That's because the front loaded horns you are looking at have very high tuning. There are tapped horns with very high tuning too, if you look at those you won't need so much excursion potential.
Besides, there are other drivers that have a lot more excursion potential than the ones you are looking at. Alpine is a good brand with healthy xmax and it's available almost everywhere.
I'm not really sure what your goals are. If you state maximum dimensions, and give a full list of drivers that are available that would help. And look a bit further for drivers, I'm sure there are more than 4 drivers that are available in your area. They don't have to be pro drivers to work well.
Almost all the drivers you listed so far are imported, probably all of them. So just because they are not made in your country doesn't mean much. I don't have a price and availability list for drivers in brazil so I can't help much with that, all I can do is make recommendations on what you say you can get and are willing to buy.
Goal:
Low frequency response = Between 30Hz and 40hz no high.
High frequency response = Between 110Hz and 150Hz (if higher is not a problem)
Weight = less then 80kg/cab
Drive amount = 1 per cab
Consider the list below with just 5 driver (all the others didn't shown a good options so lets focus on those) but to reduce your time you can try just 2: KSPA-18-1200 and SUB800
Designs that PrecisionDrive PD186, RCF 18L18P300 and 18Eighteen 18LW2400 works, those two option i have will work too.
Low frequency response = Between 30Hz and 40hz no high.
High frequency response = Between 110Hz and 150Hz (if higher is not a problem)
Weight = less then 80kg/cab
Drive amount = 1 per cab
Consider the list below with just 5 driver (all the others didn't shown a good options so lets focus on those) but to reduce your time you can try just 2: KSPA-18-1200 and SUB800
Designs that PrecisionDrive PD186, RCF 18L18P300 and 18Eighteen 18LW2400 works, those two option i have will work too.
An externally hosted image should be here but it was not working when we last tested it.
I think the only ones we've looked at so far that can meet this goal are the Keystone, TH18 and Labhorn.
The 186 isn't even close to 40 hz, the 1850 isn't either, the WSX is -3 at 38 hz (and that's the manufacturer claim, which probably isn't true) so it's close but no cigar.
So I don't know why you keep talking about these as they don't even meet your first, most important goal.
You did mention a couple of other models, but if your drawings are to scale the only one that has a chance of hitting 40 hz is the MTH.
You might have some troubles getting a smooth response out of tapped horns at frequencies that high.
If you use drivers with neo magnets and make the cab from Arauco plywood (made in Mexico) the cab will be very light, even if it's really big.
Why?
So according to your list of goals and the fact that you only have 8 drivers in your market you are going to have a tough time. You are stuck with the TH18 and the Keystone with your choice of any of the 8 available drivers.
Or like I said you can design your own and look a little harder for drivers. I'm pretty sure there are a LOT more than 8 drivers available in your area, and a lot better drivers too.
Just a guy, we may are running in circles. But I appreciate your help a lot.
You can trust in me, there is no others good drivers. Or in other words, to buy imported one i will pay 4 times more
it's not a good time here.
The SUB800 and the KSPA-18-1200 shown that are able to work well on the keystone and TH18 design, I'd love to build one of those cab.
lets look to my goal in a different way so:
How to tune TH18 and Keystone to reduce diaphragm displacement and keep the FR bandwidth? i don't mind to reduce the efficiency.
You can trust in me, there is no others good drivers. Or in other words, to buy imported one i will pay 4 times more
it's not a good time here.The SUB800 and the KSPA-18-1200 shown that are able to work well on the keystone and TH18 design, I'd love to build one of those cab.
lets look to my goal in a different way so:
How to tune TH18 and Keystone to reduce diaphragm displacement and keep the FR bandwidth? i don't mind to reduce the efficiency.
to refresh the results for TH18 and Keystone with the driver I mentioned.
An externally hosted image should be here but it was not working when we last tested it.
