same cone area. I need to model that.
IT WORKS....BUT.....!
I made a test horn today using an OB array I have. I used styrofoam and duct tape to close in the chamber and plywood to make a straight horn out of the back.
Although HornRESP told me I'd get 100db, I only got 87db at 40hz on 1 watt. The difference is due primarily to the makeshift construction. The styrofoam really vibrated and alot of sound was coming from there. Also, a straight conical horn is less efficient. In addition I wasn't using a corner or side wall and the 4ft tall narrow shape probably resulted in less benefit from the floor boundary.
Although the test was successful and I believe that 1 watt and 16 NSB's will get you 101db at 40hz in a compact but tall array horn, the limitation is that you can only put 1 watt into it. The cone excursion increases drastically below 50hz up to .7mm at 40hz, and with the xmax of 1mm even .7 is higher than you want to be.
I may go ahead and build a pair to go with my low powered tube amp, but my enthusiasm is shot because I really wanted to get to 110db on 10 watts from 40hz on up using cheap 4" drivers, but that just isn't going to happen and 50hz is too high to put a high pass filter.
I made a test horn today using an OB array I have. I used styrofoam and duct tape to close in the chamber and plywood to make a straight horn out of the back.
Although HornRESP told me I'd get 100db, I only got 87db at 40hz on 1 watt. The difference is due primarily to the makeshift construction. The styrofoam really vibrated and alot of sound was coming from there. Also, a straight conical horn is less efficient. In addition I wasn't using a corner or side wall and the 4ft tall narrow shape probably resulted in less benefit from the floor boundary.
Although the test was successful and I believe that 1 watt and 16 NSB's will get you 101db at 40hz in a compact but tall array horn, the limitation is that you can only put 1 watt into it. The cone excursion increases drastically below 50hz up to .7mm at 40hz, and with the xmax of 1mm even .7 is higher than you want to be.
I may go ahead and build a pair to go with my low powered tube amp, but my enthusiasm is shot because I really wanted to get to 110db on 10 watts from 40hz on up using cheap 4" drivers, but that just isn't going to happen and 50hz is too high to put a high pass filter.
It was really neither version because the plywood I had was in between the 2 horn lengths. I used the same throat adjusted for the 12 drivers instead of 16 and a mouth size the modelled similarly to my layouts.
It doesn't really make any difference. With that amount of Sd you need more excursion down low. You can double the drivers and enlarge the throat and chamber and get a similar response and cut excursion in half, but then you're in the same boat at 2 watts. If you drastically increase the horn length, you can get a more extended flat response and extend the controlled cone movement into lower frequencies.
I tried changes in driver specs and it made very little difference.
To get more output you need more excursion or much more total volume of your horn.
I still think that adding a bass horn is a good option for an array. I have some cheap 5" drivers that have more xmax that I'm going to try.
It doesn't really make any difference. With that amount of Sd you need more excursion down low. You can double the drivers and enlarge the throat and chamber and get a similar response and cut excursion in half, but then you're in the same boat at 2 watts. If you drastically increase the horn length, you can get a more extended flat response and extend the controlled cone movement into lower frequencies.
I tried changes in driver specs and it made very little difference.
To get more output you need more excursion or much more total volume of your horn.
I still think that adding a bass horn is a good option for an array. I have some cheap 5" drivers that have more xmax that I'm going to try.
John,
Well, at least you tried to experiment, which is more than I can say for a lot of the people. Nothing ventured...right? I think you just lost control of the cone from not being able to load the horn properly that low with that driver. Too far below resonance(Fs). You should see very little movement, if properly loaded.
If still interested in horn bass, I would try the Adire Tempest horn shown in the diy part of their site. Didn't you have a couple of subwoofers lying about? Big box, though.
Maybe just a single driver in a test horn with a suitable 90 - 100 Hz cut-off?
Tim
Well, at least you tried to experiment, which is more than I can say for a lot of the people. Nothing ventured...right? I think you just lost control of the cone from not being able to load the horn properly that low with that driver. Too far below resonance(Fs). You should see very little movement, if properly loaded.
If still interested in horn bass, I would try the Adire Tempest horn shown in the diy part of their site. Didn't you have a couple of subwoofers lying about? Big box, though.
