Hello
After read a lot of threads on diyaudio forum I became excited to replace my current T18 cabinet for another Horn type but first i will need an additional help from you.
considering those models bellows, I highlight the back chamber volume in blue from a 2D scheme to correlate with some driver options.
click here for larger image: https://s9.postimg.org/5ohtvxta7/Cabinets.png
Thinking about keep the original design or making small tuning for specific drivers I found one equation from the D.B. Keele Jr publication to design Horn using T/S parameters and doing the math I filled the table below.
click here for larger image: https://s22.postimg.org/pxim20ob5/Data.png
Taking Keele's equation in mind the question is:
What happen if the cabinet has higher Vb them the drivers need?
What happen if the cabinet has lower Vb them the drivers need?
How much can I expect to increase Vb volume using acoustic materials like fiber glass and modern ones?
After read a lot of threads on diyaudio forum I became excited to replace my current T18 cabinet for another Horn type but first i will need an additional help from you.
considering those models bellows, I highlight the back chamber volume in blue from a 2D scheme to correlate with some driver options.
An externally hosted image should be here but it was not working when we last tested it.
click here for larger image: https://s9.postimg.org/5ohtvxta7/Cabinets.png
Thinking about keep the original design or making small tuning for specific drivers I found one equation from the D.B. Keele Jr publication to design Horn using T/S parameters and doing the math I filled the table below.
An externally hosted image should be here but it was not working when we last tested it.
click here for larger image: https://s22.postimg.org/pxim20ob5/Data.png
Taking Keele's equation in mind the question is:
What happen if the cabinet has higher Vb them the drivers need?
What happen if the cabinet has lower Vb them the drivers need?
How much can I expect to increase Vb volume using acoustic materials like fiber glass and modern ones?
If the rear chamber is too small you will likely get a dip in response right above the low knee. If the rear chamber is too large you will use more excursion than you need to because the sealed chamber is too large to keep excursion in check.
When dealing with massively undersized horns like these forget about equations. Sim the box with the driver you intend to use and see how it's going to perform. The calculations are intended to be used to design full sized horns, not tiny horns like this.
When dealing with massively undersized horns like these forget about equations. Sim the box with the driver you intend to use and see how it's going to perform. The calculations are intended to be used to design full sized horns, not tiny horns like this.
Also, it appears these horns are all unnecessarily complex, they all have more wasted space than necessary, and all use unnecessary corner deflectors. IIRC the 186 and 1850 are tuned very high and I'm not familiar with the WSX but if the drawings are to scale then it is tuned very high as well.
Thanks "just a guy".
For kind of complex design it demands more complex software them hornresp and I don't know if I will be able to use it
but i will give a try.
Behind simulation there is math so simple equations can indicate some direction
WSX is bigger then 1850/186 and has small throat area and larger mouth area so it is tunned lower, I supposed something close to 40Hz once the original data sheet indicates frequency range from 30~300Hz.
For kind of complex design it demands more complex software them hornresp and I don't know if I will be able to use it
but i will give a try.Behind simulation there is math so simple equations can indicate some direction
WSX is bigger then 1850/186 and has small throat area and larger mouth area so it is tunned lower, I supposed something close to 40Hz once the original data sheet indicates frequency range from 30~300Hz.
Thanks "just a guy".
For kind of complex design it demands more complex software them hornresp and I don't know if I will be able to use it
If it's the fact that hornresp doesn't have enough segments you can use TL.app or Akabak. If you don't want to do that this is another method to use hornresp's limited amount of segments to give the most accurate sim possible.
Analyze the cab.
Redraw as a straight horn.
Draw straight lines across the top of the redrawn straight horn to determine the best fit spots for the segment markers.
And if you don't want to do any of that just look at the measured response of the 186 and 1850. I couldn't find a measured response of the WSX but it is specified at -3db at 38 hz and the manual specifically said it's meant to be used in stacks of 4 units or more.
Behind simulation there is math so simple equations can indicate some direction
If you REALLY know what you are doing you might be able to get a bit of insight from simple equations but it's like building a go cart based on the math a car designer would use. These undersized horns are not the same thing that the simple equations are made for.
