Your driver can handle 22 hz just fine, not sure what you mean by response becoming unstable.
I'm unclear of what your goals are. If it's pancake flat response over the widest bandwidth and with the lowest possible extension you can do better. Especially if you don't mind having a really weird flare shape and intend to use 8th order crossover and don't care about it going really loud.
Here's an example, it's about +/- 0.75 db from 15 - 90 hz. Shown here at xmax with no hpf. It only takes 100 watts. It doesn't go nearly as loud as yours but that's to be expected since it's 150 liters smaller than yours, tuned a lot lower and the huge rear chamber size is required as it's largely responsible for a large portion of the flat bandwidth.
This is just a quick 5 minute sim, I could get even flatter response over a wider bandwidth, and probably with a bit more peak spl. This is not a good design by any means, it's just to show that you can get any response curve shape you want and that your driver is just fine down to 22 hz and lower with flat response and wide bandwidth.
This is probably a good time to note that the flattest possible response over the widest possible bandwidth might not be a great goal if high performance is also a goal. Double up Re (or add Rg = Re) in your sim. That will give a very crude approximation of what response will look like under heavy power compression. Then do the same with a more normal horn example like the ones at the beginning of this thread, one with a more common flare shape, a bit of rising response and a bit of response ripple. You might find that the more normal flared horn suffers heavy power compression a bit better. This won't be a huge deal at the power levels you are simulating at, but if you ever do decide to add a high pass filter and turn it up this could become very important. And since you have dsp capability you can eq the response to be as flat as you want, it doesn't have to be flat right out of the box. A bit of rising response and response ripple isn't necessarily a bad thing in horns.
I'm unclear of what your goals are. If it's pancake flat response over the widest bandwidth and with the lowest possible extension you can do better. Especially if you don't mind having a really weird flare shape and intend to use 8th order crossover and don't care about it going really loud.
Here's an example, it's about +/- 0.75 db from 15 - 90 hz. Shown here at xmax with no hpf. It only takes 100 watts. It doesn't go nearly as loud as yours but that's to be expected since it's 150 liters smaller than yours, tuned a lot lower and the huge rear chamber size is required as it's largely responsible for a large portion of the flat bandwidth.
This is just a quick 5 minute sim, I could get even flatter response over a wider bandwidth, and probably with a bit more peak spl. This is not a good design by any means, it's just to show that you can get any response curve shape you want and that your driver is just fine down to 22 hz and lower with flat response and wide bandwidth.
This is probably a good time to note that the flattest possible response over the widest possible bandwidth might not be a great goal if high performance is also a goal. Double up Re (or add Rg = Re) in your sim. That will give a very crude approximation of what response will look like under heavy power compression. Then do the same with a more normal horn example like the ones at the beginning of this thread, one with a more common flare shape, a bit of rising response and a bit of response ripple. You might find that the more normal flared horn suffers heavy power compression a bit better. This won't be a huge deal at the power levels you are simulating at, but if you ever do decide to add a high pass filter and turn it up this could become very important. And since you have dsp capability you can eq the response to be as flat as you want, it doesn't have to be flat right out of the box. A bit of rising response and response ripple isn't necessarily a bad thing in horns.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Yeah, I know its not often that I'll hear low bass. If its in there though, I'd like to know about it. Since OSB and PL Premium are cheap and about all I'll have to buy, this sort of project is OK by me. Others wouldn't put up with it, but I don't mind the large size at all. I may build something to put on top of it, or maybe not. Its gonna have some of those small, strong rubber cabinet wheels under it for sure.
It may be a cheap project compared to your Keystone half the size and double the output since no need to buy the driver. Maybe not, depends on the driver cost.
My driver won't go any lower stably, or this cabinet would be even bigger. This horn seems to be near a minimum size to reach the lowest extension my driver will stably perform at while allowing a flat response with a fair amount of sensitivity. Like I said I can get more sensitivity from this size cabinet and smaller, but then I'll have to abandon some of the low end extension thats essentially already free. I'm trying to get everything thats down there, flat across the usable band, with low power demands, and heavy on the cheap.
I will be building a pair of full range cabs later on, I'll make those a lot smaller. 🙂
It may be a cheap project compared to your Keystone half the size and double the output since no need to buy the driver. Maybe not, depends on the driver cost.