Well you don't want much, do you. If you want the same low knee AND less excursion why not make it bigger? About 220 liters bigger should do. It improves efficiency too.
I used the same 18 sound driver as the sim you showed inputs for earlier, and the same power level. All I changed was a few cross sectional areas and lengths. Shown with your original design with the 18 sound vs the new larger design overlaid.
I keep telling you to play with the simulator, you could answer a lot of these questions yourself. It took me 30 seconds to get this result. I see you asking about making a very small change to the segment before the last bend in the TH18 thread. Little changes like that are not going to do what you want. Big changes like this will do it. This sim isn't optimized. If you sim one of the better drivers in this it should look better, but still not optimized.
All designs are a compromise. If you want more of something you have to give up on something else or add something new to the mix (like and extra 220 liters).
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Yup. THs are basically like bandpass boxes - count on about 2.5 octaves of bandwidth of good performance under the best of circumstances. It does appear to be possible to design one on paper that has a wider FR (see this thread: http://www.diyaudio.com/forums/subwoofers/296731-ooh-looky-th-does-42hz-500hz-3db.html), but that's only on paper - I haven't seen any DIY designs "in the wild" that claim and demonstrate smooth wideband performance, and a look at the phase response above the third major impedance peak in their HornResp sims suggests that it's probably not a good idea to use them that high anyway.
My POC4 design here (The Subwoofer DIY Page v1.1 - Projects : "Proof of Concept #4") comes pretty close to flat wideband performance, but there is that 5dB peak in the upper bass response. Work on that design is basically halted for the moment as the owner has his hands full with other projects
.Checking the time was consumed to develop the TH18 and others designs on this forum with person with high knowledge, I conclude it will take too much time from them I have to do the same. Why i need to reinvent the wheel? I can just choose one using engineering criteria to feet my car, right?
Brian Steele,
Thanks to share you design, the drive you chosen is very close from what I have available but redoing your simulation as you can see below i reach the same problem, the Xmax isn't being enough. Could you check it?
As indicated on the B&C 18TBX100 datasheet below, it has Xmax = 9mm
https://www.google.com.br/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&cad=rja&uact=8&ved=0ahUKEwjK2tS4qtvPAhXCiJAKHRugACAQFggmMAE&url=http%3A%2F%2Fwww.bcspeakers.com%2Fproducts%2Flf-driver%2F18-0%2F8%2F18tbx100.pdf&usg=AFQjCNGMWPrri6Xs6Zf9_AnCjjG4f4sAkA
But applying 1200W make the diaphragm displacement = 13,7mm that is too much.
What we are missing? 
The POC#3 is just a concept or the plan is available?
Other note.: All simulations i've seeing for TH they are using CON and your simulation use PAR
Brian Steele,
Thanks to share you design, the drive you chosen is very close from what I have available but redoing your simulation as you can see below i reach the same problem, the Xmax isn't being enough. Could you check it?
As indicated on the B&C 18TBX100 datasheet below, it has Xmax = 9mm
https://www.google.com.br/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&cad=rja&uact=8&ved=0ahUKEwjK2tS4qtvPAhXCiJAKHRugACAQFggmMAE&url=http%3A%2F%2Fwww.bcspeakers.com%2Fproducts%2Flf-driver%2F18-0%2F8%2F18tbx100.pdf&usg=AFQjCNGMWPrri6Xs6Zf9_AnCjjG4f4sAkA
But applying 1200W make the diaphragm displacement = 13,7mm that is too much.
What we are missing?
The POC#3 is just a concept or the plan is available?
Other note.: All simulations i've seeing for TH they are using CON and your simulation use PAR
An externally hosted image should be here but it was not working when we last tested it.
If you are using flat sheet wood to build the sub and you build it with 2 parallel walls then you need to use PAR. If you build it expanding in all 4 directions then you can use CON or EXP.
All the design you have looked at so far use PAR segments.