Based on the math a 40 hz horn needs a mouth size about 30000 sq cm and the horn will have a volume of about 2000 liters. So unless you are planning to design the horn and make it HUGE or modular the simple equations are not really related to what you are trying to do.
The first thing you should do is lay out some goals. Why do you want a different cab than the one you have? What frequency response do you want? Those two questions alone will narrow down the potential designs by a huge amount.
hey just a guy,
I will give a try going deep on the simulations, i also found a different software "AJHorn" looks like it was made by Germany guys. It also have a nice tutorial, take a look.
AJHorn Manual
My goal is to have a cab with high efficiency as possible to create high SPL using less energy (eletric watts) using 1 driver/cab.
I intent to use 2 or 4 cabs with frequency response < 150Hz and low cut off lowest as possible, 40Hz would be enough.
I will give a try going deep on the simulations, i also found a different software "AJHorn" looks like it was made by Germany guys. It also have a nice tutorial, take a look.
AJHorn Manual
My goal is to have a cab with high efficiency as possible to create high SPL using less energy (eletric watts) using 1 driver/cab.
I intent to use 2 or 4 cabs with frequency response < 150Hz and low cut off lowest as possible, 40Hz would be enough.
- 186 offer 64Hz
- 1850 offer 54Hz with the same size as 186
- WSX is little bigger but looks like offer lower response 38Hz, is a good trade off
- Super Bass Horn (looks like it's a 1850/186 modified by French guys) is even bigger then WSX so i don't think it worth.
- Labhorn also the bigger one i found but it uses 2 drivers so it don't fit my goal.
Last time I looked at AJHorn was about 10 years ago and it was all in German so I ended up going with Hornresp instead and I'm very happy with that decision.
A quick scan through the linked manual suggests that Hornresp can do everything AJHorn can do and a whole lot more. In fact it looks like AJHorn can only do one segment, so Hornresp is exponentially better on that front.
It seems you are destined to go your own way on this but if you want a very easy to use program that has virtually unlimited amounts of segments you want to try TL.app by Leonard Audio. Just don't use it for tapped horns, that part wasn't working last time I checked.
Transmission Line | Leonard Audio
I'm not familiar with any of these except Labhorn. Do you even have plans to build the Martin Audio stuff? The MTH looks like it has a huge amount of wasted space inside the cab, this is just stupid. I wouldn't trust anything on the freespeakerplans website without simulating it first, and if you can sim it you can design your own (if you want to).
Efficiency is largely a matter of cab volume. The bigger the cab the more efficiency you can get out (generally, assuming good design). There is no magic alignment or design that does way better than others. They all follow the rules of physics. For a given cab volume and low knee frequency you will get roughly the same efficiency no matter what type of cab you make. (This includes multi resonant types, not sealed.)
The Labhorn is a good design but could use a refold and a modern driver to get rid of the big dip right above the low knee (from a too small rear chamber) and to give it dramatically higher power handling. You could pretty easily design your own Labhorn style horn - meant to be used in a stack for highest efficiency and smoothest response but still usable as a single cab.
None of this stuff is all that hard, you just need a lot of practice with a simulator.
If designing your own isn't in the cards I would look at some of the designs on this forum - the Xoc1 TH 18 35 hz tapped horn which is VERY similar to a commercial Danley design, or maybe the Keystone. These designs are well vetted, they use modern easily accessible drivers, they can handle a ton of power and they can get very loud.
And I would finish your T18 cab or whatever it's called, at least so you have some experience and something to compare other designs against.
Hmm.. maybe the LH could be reverse-engineered to see if any existing drivers would be a better match for its characteristics than the Eminence driver. HornResp does provide the option of varying the driver's parameters to see the impact that would have on the system's performance.
I'm planning to improve my knowledge about cab design before build new ones. I'm an enthusiastic and a like build cabs but till now i was not using enough engineering.
The T18 project in fact is done since 2 years ago and i'm using it to learn how to simulate and to check if i chosen a good driver for it.
