USRFobiwan:
I am comparing 16 boxes of bass reflex totalling 2784 liters + material with one TH at 4817 + material.
Ofcourse we can look at the other aspects.
Power:
The maximum power seen by each driver is 32watts, vs 450watts for the TH.
Both can be powered by a reasonably well built amplifier.
But if you wish to do so you can power the BR's by 8 pieces of stereo flea watt amps. Then you have some redundancy, should one fail the rest will continue.
Practical sides:
If you wish to move the BR's you need a friend and a small truck. You can also divide the speakers pending on your need/application.
The TH must be either built inside the room it is intended for, or it must be a complicated puzzle at best.
Build:
The BR boxes are easy to build and will require moderate bracing.
The TH will be more difficult to build and require creative bracing.
I do not mind comparing an apple to a watermelon, it is merely the basis of the comparison that troubles me.
I am comparing 16 boxes of bass reflex totalling 2784 liters + material with one TH at 4817 + material.
Ofcourse we can look at the other aspects.
Power:
The maximum power seen by each driver is 32watts, vs 450watts for the TH.
Both can be powered by a reasonably well built amplifier.
But if you wish to do so you can power the BR's by 8 pieces of stereo flea watt amps. Then you have some redundancy, should one fail the rest will continue.
Practical sides:
If you wish to move the BR's you need a friend and a small truck. You can also divide the speakers pending on your need/application.
The TH must be either built inside the room it is intended for, or it must be a complicated puzzle at best.
Build:
The BR boxes are easy to build and will require moderate bracing.
The TH will be more difficult to build and require creative bracing.
I do not mind comparing an apple to a watermelon, it is merely the basis of the comparison that troubles me.
Going back to my original post:
It's clearly a case of: you have 100 points to distribute and each one you give won't be available to others.
A) Extension
B) Maximum SPL
C) Lowest Cabinet volume
So, to translate my first post into numbers:
A) 45%
B) 20%
C) 35%
Basically, the sacrifice would be the SPL*. On the other hand, Extension is basically a compromise by itself: 28hz is somewhat reasonnable and very far from the HT aficionados usual target...
Finally, cabinet size.
It's in fact very simple: my questionning is about the sweetspot. WHERE the law of diminishing return kicks in ?
(Any enclosure type)
is going from 20 liters to 40 liters would give 10db gain (best case scenario) while going from 40 to 80 liters (doubled again) will only give 8db gain (best case scenario again).... While at the extreme, going from 2000 liters to 4000 liters will only give 4db gain ?
ALL things considered, of course! Which means xmax/xmech of drivers, drivers available on the market right now, realistic power handling, and of course the 28hz target...
*A forced sacrifice... But still a target nonetheless.
It's clearly a case of: you have 100 points to distribute and each one you give won't be available to others.
A) Extension
B) Maximum SPL
C) Lowest Cabinet volume
So, to translate my first post into numbers:
A) 45%
B) 20%
C) 35%
Basically, the sacrifice would be the SPL*. On the other hand, Extension is basically a compromise by itself: 28hz is somewhat reasonnable and very far from the HT aficionados usual target...
Finally, cabinet size.
It's in fact very simple: my questionning is about the sweetspot. WHERE the law of diminishing return kicks in ?
(Any enclosure type)
is going from 20 liters to 40 liters would give 10db gain (best case scenario) while going from 40 to 80 liters (doubled again) will only give 8db gain (best case scenario again).... While at the extreme, going from 2000 liters to 4000 liters will only give 4db gain ?
ALL things considered, of course! Which means xmax/xmech of drivers, drivers available on the market right now, realistic power handling, and of course the 28hz target...
*A forced sacrifice... But still a target nonetheless.
Last edited:
Ok, you clearly missed the point of the example completely. I was just showing that resonant enclosures achieve sensitivity gains through enclosure size. I was NOT suggesting that anyone should consider building a TH like that, there are clearly diminishing returns and they kick in well before the horn gets anywhere near that size.
But since the response is almost ruler flat phase and group delay won't be nearly as bad as you might think.
No, I got the point regarding increase of SPL from increased enclosure volume.
What I do not get is why you'd choose to use a TH as an example, if going for a FLH with a size of 2785 liters you get nearly the same performance as the TH you posted, granted that you'd need steep HP filters on that also, and the delay is still there, but it's more constant, going to a 4804 liter FLH you get 133db down to 34 hz but you'd need a steep HP filter because it handles max 165 watts before reaching xmax at 30hz. I just can not understand why TH is more popular than FLH, or 1/4 wave horns for that matter.