It's a good point. Coming up with your own DIY pro audio subwoofer design could involve quite a bit of "debugging" the design, particularly to address those issues that are not predicted by sim software like HornResp. Panel flex issues, for example. Or how the system acts under high input power levels.
To be honest, I won't worry TOO much about that. You're not (well, I hope you're not!) going to be feeding your design sine waves @ 1.2kW, right? In real life, the TBX100's suspension is going to stiffen up as you pass Xmax so the end result is likely to be just a little more distortion at high SPL for frequencies around that excursion peak. And I think that its Xmech is greater than 14mm, so little or no chance of destroying it mechanically if you do manage to hit it with 1.2kW within the TH's passband (outside the passband is another story). If it really becomes an issue, the TBW100 can be substituted and that has a longer Xmax (12mm?) and I'll bet you'll never notice that driver exceeding Xmax with a 1.2kW input.
The POC#4 (not 3) is basically stuck in development at the moment. It's actually a relative of mine that's building it, and he has no time to work on it at the moment. And he went with less bracing than I recommended, the result being more panel flex than I'd like. In any case, I'd like to see it run for a few months in a live environment before I publish the full design, to confirm that both the sound and durability are good enough. The panel dimensions are published on my page however if you want to take a stab at it. Note that they are all cut square - best approach would be to layout the horn as indicated in the cross-section and trim the internal panels to suit.
Using PAR (short for parabolic) expansion is the correct approach for any horn design that involves a layout that has two parallel sides, i.e. almost every horn subwoofer folded up into a box
.You may or may not know this but xmax and power handling are not absolute limits.
It's ok to exceed xmax by a few mm, but your sims should keep within a realistic range, usually within xmax. But as the other guy in the other thread said, if you are running the subs hard there will be power compression, so you will need more power than expected to hit any given excursion limit.
Power handling is a very tricky subject. First, manufacturers rate power handling in different ways so you should read up on how each manufacturer you are considering is rating the power handling, as they may be based on very different methods and mean totally different things.
Furthermore, depending on what type of signal you play, the driver may be able to handle several times it's rated power or it may be only able to handle a fraction of it's rated power.
The type of signal is incredibly important. The frequency, duty cycle and crest factor all play a role in how much your sub is actually consuming over time and different signals will give dramatically different amounts of power over time even if the voltage level on the amp is held constant.
You saw a driver power graph that I posted earlier, it was swinging wildly between 100 and 1900 watts. If your signal contains a note at a certain frequency the driver will only be consuming 100 watts, and yet if you play a different frequency just a few hz away it will be consuming 1900 watts, and yet both notes will be played at the same spl. That's the frequency component of the signal. So if you play a note that's at an impedance peak frequency you can play it very loud all day long and the driver will never heat up because it isn't consuming much power. But other notes could cause big problems in a short time.
Next, the duty cycle. If your signal contains no bass at all, obviously the duty cycle is zero and the sub will never heat up. If the signal contains only infrequent bass drum hits and has lots of time to cool off between hits then you can give the driver several times it's rated power, especially if the bass drum hits are at an impedance peak frequency. But if you are playing a signal that has heavy bass that doesn't stop, that is a 100 percent duty cycle, and you could fry the driver quickly especially if the frequency of the bass is in between impedance peaks. That's the duty cycle component.
And finally you have the crest factor or dynamic range. Uncompressed bass drum hits are very dynamic, they hit hard for a fraction of a second and then get lots of time to cool off before the next hit. This is by definition a high crest factor, very dynamic. But continuous sine waves are the opposite, they don't stop and the dynamic range is very low, only 3 db. So if you are playing classic rock the driver might be able to handle several times it's rated power but if you are playing EDM with continuous bass notes similar to sine waves in dynamic content you may overheat the driver with 1/2 the driver's rated power handling or less.
