Are you out of your mind? Cant you scale picture? You got everything black and white. Picture of size 2487 x 1620 and you cant scale length?
I give you a hint as I see its not obvious to you... scale along x and along y is the same.
Width is 622mm(S=a*b
)Ofc the surfaces 3000cm2 and 8000cm2 are in cm2 not cm3 my mistake.
I have to quote myself:
Are you out of your mind? Cant you scale picture? You got everything black and white. Picture of size 2487 x 1620 and you cant scale length?
I give you a hint as I see its not obvious to you... scale along x and along y is the same.
Width is 622mm(S=a*b
Ofc the surfaces 3000cm2 and 8000cm2 are in cm2 not cm3 my mistake.
I could scale the picture if I wanted to count pixels but I don't really want to. And the burden of proof is on you. And it's WAY easier for you to post a picture with proper dimensions than it is for me to count pixels.
Your fold is not accurate to the Hornresp model and it's not efficient.
I already addressed this in post 39. Here's some quotes for you.
"Looking at your inputs, you've used a pure EXP segment for the bulk of your horn flare. That type of pure EXP segment flares out way too fast at the mouth which is going to make it nearly impossible to fold your design up in an efficient way."
"It would be quite difficult to fold this flare shape into an efficient square or rectangle shape. Usually the rear chamber helps to make things fit because it's shape is flexible (whatever shape you want) but in this case you 30 liter rear chamber is very small, about the size of the red square on the graph."
"... it is very clear that the rear chamber is WAY to big in relation to the last segment of the flare.
You specifically said earlier that you could increase the rear chamber size and it wouldn't affect the design much. Maybe that's what you did in this fold job. Regardless, increasing the rear chamber size does have an effect AND adding volume to the rear chamber just to make the box a rectangular shape is not an efficient fold as it obviously makes the entire enclosure larger.
This is what I said in the first place - it's nearly impossible to efficiently fold this type of flare into a rectangle.
Now notice that every one of those quotes says it would be hard or impossible to fold your design in an EFFICIENT way.
What you have done to make your design fit into a rectangular shape is increase the size of the rear chamber. A quick estimate suggests it's about 80 liters in your fold pic but your design called for a rear chamber size of 30 liters. You made your rear chamber almost 3x larger just so you could fit your flare into a rectangle.
THAT IS NOT EFFICIENT FOLDING. THIS IS WHAT I'VE SAID WAS GOING TO HAPPEN SEVERAL TIMES NOW.
You've also got another dead area in the bottom right corner that's another 15 or 20 liters.
So in total you've added somewhere around 70 liters or so to the design JUST TO MAKE IT FIT INTO A RECTANGULAR SHAPE.
This is extremely inefficient, and worse, that space could be used to good effect if the horn had been planned properly from the beginning.
So let's look at the effect of making the rear chamber 80 liters. Light grey trace is the original design, bold black track is overlay of the design with rear chamber increased to 80 liters. Showing frequency response and excursion with same power input.
An externally hosted image should be here but it was not working when we last tested it.
So we can see from the frequency response that you've raised the low knee from the original 50 hz up to around 65 hz.
And from the excursion graph we see that you are using an extra mm inside the passband and an extra 10 - 15 mm outside the passband (outside doesn't matter as much if you use a high pass filter).
The original design was decent but increasing the rear chamber size to 80 liters completely wrecks it.
So in order to fit your design into a rectangle you've done an incredibly inefficient fold resulting in terrible performance and a bad design.
Also, this obviously makes the whole enclosure larger too, making it heavier and larger and generally less easy to work with. In other words less efficient.
Congratulations on proving my point AGAIN.
And I'm still not sure you've even folded the flare correctly. You butchered the attempt at folding the rear chamber so I would not be surprised if the rest of it was equally incompetently laid out.
It's time for you to stop. You've already ruined a very nice thread. I can continue to point out your errors - in fact everything you've said so far is wrong - but I'd rather have you just stop posting here.
So, I kinda tried to scale it, curious as I am, simulated it with a BC 21SW115-4...
105dB sensitivity @ 2V, 54Hz F3, reasonably efficiënt, good cone control and so on!
Exactly the charactaristics of a fullsized horn as JAG described, right? So what are you on about?
You haven't designed a horn the same size and with the same Fc, that has charactaristics of a fullsized horn. You designed a larger horn with a way higher Fc. Ofcourse it will behave like a fullsized horn...