And regarding the Group delay, it's 46ms at 36hz which might not matter too much, but the 17ms peak at 200hz is right in the xo frequency and that is certainly noticeable. It varies between 11-17ms between 80hz up to as far as you'd ever want to go in frequency.
That aside:
If you have just one single driver, you want max SPL, have a lumber yard and there is a snowballs chance in hell you'd get another driver, sure, build a 5000 liter horn.
Edit:
So yeah, maybe I missed some humorous point you had or something, just didn't see it.
What I do not get is why you'd choose to use a TH as an example, if going for a FLH with a size of 2785 liters you get nearly the same performance as the TH you posted, granted that you'd need steep HP filters on that also, and the delay is still there, but it's more constant, going to a 4804 liter FLH you get 133db down to 34 hz but you'd need a steep HP filter because it handles max 165 watts before reaching xmax at 30hz. I just can not understand why TH is more popular than FLH, or 1/4 wave horns for that matter.
And regarding the Group delay, it's 46ms at 36hz which might not matter too much, but the 17ms peak at 200hz is right in the xo frequency and that is certainly noticeable. It varies between 11-17ms between 80hz up to as far as you'd ever want to go in frequency.
That aside:
If you have just one single driver, you want max SPL, have a lumber yard and there is a snowballs chance in hell you'd get another driver, sure, build a 5000 liter horn.
Edit:
So yeah, maybe I missed some humorous point you had or something, just didn't see it.
Last edited:
Again, the ONLY point of that post was to show that resonant sensitivity is all about enclosure size. Nothing else. You would have to be some kind of idiot to build the sim I showed. I was not suggesting it is a viable alternative, just showing sensitivity gains of a large box.
I showed a tapped horn instead of a front loaded horn because that's what was being discussed, that's the only reason.
200 hz is well over an octave above a normal 80 hz crossover, a group delay peak at 200 hz wouldn't make any audible difference.
I'm not sure why you want to pick apart a 2 minute sim that was just posted as an example of sensitivity vs size, it makes no sense. No one is ever going to build that or even consider building something similar to that.
I showed a tapped horn instead of a front loaded horn because that's what was being discussed, that's the only reason.
200 hz is well over an octave above a normal 80 hz crossover, a group delay peak at 200 hz wouldn't make any audible difference.
I'm not sure why you want to pick apart a 2 minute sim that was just posted as an example of sensitivity vs size, it makes no sense. No one is ever going to build that or even consider building something similar to that.
An externally hosted image should be here but it was not working when we last tested it.
That's what i came up with so far.
Bass reflex, front slotted port, total system volume 69 liters (54 liters cabinet), just on the 19mm xmax/500w limits: 123-124db @ 30hz @ 1,0 pi
Honestly, I did get it.
i'm just having a complete lack of humour today, and that sim was so silly it was, well.. annoying.
On any other day I might have smiled.
I apologize.
Will try to read things in a more positive light.
Back to the matter at hand.
I think the point of diminishing returns is somewhere well before 1000 liters. I'd never go over 350 liters.
The main advantage of BR is that you just find the space you want to fill, choose a suitable driver (FS not more than 10hz over resonance of enclosure, qts 0,3 to 0,45) make the enclosure as much as a cube as possible (screw the "Golden Ratio", think more "near equal in all directions from center -port"), because the internal reflections will be too short to disturb the passband you intend to use it in (bass). This makes it easier to integrate in a system. Because there are no parts long enough to cause internal reflections of significance you will also have less phase variation, it will also have lower group delay and therefore sound more like a sealed box, this will not work quite as well if you sim a tall slim tower because you will get internal reflections that may interfere with the midrange and possibly reverberate through the cone. It is better to make many small cubes and stack them. You tailor the port to be big enough to have low enough particle velocity with the power intended for the design, and adjust length according to power response and practicality. I often choose to let the port be about the same length as the depth of the enclosure.
i'm just having a complete lack of humour today, and that sim was so silly it was, well.. annoying.
On any other day I might have smiled.
I apologize.
Will try to read things in a more positive light.
Back to the matter at hand.
I think the point of diminishing returns is somewhere well before 1000 liters. I'd never go over 350 liters.