Take a look. It sound really good, mainly between 60Hz~100Hz, but you know, we always wanna improve things and i still miss some low frequencies in addition to increase the upper frequency till 150Hz.
Click here for larger image: https://s12.postimg.org/3xklmf3ot/P_20161009_153907.jpg
Regarding the plans i have all plans and all 3D models already. For sure i will simulate before build anything and the idea is to replace my T18.
Click here for larger image: https://s16.postimg.org/e7heocnd1/all_models.png
Regarding the models I've been looking for a while i can comment:
MTH-4664 - Tapped horn (no rear chamber)
MHB-46 - Front loaded horn (tight throat) with vented rear chamber to improve low frequencies thru the ports.
WMX - front loaded horn (tight throat) with vented rear chamber to improve low frequencies thru the ports. (Martin Audio).
T18 - Looks like it's a hybrid solution too, front band pass loading a horn without any rear chamber
1850/186 - Front loaded horn with rear chamber
AB-36C - Front loaded horn (tight throat) with rear chamber (Cerwin Vega)
WSX - Front loaded horn (tight throat) with rear chamber (Martin Audio)
LabHorn - Front loaded horn (tight throat) with rear chamber using 2 12" drives.
There is also the new cab using what they call as k-horn
I've also collecting pros and cons from each model but it is a hard work so i'm trying ti learn as much as I can with you.
The T18 project in fact is done since 2 years ago and i'm using it to learn how to simulate and to check if i chosen a good driver for it.
Take a look. It sound really good, mainly between 60Hz~100Hz, but you know, we always wanna improve things and i still miss some low frequencies in addition to increase the upper frequency till 150Hz.
An externally hosted image should be here but it was not working when we last tested it.
Click here for larger image: https://s12.postimg.org/3xklmf3ot/P_20161009_153907.jpg
Regarding the plans i have all plans and all 3D models already. For sure i will simulate before build anything and the idea is to replace my T18.
An externally hosted image should be here but it was not working when we last tested it.
Click here for larger image: https://s16.postimg.org/e7heocnd1/all_models.png
Regarding the models I've been looking for a while i can comment:
MTH-4664 - Tapped horn (no rear chamber)
MHB-46 - Front loaded horn (tight throat) with vented rear chamber to improve low frequencies thru the ports.
WMX - front loaded horn (tight throat) with vented rear chamber to improve low frequencies thru the ports. (Martin Audio).
T18 - Looks like it's a hybrid solution too, front band pass loading a horn without any rear chamber
1850/186 - Front loaded horn with rear chamber
AB-36C - Front loaded horn (tight throat) with rear chamber (Cerwin Vega)
WSX - Front loaded horn (tight throat) with rear chamber (Martin Audio)
LabHorn - Front loaded horn (tight throat) with rear chamber using 2 12" drives.
There is also the new cab using what they call as k-horn
I've also collecting pros and cons from each model but it is a hard work so i'm trying ti learn as much as I can with you.
I found also the project you mention TH18 Xoc1 with the detail to simulation on hornresp, it looks an interesting solution. Is it true that the TH there is no delay?
Thanks too to Brian Steele once i found his excel file "Stepped TH"
Those two things together will allow me to really speed up the knowledge
Thanks too to Brian Steele once i found his excel file "Stepped TH"
Those two things together will allow me to really speed up the knowledge
Most of the output is coming from the throat side tap, the throat side tap is several feet from the mouth so there is going to be delay.
If you are looking for resources to learn about horns, also check out Soho54's "Eazy Horn" spreadsheet, the "System Design - With Driver" tool in Hornresp (a particularly good implementation of Leach's math from the "On the specification of moving coil drivers for low frequency horn loaded loudspeakers" paper, which is generally considered one of the best horn models available - http://thehearingblog.com/wp-conten...or-Low-Frequency-Horn-Loaded-Loudspeakers.pdf
I can't recommend strongly enough that practice with the simulator will help you learn more than reading ever will. Design your own, learn to fold and you will never have to rely on designs from other people again - unless you find one that meets your goals and you want to use it.
Hey just a guy, thanks a lot for you support, all new reply you do improves my knowledge.