Maybe just a single driver in a test horn with a suitable 90 - 100 Hz cut-off?
Tim
Thanks Tim,
Using HornResp, I've plugged a bunch of different drivers in the same horn with pretty similar results, even single 15's. Without the benefit of a much bigger horn to tune lower, it takes .5-.7mm of excursion to get to 100db at 40hz, so you need more excursion than a 1mm xmax to get more higher output and still sound good.
My goal was to try to do it just with NSB's, but a 100db limit isn't enough. I have a bunch of $2.50 5"ers that sound almost as good as my 4's. It looks like they have about 3 times the xmax and the modelling is showing I can get them up to 105db at 40hz and less than .5mm excursion with a 16 driver array in a total footprint of about 18"x18". Half of that is the final flare, so we're still talking about small, but tall.
There's no way to get into subwoofer territory without a big horn. I can't wait to have an attic or basement to hide one in.
Using HornResp, I've plugged a bunch of different drivers in the same horn with pretty similar results, even single 15's. Without the benefit of a much bigger horn to tune lower, it takes .5-.7mm of excursion to get to 100db at 40hz, so you need more excursion than a 1mm xmax to get more higher output and still sound good.
My goal was to try to do it just with NSB's, but a 100db limit isn't enough. I have a bunch of $2.50 5"ers that sound almost as good as my 4's. It looks like they have about 3 times the xmax and the modelling is showing I can get them up to 105db at 40hz and less than .5mm excursion with a 16 driver array in a total footprint of about 18"x18". Half of that is the final flare, so we're still talking about small, but tall.
There's no way to get into subwoofer territory without a big horn. I can't wait to have an attic or basement to hide one in.
The old rules of thumb for full sized horns: axail length = wavelength (HZ) at cut off. Mouth circumference equals wavelength at cut-off. 1140 ft/sec speed of sound/40 Hz=28.5
40 Hz = 28.5 Ft wavelength divided by 4 for 1/4 length= 7.125 ft. both in pathway and mouth circumference. 14.25 ft for 20 Hz.
These have been replaced by Marshall Leach's math, Horn Resp, and some other guidelines, but it will give you an approximation of the scale of things, as you probably now know.
They do grow rapidly with lower frequency. Foreshortening just adds irregularities to response.
Nice try, anyway.
Tim
40 Hz = 28.5 Ft wavelength divided by 4 for 1/4 length= 7.125 ft. both in pathway and mouth circumference. 14.25 ft for 20 Hz.
These have been replaced by Marshall Leach's math, Horn Resp, and some other guidelines, but it will give you an approximation of the scale of things, as you probably now know.
They do grow rapidly with lower frequency. Foreshortening just adds irregularities to response.
Nice try, anyway.
Tim
I cheated down on the mouth size for the corner loading (1/8 space). This would be for each driver, and you would gain some close coupling numerous drivers with the added total mouth size increased proportionately.
Fact is, a horn starts loading at 1/4 wavelength, but does an increasingly better job as you approach 1/2 wavelength, with mouth size about 1/3 wavelength or greater.
Most of the bass designs of recent release use the room boundaries to effectively extend the mouth, as you had originally planned with corner loading.
Tim
Fact is, a horn starts loading at 1/4 wavelength, but does an increasingly better job as you approach 1/2 wavelength, with mouth size about 1/3 wavelength or greater.
Most of the bass designs of recent release use the room boundaries to effectively extend the mouth, as you had originally planned with corner loading.
Tim
Tim,
By pathway do you mean horn length. I'm getting a good graph with a 143cm (4.7ft) horn length and a 6,368 sq cm (6.85 sq ft) mouth. Do you think that's compromising too much to sound good with corner loading?
I do think the actual results may be somewhere between 1/8 and 1/4 space due to the tall narrow shape, but the 1/4 space graph looks pretty good too.
By pathway do you mean horn length. I'm getting a good graph with a 143cm (4.7ft) horn length and a 6,368 sq cm (6.85 sq ft) mouth. Do you think that's compromising too much to sound good with corner loading?
I do think the actual results may be somewhere between 1/8 and 1/4 space due to the tall narrow shape, but the 1/4 space graph looks pretty good too.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.