My driver won't go any lower stably, or this cabinet would be even bigger. This horn seems to be near a minimum size to reach the lowest extension my driver will stably perform at while allowing a flat response with a fair amount of sensitivity. Like I said I can get more sensitivity from this size cabinet and smaller, but then I'll have to abandon some of the low end extension thats essentially already free. I'm trying to get everything thats down there, flat across the usable band, with low power demands, and heavy on the cheap.
I will be building a pair of full range cabs later on, I'll make those a lot smaller. 🙂
@ just a guy,
I have no clue how you did that. All my sims have a 22 hz dip in them when I go that low.
Also, I thought a flat response was the best response, but maybe rising is better?
I have no clue how you did that. All my sims have a 22 hz dip in them when I go that low.
Also, I thought a flat response was the best response, but maybe rising is better?
I see it now after entering your sim. Just forget about high sensitivity and it works... Hmm...
Big rear chamber.
If you have dsp it really doesn't matter either way. Ideally you design the response to be the inverse of the environmental loading so it sums to flat. But in this case it's going to be used inside, outside and probably other places so the target frequency response goal is whatever you want it to be.
The big point here is that in undersized front loaded horn a wild, spiky response is more likely to handle power compression more gracefully than a super flat response. Assuming your cab is undersized (but still reasonably sized), making it larger will make the resonance peaks at or near the impedance peak frequencies larger. These larger resonant peaks are an indication that the response is relatively undamped and acoustically large enough.
Power compression presents as an increase in Re (it's more complex than that but that's the main problem). When you increase Re and keep the cab size the same the response becomes more damped, sensitivity goes down, and as Re increases to double it's orginal value the response across the entire passband droops. The impedance peaks keep the response propped up (somewhat like telephone poles) but between the impedance peaks the response droops down (somewhat like the droopy wires between the poles).
The higher Re wants a much larger cab. If you start with a larger cab in the first place it settles into power compression much more gracefully.
And forget about the integrity of a true front loaded horn that you were so worried about in the beginning. At this point it's really got not much resemblance to a front loaded horn at all. It's just a tl with a wacky line shape and a closed rear chamber.
Anyway, if this kind of weird flare shape is interesting to you, make sure to check velocity at the mouth, it can get really high if the mouth gets too small. I checked on the sim I showed, it's still down under 13 m/s but if you don't watch out it can get out of hand.
Yeah I get it now, but still don't get it enough to be able to go out and do it better. Thing is I don't know how to solve enough of these problems at once. Dumb rookie has hit the wall... 🙁
Still having fun though. Not giving up.
Edit: Low sensitivity not on my wish list. Seems to me that high sensitivity helps a lot to save power.
Still having fun though. Not giving up.
Edit: Low sensitivity not on my wish list. Seems to me that high sensitivity helps a lot to save power.
Last edited:
This is what the Hornresp sliders are for. You can try a tremendous amount of tweaks in a very short time. How do you think I came up with that design? There's no mathematical formula, I played with the sliders until I got the response I wanted. I have simulated this type of ultra flat response before so I knew the general direction to take (large rear chamber, cigar shaped flare with a small mouth) and it took about 5 minutes.
The cure for low sensitivity is cab size or higher tuning. No way around that. To bring that Frankenstein cab I posted up to a higher sensitivity with the same low knee frequency and a flat passband is going to take a monster size cab.
Define your goals and pick your compromises. It's a game, have fun with it.
To address this specifically - what does save power mean to you? My sim only takes 100 watts to hit xmax. High sensitivity means more spl per watt, but it does not guarantee that you can use up your excursion with a small amount of power.
The key to a good design is to use up all your usable excursion with a small amount of power to reduce power compression. There's two ways to use up excursion fast - low tuning and/or large cab size.
If you want high sensitivity AND the ability to use all your excursion with low power levels, they only way to do that is with a relatively large cab and a relatively high tuning.
Yep, yep! Keep smiling!
I am happy with the response, the passband width, and the sensitivity of the horn I just posted. I am also happy that with the size of that beast it may be easy to change a lot of stuff later if there are any problems folding it up. But I didn't address any power compression issues.
The first thing I would try to do to improve that horn would be to try for less power, for the same level of performance.