I found everything i need to improve simulations.
If possible i'd like to ask you one more information considering the TH18 Xoc1 as the design reference once i got the model and my result is exactly the same as the original one.
What would be the advantage or disadvantage to add throat adapter on this cab design? like Martin Audio WSX has and as indicated on the image below.
Click here for larger image: https://s4.postimg.org/kq670osvx/throat_adapter.png
And in terms of trade of, looks like the point for subwoofer is trade SPL for Extension right? so which tips you could give me? Path length means low frequency cut off? is there any other?
I found everything i need to improve simulations.
If possible i'd like to ask you one more information considering the TH18 Xoc1 as the design reference once i got the model and my result is exactly the same as the original one.
What would be the advantage or disadvantage to add throat adapter on this cab design? like Martin Audio WSX has and as indicated on the image below.
An externally hosted image should be here but it was not working when we last tested it.
Click here for larger image: https://s4.postimg.org/kq670osvx/throat_adapter.png
And in terms of trade of, looks like the point for subwoofer is trade SPL for Extension right? so which tips you could give me? Path length means low frequency cut off? is there any other?
I don't know, I'd have to simulate it and find out. But I do know that the Danley TH118 has a similar restriction but it's not in the middle, it's on both sides of the driver. Something like this -
An externally hosted image should be here but it was not working when we last tested it.
Basically, yes.
Path length has a lot to do with flare shape too, not just length. Read MJK's Alignment Tables paper for a quick look at how flare shape and size affect tuning - the shape and the size is at least as important as length in tl tuning.
The best advice I have is to play with a simulator. Especially one like Hornresp that can make changes quickly with the sliders. It's my experience that you can learn a lot more a lot faster that way than by reading.
If you have specific questions it's a lot easier to answer than giving a list of tips.
backing to share additional results
First and must important one.: The best driver for LF presents the best results for the major cabs So it simplify my life once I can reduce the simulations just for those good drivers.
I got the TH18 Xoc1 plan/input data for hornresp and did simulations for some drivers available at my market.
click here for larger image: https://s10.postimg.org/g3yz1q7x5/TH18_results.png
I also got the OTHorn plan/input data for hornresp and did simulations for some drivers available at my market.
Click here for larger image: https://s14.postimg.org/nor2gttcx/OTHorn_results.png
overwriting those two results
Click here for larger image: https://s16.postimg.org/lj2tx93o5/TH18_vs_OTHorn_results.png
For both designs the driver Snake HPX2180 presents little more SPL close to 40Hz/53Hz. I've seeing guys desiring flat response for cabs, but considering the loudness effect for human ears wouldn't be better the HPX2180 response (less flat)?
and simplifying the results for comparison we have:
blue = TH18 performance
green = OThorn performance
gray area = Efficiency lost over 40Hz
red area = Efficiency gain under 40Hz
so simulation confirms what just a guy confirmed too regarding subwoofers being a trade off between SPL and Extension. Maybe we could just add the fact that we will pay the price to go down with the cab volume increasing. No magic right?
Click here for larger image: https://s14.postimg.org/mq10gefm9/FR_TH_comparisson.png
Do you guys knows other designs for 18" drivers that are able to go down to 30Hz as OTHorn do?
First and must important one.: The best driver for LF presents the best results for the major cabs
So it simplify my life once I can reduce the simulations just for those good drivers.I got the TH18 Xoc1 plan/input data for hornresp and did simulations for some drivers available at my market.
An externally hosted image should be here but it was not working when we last tested it.
click here for larger image: https://s10.postimg.org/g3yz1q7x5/TH18_results.png
I also got the OTHorn plan/input data for hornresp and did simulations for some drivers available at my market.
An externally hosted image should be here but it was not working when we last tested it.
Click here for larger image: https://s14.postimg.org/nor2gttcx/OTHorn_results.png
overwriting those two results
An externally hosted image should be here but it was not working when we last tested it.