The main advantage of BR is that you just find the space you want to fill, choose a suitable driver (FS not more than 10hz over resonance of enclosure, qts 0,3 to 0,45) make the enclosure as much as a cube as possible (screw the "Golden Ratio", think more "near equal in all directions from center -port"), because the internal reflections will be too short to disturb the passband you intend to use it in (bass). This makes it easier to integrate in a system. Because there are no parts long enough to cause internal reflections of significance you will also have less phase variation, it will also have lower group delay and therefore sound more like a sealed box, this will not work quite as well if you sim a tall slim tower because you will get internal reflections that may interfere with the midrange and possibly reverberate through the cone. It is better to make many small cubes and stack them. You tailor the port to be big enough to have low enough particle velocity with the power intended for the design, and adjust length according to power response and practicality. I often choose to let the port be about the same length as the depth of the enclosure.
Last edited:
What driver are you using again? It looks like it might benefit from using the Lossy Le tool in Hornresp.
Why 1 pi?
Why the big peak at tuning?
What's the port velocity?
If you use all of Hornresp's tools you can design a flared port which will save space and keep velocity at the port ends low. If you are interested in that but need help let me know.
There's probably no avoiding the "peakyness" because of the driver chosen.
Large sealed box, maybe a optimalized TL.
Large sealed box, maybe a optimalized TL.
That peak would normally be a problem but it's not for me since it's @ 30hz. I would have prefered the peak to be somewhere in the 26-28hz but that requires an even larger enclosure and port.
Why 1pi? I'm always looking at both 1pi and 2pi..
I'm a beginner with hornresp, i don't know how or if i can look for velocity. Should be ok but i should double-check with term-pro.
Yes, driver chosen (JL audio 8W7) is just an example. So far that's the best i found for the max SPL + lowest cut-off + smallest volume.
Why 1pi? I'm always looking at both 1pi and 2pi..
I'm a beginner with hornresp, i don't know how or if i can look for velocity. Should be ok but i should double-check with term-pro.
Yes, driver chosen (JL audio 8W7) is just an example. So far that's the best i found for the max SPL + lowest cut-off + smallest volume.
Check velocity by going to acoustic power window, then -
- tools, output, port, then
- tools, particle velocity
That's from memory, I don't have Hornresp open right now but that should be close enough to get you there.
You can probably do better for your goals with a different driver. Unless you really WANT that big peak at tuning.
- tools, output, port, then
- tools, particle velocity
That's from memory, I don't have Hornresp open right now but that should be close enough to get you there.
You can probably do better for your goals with a different driver. Unless you really WANT that big peak at tuning.
Thank you, justaguy, that's good to know.
Here is the results (at max power, 1pi)
Here is the results (at max power, 1pi)
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
That port velocity is too high. I like to keep it at or under 10 m/s, although other people are happy to go much higher. But over 10 m/s you start to get into compression losses and if you let it get too high you get chuffing and severe compression.
Like I said, a flared port will be smaller and have lower velocity.
If you want other driver choices just pick something that has qts around .45 and will work in the space you have designated. A different 8 or maybe even a 10 or a 12. Just do some sims and see what works well in the space you can afford.
Like I said, a flared port will be smaller and have lower velocity.
If you want other driver choices just pick something that has qts around .45 and will work in the space you have designated. A different 8 or maybe even a 10 or a 12. Just do some sims and see what works well in the space you can afford.
Based on my last build and comparison to the HornResp sim, 18 m/s seems to be Ok, based on the linearity tests that I did. Above that, measurable compression starts to set in and above 22 m/s, you're into "chuffing" territory. But that's just one data point, based on one build of a vented alignment.
That's at max power and only at the peak 30hz, though. But thanks, i will check for a flared port.
Problem with other drivers is i can't find any that is as well-balanced SPL/F3/Volume.
Since the max SPL is the sacrifice here, i need at least a driver with high power handling (from a trustworthy manufacturer in that regards) and also high xmax/xmech.
Also, i'd rather go with a 8in. if that is possible, unless a better 10-12in could do the job in the same volume.
Yeah, I know it's at max power and only at the peak at tuning. That's where it always happens and it's a problem. The whole point of the port is to provide a boost at tuning, if you have chuffing and compression you might as well not have a port at all. If you intend to use a high pass filter in real life you should include it in the sim and that will help lower the velocity.
What do you mean "check for a flared port"? If you want to use a flared port you can design it in Hornresp and build it right into the box. Here's how Danley did it.
Based on the big peak at tuning in your sim, it appears that the driver you chose is anything but well balanced.
I don't have time to sim or recommend anything at the moment (I have to go to work) but you can try simulating some other drivers in the box you already designed and see how they work.