Edit: Less power for the same or better level of performance. I wouldn't waste time on it though, if sensitivity could not be increased.
I'm not into a rippled response. I drink too much coffee as it is trying to knock that down. I have no problem with a slightly rising response though, as long as its smooth, because the room will reinforce somewhat. Since I'll be shoving the cab out onto the porch often, flat should an OK compromise shouldn't it?
Likely the funnest stuff are yet to come.
Last edited:
Less power required to reach xmax with the same low knee and more sensitivity = bigger. Simple as that.
You bring up a good point here. The room is going to chop that pretty response right up with peaks and dips that could end up being +/- 15 db in room response. Compared to that what's a bit of 3 db ripple in the horn response? Seriously you won't even notice it, even when it's outside. And if you did happen to notice it you have fully featured dsp to eq it flat.
But it's up to you. Do what you like. This type of ultra flat response is very easy to sim so maybe ask yourself why you've never seen anything like it before in any commercial or diy project. (I did one a long time ago just for fun to see how flat I could get it but I wouldn't do it again.)
Here's the measured response of the one I made, no filters or eq applied, top line light green. It was a 6 inch driver, the horn was front loaded maybe 60 - 80 liters or something and the dip at 90 hz didn't sim as deep as it measured so maybe I screwed something up in the construction, I didn't bother to investigate as it didn't really matter. It was winter so it was an in room measurement with the mic right at the horn mouth so maybe that has something to do with the dip. It stomps all over a ported box with the same driver and low knee but it could probably be beat by a well designed horn with a more traditional layout at the spl limits.
An externally hosted image should be here but it was not working when we last tested it.
After rereading everything tonight I'd like to ask a bit more about this. I'd better explain my angle of the dangle first though.
When I speak of my desire for high sensitivity that means the same to me as high SPL from per watt power. Finding higher sensitivity seems like a shortcut to greater SPL from lower power levels.
I understand that its desirable to use up all your excursion with the least power input. Thats something I hope for too. It seems to me that high sensitivity helps you to reach xmax sooner, on less power.
What I don't know is where will power compression become an issue? I tried inputting various figures into the EG field in Hornresp and clearly see the results.
Will it be a problem at 10 watts, or 100 watts, or 300 watts etc...? I'm not asking you to tell me that figure.... but are there any methods for guesstimating when power compression may occur in the absence of dynamic measurement?
Thanks,
FWIW the bottom line is a tapped horn with the same driver and I like the sound of that horn MUCH better. It's a completely different (lower) tuning, it's probably 20 liters smaller, it's made of 5/8 mdf (the flh is 3/8 plywood), it doesn't go as loud (due to the lower tuning and smaller size) and the response isn't anywhere near flat. But it sounds fantastic and it's the one I pull out of the closet and fire up if max spl is not a requirement.
High sensitivity = more spl per watt. It DOES NOT help you reach xmax sooner. Excursion depends on tuning and enclosure size, it has nothing at all to do with sensitivity directly.
Eg is applied voltage, it won't show you anything about power compression. Re or Rg is what you need to increase to crudely simulate power compression.
Will it be a problem at 10 watts, or 100 watts, or 300 watts etc...? I'm not asking you to tell me that figure.... but are there any methods for guesstimating when power compression may occur in the absence of dynamic measurement?
Thanks,[/QUOTE]
As I explained before, it depends totally on the design and the music you play. If you play girls and guitars music with no bass it will never be an issue. if you play classic rock with low level bass guitar and sporadic bass drum hits it won't be an issue until you get to high power levels, maybe up near the driver's rated power handling or even higher.
If you play a track like this one, with an almost constant bass note that has all the intensity of a pure constant sine wave, and that note is at a low efficiency point in the passband, and you play that track on repeat for a few hours, then you might get power compression kicking in at power levels as low as 1/4 the rated driver power handling.
You have to analyze the enclosure efficiency and driver power consumed at various points in the passband and you have to analyze your tracks to find out what notes the bass hits at, what level the bass is at, what frequency the bass line plays, what intensity (sine wave, square wave, etc) and what duty cycle (dynamic or constant and for how long).
And even then you can only get a rough estimate because you can't easily guess how effective the driver cooling is, how much the cone motion will "fan" away the heat what the ambient temperature will be on any given day, and a bunch of other things.