Click here for larger image: https://s16.postimg.org/lj2tx93o5/TH18_vs_OTHorn_results.png
For both designs the driver Snake HPX2180 presents little more SPL close to 40Hz/53Hz. I've seeing guys desiring flat response for cabs, but considering the loudness effect for human ears wouldn't be better the HPX2180 response (less flat)?
and simplifying the results for comparison we have:
blue = TH18 performance
green = OThorn performance
gray area = Efficiency lost over 40Hz
red area = Efficiency gain under 40Hz
so simulation confirms what just a guy confirmed too regarding subwoofers being a trade off between SPL and Extension. Maybe we could just add the fact that we will pay the price to go down with the cab volume increasing. No magic right?
An externally hosted image should be here but it was not working when we last tested it.
Click here for larger image: https://s14.postimg.org/mq10gefm9/FR_TH_comparisson.png
Do you guys knows other designs for 18" drivers that are able to go down to 30Hz as OTHorn do?
That curve shape that the HPX2180 is showing in both cabs is a clue that the enclosure is too small for that driver. I would say without a doubt the HPX2180 is by far the WORST choice based on these graphs alone.
I would suggest that you add a few ohms extra resistance to get a crude idea of what heavy power compression is going to do to your pretty graphs. Either double Re or add Rg = Re and rerun the sims. The HPX2180 is going to completely fall apart with a terrible response shape while some of the others will fare quite a bit better. This extra resistance is a window into what it's going to look like if you run the cabs hard and get the drivers hot.
The next thing I would suggest is that you simulate at the spl limits, not at 1 watt or whatever power level you are using. Take the xmax and power handling into consideration and increase power for ALL the cabs with ALL the different drivers until you reach the power handling or xmax limit. That's much more useful that what you are doing. You are showing sensitivity based on equal power input (and it might not even be equal power input unless you are resetting the power for each driver by actually giving it 1 watt instead of 2.83 V or whatever power level you are using). It's more useful to show max spl instead. If one or two of those drivers have very small xmax or power handling they won't get nearly as loud as the others regardless of their equal power sensitivity.
For equal cab volume, lower tuning will reduce sensitivity. If you make the cab larger you can gain back some or all of the lost sensitivity - but you have to make it a lot bigger to gain significant sensitivity.
I'll do a sim so you can look at it. Not an existing design, just a sim with a 30 hz low knee.
I just happened to have this in my files, so this is what you get. If you want to see one that's flat to 30 hz with the high pass filter in place just ask, it only takes 30 seconds to run the sim.
This sim is using the Hornresp System Design - With Driver tool. This is designed as a full size horn and then chopped up into modular cabs. 32 hz horn, 8 cabs using the B&C 18tbw100 driver.
First is a shot of the entire stack of 8 cabs. Power input is a bit high because I cheated. I did 12000 watts in reference to 1 ohm instead of the actual 0.66 ohms presented by the drivers in parallel. Compression ratio is 2.42:1. Each cab will be about the size of a Labhorn, maybe a bit smaller in volume but the same mouth size. Shown with and without a high pass filter - you could run this without a high pass filter if you wanted, at this power level unfiltered excursion peaks at 24 mm which is well within this driver's xlim (and I don't think the suspension will stretch further than about 20 mm anyway), but I've shown it filtered to 14 mm which is this driver's Xvar (xmax is 12 mm but this driver is rated to sound fine to 14 mm).
B&C Speakers
Note that this design is not optimized in any way, I just used the System Design tool to give a horn design for this driver that had a reasonable compression ratio and rear chamber volume. It literally takes 30 seconds to design this way.
Remember, if you want to see a design that's flat to 30 hz with the high pass filter in place just let me know.
Now let's break that down into single cabs and look at single cab performance. This time it's 1500 watts into 8 ohms (instead of 5.3 ohms Re, so I'm cheating again and overpowering by the same amount).
This sim is using the Hornresp System Design - With Driver tool. This is designed as a full size horn and then chopped up into modular cabs. 32 hz horn, 8 cabs using the B&C 18tbw100 driver.