If you need help implementing the flared port in Hornresp just ask, I can help with that later. It's really easy.
What do you mean "check for a flared port"? If you want to use a flared port you can design it in Hornresp and build it right into the box. Here's how Danley did it.
An externally hosted image should be here but it was not working when we last tested it.
Based on the big peak at tuning in your sim, it appears that the driver you chose is anything but well balanced.
I don't have time to sim or recommend anything at the moment (I have to go to work) but you can try simulating some other drivers in the box you already designed and see how they work.
If you need help implementing the flared port in Hornresp just ask, I can help with that later. It's really easy.
Ok, I don't have a lot of time today but I did do a quick sim with a 12 inch driver and this sim will also show you how to sim flared ports.
The driver is a 12 inch pro driver, I forget which one it was but it was pretty unique. You should be able to get similar results with other drivers though but you might hove to step down to 10 inch. If you really want to know what this driver is I can look it up but it won't be easy to find.
This sim uses a lot of power, 66 volts into 5 ohms is about 870 watts. That's the trade off when you have a box this small.
The box AND port in this design are 48 liters, well under the size of your sim.
I used 1 pi in this sim, not because I wanted to but to make it comparable with the sim you showed.
In your sim, if you add a high pass filter you will lose a couple of db at the peak at tuning. This design doesn't have a peak at tuning but it's 121 db at the lowest point in the passband and quite a bit more at the top of the passband so it beats the snot out of your sim. This design has as much spl as the PEAK in your design and more at higher frequencies.
The high pass filter is 4th order at 26 hz.
I tried to keep the tuning about the same as yours, the low knee is around 30 hz.
It should be easy to get more spl out of this if you make the box bigger, but as it is this design is louder and smaller than yours.
I used Flare It to determine the acceptable core velocity for the narrow part of the port. Flare It said 30 m/s was fine in the narrow part and I kept the port exit velocity to near 10 m/s as I like to do.
If you have any questions just ask.
The driver is a 12 inch pro driver, I forget which one it was but it was pretty unique. You should be able to get similar results with other drivers though but you might hove to step down to 10 inch. If you really want to know what this driver is I can look it up but it won't be easy to find.
This sim uses a lot of power, 66 volts into 5 ohms is about 870 watts. That's the trade off when you have a box this small.
The box AND port in this design are 48 liters, well under the size of your sim.
I used 1 pi in this sim, not because I wanted to but to make it comparable with the sim you showed.
In your sim, if you add a high pass filter you will lose a couple of db at the peak at tuning. This design doesn't have a peak at tuning but it's 121 db at the lowest point in the passband and quite a bit more at the top of the passband so it beats the snot out of your sim. This design has as much spl as the PEAK in your design and more at higher frequencies.
The high pass filter is 4th order at 26 hz.
I tried to keep the tuning about the same as yours, the low knee is around 30 hz.
It should be easy to get more spl out of this if you make the box bigger, but as it is this design is louder and smaller than yours.
I used Flare It to determine the acceptable core velocity for the narrow part of the port. Flare It said 30 m/s was fine in the narrow part and I kept the port exit velocity to near 10 m/s as I like to do.
If you have any questions just ask.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Also if this port stuff is getting too much for you just use passive radiators, that will save a bunch of space too. The amount of space saved will be the volume of the port. I don't like passives personally, I don't like paying for something that does the same job as a slug of air but YMMV especially if size is a major concern.
Thank you, Just a guy, that's great to see a comparison sim with a pro driver in such small volume! Also teaches me how to make flare port.
One thing, though: I personnally value better the 28-35hz. In fact, i tend to EQ-BOOST that area almost all the time (whenever possible), by at least few dBs
Now, what i see, is a potential limited at 121db, even though the unobtainium 12inch, even though the massive wattage.
Correct me if i'm wrong but:
1. A single 8W7 would exceed that 121db as soon as 28-30hz
2. With passive radiator(s), i will fix the velocity problem and get an even lower volume OR lower F3.
One thing, though: I personnally value better the 28-35hz. In fact, i tend to EQ-BOOST that area almost all the time (whenever possible), by at least few dBs
Now, what i see, is a potential limited at 121db, even though the unobtainium 12inch, even though the massive wattage.
Correct me if i'm wrong but:
1. A single 8W7 would exceed that 121db as soon as 28-30hz
2. With passive radiator(s), i will fix the velocity problem and get an even lower volume OR lower F3.
- Status
- Not open for further replies.