This is just a reality check to show that xmax and power handling are not absolute values, they can be used to good effect to predict how much power a driver can take and how much spl it will give, but without knowing a lot about the driver design (suspension and Bl limts) and the signal you will be playing (frequency, duty cycle, crest factor) the xmax and power handling limits are somewhat obscure concepts.
I wouldn't be so sure of that. Most people looking into these types of designs are playing bass heavy EDM, stuff like Hunter, which as sine143 has pointed out, is basically the equivalent of a sine wave throughout most of it's duration.
And IIRC, Hunter is hitting hard (like a sine wave) right around 35 hz or so, right in between the impedance peaks for a sub tuned in this range, right where the driver is sucking up full power.
Here's the sim I did from before of the TH18 with the 18 Sound driver. If you play Hunter on this sub at this power level the driver will probably be cooked before the end of the track.
An externally hosted image should be here but it was not working when we last tested it.
Don't forget that the driver's magnet structure is sitting at probably the best place for maximum cooling at low frequencies - right in the mouth of the horn
. My own high volume testing with the 18TBX100 in the POC4 using EDM (admittedly not that Hunter stuff) resulted in the magnet structure being only slightly warm after the testing. In the case of my POC3 design, it has survived several long-duration high power tests with the driver's magnet getting only slightly warm. Hmm... as I still have my POC3, maybe I should try running a high-power test with a pure sine wave at the lower resonance frequency to see what happens...
Now, if it was getting that 1.2kW at its other impedance minimum (around 100Hz), the end result might be a little different
Don't forget that the driver's magnet structure is sitting at probably the best place for maximum cooling at low frequencies - right in the mouth of the horn
I'm not forgetting, that's definitely an advantage. How much that advantage matters with sine waves (or tracks similar to sine waves) playing at impedance minimum frequencies I'm not sure. The cooling from the "mouth wind" is cooling the outside of the magnet. The heat is being generated in the voice coil. There's a lot of steel and magnet in between the coil and the windy mouth. It takes time to dissipate heat by external cooling (blowing on the motor). So unless the driver is very well vented I'm not sure how much the breeze will affect things in the short term.
I don't doubt it. You are using a good driver and from random track selection you probably won't hit a brutal track like Hunter (that's basically a sine wave at an impedance minimum frequency) very often, if at all.
OP is not willing to buy B&C, so probably will not end up with a state of the art driver.
Be careful - by the time the magnet gets warm to the touch you could have fried the voice coil. It's probably unlikely in a driver as well designed as yours, but still be careful as it could be a costly experiment.
You both guys are completely right.
I din't go so deep regarding the signal because during the discussion about the TH design no one talked about it, so they were just using the T/S parameter for that, and looks like TH are more driven by BL and Qts. This is the reason i got surprised with the Xmax results. Once all my simulations are aligned I'm relaxed now and can back to plan to build the cab
I din't go so deep regarding the signal because during the discussion about the TH design no one talked about it, so they were just using the T/S parameter for that, and looks like TH are more driven by BL and Qts. This is the reason i got surprised with the Xmax results. Once all my simulations are aligned I'm relaxed now and can back to plan to build the cab
The POC3 actually uses a Dayton PA310 driver, which is pretty cheap - certainly not in the B&C TBX bracket. I'll likely perform the test at different power levels and monitor SPL output. When the change starts to get noticeably non-linear, that would mean that the coil is starting to heat up noticeably.
Particle velocity at the mouth peaks just above where the impedance minimum occurs - an almost perfect relationship
Attachments
The driver I'm willing to buy are very well aligned with B&C 18TBX100 and with 18sound 18w1000.
The problem is to align with B&C 18SW115, B&C 21WS152 and Void V18-1000.
Now is the time to to decide between Keystone and TH18, there are a lot of pros and cons for both sides but this is the good part once the driver was solved.
The problem is to align with B&C 18SW115, B&C 21WS152 and Void V18-1000.
Now is the time to to decide between Keystone and TH18, there are a lot of pros and cons for both sides but this is the good part once the driver was solved.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.