First is a shot of the entire stack of 8 cabs. Power input is a bit high because I cheated. I did 12000 watts in reference to 1 ohm instead of the actual 0.66 ohms presented by the drivers in parallel. Compression ratio is 2.42:1. Each cab will be about the size of a Labhorn, maybe a bit smaller in volume but the same mouth size. Shown with and without a high pass filter - you could run this without a high pass filter if you wanted, at this power level unfiltered excursion peaks at 24 mm which is well within this driver's xlim (and I don't think the suspension will stretch further than about 20 mm anyway), but I've shown it filtered to 14 mm which is this driver's Xvar (xmax is 12 mm but this driver is rated to sound fine to 14 mm).
B&C Speakers
Note that this design is not optimized in any way, I just used the System Design tool to give a horn design for this driver that had a reasonable compression ratio and rear chamber volume. It literally takes 30 seconds to design this way.
Remember, if you want to see a design that's flat to 30 hz with the high pass filter in place just let me know.
An externally hosted image should be here but it was not working when we last tested it.
Now let's break that down into single cabs and look at single cab performance. This time it's 1500 watts into 8 ohms (instead of 5.3 ohms Re, so I'm cheating again and overpowering by the same amount).
An externally hosted image should be here but it was not working when we last tested it.
I'm going to be a bit sacrilegious here and suggest that both the TH18 and Othorn designs probably can benefit from a bit of further optimization. On examining some of the "commercial" TH designs, it seems the designers instead of opting for a flat response at low frequencies and fairly large response aberrations just above the passband instead seem to shoot for an upwardly sloping response from the horn's corner frequency which also tends to minimize the impact of those aberrations.
Here's one more sim just to get you thinking. Once you start shooting for a low knee below 30 hz it starts to make more sense looking for drivers with more excursion potential than your average pro audio driver. And it makes less sense to use the System Design tool, as the design just ends up way too big.
Here's a different kind of design - it uses 4x Sundown ZV3 18s as measured by data-bass. This driver has a claimed xmax of 30 mm. In this sim I've detuned Bl to account for lossy inductance issues.
Data-Bass
Shown at 6000 watts (1500 per driver) in 3648 liters total (912 liters net per cab, about 1.5x larger than Labhorn per cab). Compression ratio is a comfortable 2.39:1. This design doesn't need a high pass filter, as shown it's right at it's rated power handling and 4 mm past xmax. I'm not sure what xlim is for this driver but I'm sure it's fine as shown.
The low knee is lower than anything you've looked at yet, but it's more my style. And with 4 cabs as shown it's probably louder than you need it to be so why not go low?
Playing with the simulator is endless fun. i encourage you to break out on your own instead of always simulating existing sub designs. Pick a benchmark design that has the low knee and max spl you require (make sure you have the appropriate amount of cabs in the sim, high pass filters are used if necessary and you stay reasonably close to xmax and power handling limits) and see if you can beat your benchmark. It's fun.
Here's a different kind of design - it uses 4x Sundown ZV3 18s as measured by data-bass. This driver has a claimed xmax of 30 mm. In this sim I've detuned Bl to account for lossy inductance issues.
Data-Bass
Shown at 6000 watts (1500 per driver) in 3648 liters total (912 liters net per cab, about 1.5x larger than Labhorn per cab). Compression ratio is a comfortable 2.39:1. This design doesn't need a high pass filter, as shown it's right at it's rated power handling and 4 mm past xmax. I'm not sure what xlim is for this driver but I'm sure it's fine as shown.
The low knee is lower than anything you've looked at yet, but it's more my style. And with 4 cabs as shown it's probably louder than you need it to be so why not go low?
Playing with the simulator is endless fun. i encourage you to break out on your own instead of always simulating existing sub designs. Pick a benchmark design that has the low knee and max spl you require (make sure you have the appropriate amount of cabs in the sim, high pass filters are used if necessary and you stay reasonably close to xmax and power handling limits) and see if you can beat your benchmark. It's fun.
An externally hosted image should be here but it was not working when we last tested it.
IIRC, somebody measured a diy TH18 and a commercial Danley TH118 and the results were almost a perfect overlay. Near the end of the TH18 thread maybe? I'm not saying the TH18 is perfect but it's really close to the TH118.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.