To keep things really simple - like I already said - at the power levels you showed in your last sim, power compression is not likely to be a big issue. If you want a more detailed answer, good luck trying to figure it out.
Hmmm, there is no way to know in advance what is going to happen to response in room, so its better to go for extension and/or max spl or other goal?
If no one knows what they'll get why bother with any of it? Can't a 3 Db ripple left purposely in the response be turned into a 12-15 Db dip in room?
Does a rising response strategy improve odds that the room won't put a null in the middle of your passband?
If no one knows what they'll get why bother with any of it? Can't a 3 Db ripple left purposely in the response be turned into a 12-15 Db dip in room?
Does a rising response strategy improve odds that the room won't put a null in the middle of your passband?
I'm just saying 3 db ripple won't be a big problem, that has nothing to do with your other goals like extension, max spl, etc (at least not directly).
Who says no one knows what you will get? You can measure the room response and then make the sub's frequency response the inverse of that so it sums flat. But if you plan to use it in multiple locations you can't do that quite as effectively. The environmental acoustic effect of ANY given room vs outside will be dramatically different so there's no perfect response to shoot for.
For example, if the sub is a permanent build in unit and won't ever move you can measure the room response, figure out where the room modes are and make sure the horn resonances are not at the same frequencies. You make the sub's response compliment the room's response. Ideally if there is a peak in room response you would have a dip in the subwoofer response at the same frequency and vice versa. You won't get a perfectly flat summed response but you can do better than random.
No. Room gain is usually starting at 30 hz and increases up to 12 db/oct below 30 hz. There's a practical limit, if the walls are drywall on stud with vinyl exterior they won't contain much low bass so there won't be as much gain at low frequencies, so the room gain curve isn't linear. If the room is the size of a large warehouse and not well sealed there will be no room gain at all. There is obviously no room gain outside. Room gain generally only happens at low frequencies and will not cause sharp peaks or nulls - room modes are higher in frequency and those are what cause the sharp peaks and nulls.
A rising response in the sub will often compliment room gain at low frequencies.
And none of this really matters much anyway. You have fully featured dsp so you can use it to get perfectly flat response in any environment.
Last edited:
These are kickbutt in depth answers just a guy. All of it tonight! Thanks a million for taking the time and effort to help. Much gratitude.
To expand on this and maybe make it more clear - high sensitivity and using up the usable excursion are actually conflicting goals.
Tuning low to use up excursion actually hurts the low frequency sensitivity, a lot.
The only way you can keep the low tuning and use up excursion is a very large cab, which can get out of hand very quickly as I'm sure you have noticed.
Very high excursion drivers like this one need a low tuning to use up their excursion before hitting power limits. Or extremely large enclosures. Or both.
If I don't use a HPF excursion will be used up at a lower power level, so in my mind, why not go for higher sensitivity, so a higher SPL level will be reached at lower power.
If a higher sensitivity allows SPL levels to become extreme (disassembling the house and driving humans out) enough, no need to worry about max excursion, because no one dares to crank the volume knob up.
No worries for power compression either. In fact, by hooking up a small enough amp (or limiting the signal to the amp via DSP) to match the maximum SPL violence level you are willing to tolerate, none of those bad things are likely to happen.
This can be a design strategy. Err... is.
Again, thanks just a guy. You spent a lot of your time helping me tonight and have been tirelessly doing so throughout this entire thread. I want you to know that I am appreciative.
Edit: Or are you saying that excursion tracks with sensitivity, so can occur at moderate SPL levels.
Edit 2: Nevermind I will test for it.
If a higher sensitivity allows SPL levels to become extreme (disassembling the house and driving humans out) enough, no need to worry about max excursion, because no one dares to crank the volume knob up.
No worries for power compression either. In fact, by hooking up a small enough amp (or limiting the signal to the amp via DSP) to match the maximum SPL violence level you are willing to tolerate, none of those bad things are likely to happen.
This can be a design strategy. Err... is.
Again, thanks just a guy. You spent a lot of your time helping me tonight and have been tirelessly doing so throughout this entire thread. I want you to know that I am appreciative.
Edit: Or are you saying that excursion tracks with sensitivity, so can occur at moderate SPL levels.
Edit 2: Nevermind I will test for it.
Last edited:
- Status
- Not open for further replies.