EDIT: about the folding, with all the parabolic segments it's hardly a true exponential horn, especially considering the last part... However, it's reasonably close
105dB sensitivity @ 2V, 54Hz F3, reasonably efficiënt, good cone control and so on!
Exactly the charactaristics of a fullsized horn as JAG described, right? So what are you on about?
You haven't designed a horn the same size and with the same Fc, that has charactaristics of a fullsized horn. You designed a larger horn with a way higher Fc. Ofcourse it will behave like a fullsized horn...
EDIT: about the folding, with all the parabolic segments it's hardly a true exponential horn, especially considering the last part... However, it's reasonably close
Last edited:
It appears to be designed for 18nw100 (at least that's what it says in the Hornresp inputs in post 17 and 19), not that it matters, it's a terrible design as folded. The original (before folding) was ok (for it's size and low knee, but not compared to a full size horn), but like I said, it can't be efficiently folded.
EDIT: about the folding, with all the parabolic segments it's hardly a true exponential horn, especially considering the last part... However, it's reasonably close
True, but this we can let slide, it's the least of the problems with this mess. My designs will all come in around T=0.5, sometimes 0.6 but never as high as 1 (pure EXP) as this guy has used because you can't fold it. For people doing designs like the concrete horn thread guy it doesn't matter because there's no intent to fold it but for regular sub use ALL these little designs need to be foldable, otherwise they are just paper tigers and have no relevance to the real world.
When I use a hyp/ex flare like my usual 0.5T it can't be perfectly folded with flat sheet goods either, but at least you can get a lot closer with a 0.5T flare than with pure EXP. And as always the more segments you use the closer you can get. But he used very few segments in his fold.
I would be more worried if the flare was just not at all similar to the Hornresp model, which is probably the case anyway, as he doesn't want to provide any data except for a picture of a fold with no dimensions shown, and as I've already pointed out it's very clearly not accurate, it's so bad you can tell it's wrong just by looking at it.
Last edited:
You were promising your fans how I made wrong fold, so its up to you to prove it. Who knows, maybe if you shout loud enough you will shout me down.
I already did prove it's folded wrong, you can tell just by looking at it the rear chamber is WAY to big. I'm just curious to see how badly the rest of it is folded but it doesn't really matter. I told you before you attempted to fold it that it couldn't be done efficiently and then you proceeded to post a fold pic that clearly showed I was correct but you claimed success anyway.
I've expressed my opinion that I would like you to quit. I've also told you that I can continue to point out your errors if you choose to continue. It won't be hard since you have no clue what you are talking about and everything you have said so far is incorrect but it does waste my time so it is irritating. But go ahead, please continue. I'll just continue proving that what you are saying is wrong. I'm not sure why you would want to continue though, it would be a lot easier to take your lumps, go cry in a corner for awhile if necessary, study hard for two or three years and then come back and try a bit harder to fit in - aka drop the hubristic and insane idea that you can show up in a fairly advanced forum and teach the members how things work. You are WAY out of your element here and it would appear that you are the only one that doesn't realize it.
Hi Y'all,
Maybe genOme can find a better place for his fold and discussion in the Klipsch Community thread about the University Classic?
https://community.klipsch.com/index...sicdean-horn-profile-need-driver-choice-help/
May just be my old eyes, but his simulation and fold look like a modified version of the Classic.
It doesn't look like the discussion since Post #17 has been very productive. But hey - 2017 - It's gonna be a great year. Happy New Year.
Regards,
Maybe genOme can find a better place for his fold and discussion in the Klipsch Community thread about the University Classic?
https://community.klipsch.com/index...sicdean-horn-profile-need-driver-choice-help/
May just be my old eyes, but his simulation and fold look like a modified version of the Classic.
It doesn't look like the discussion since Post #17 has been very productive. But hey - 2017 - It's gonna be a great year. Happy New Year.
Regards,
You guys gonna hate me. But actually this annoying linguist starts to make me laugh.
BTW nice topic about reverse engineering.
How can it have the same suspension as 9601?? It has 4" VC.
Maybe you just don't know what is exponential horn. So you try to avoid it this way? Btw even kids at kindergarten can scale length from picture.
BTW nice topic about reverse engineering.
How can it have the same suspension as 9601?? It has 4" VC.
Last edited:
Kindergarten kids don't have to scale length from pictures if the person that provides the picture has the common courtesy to include dimensions on the picture. You've already drawn it in some type of CAD software, it would take a few seconds at most to add the measurements to the picture. Why would I try to scale it just to prove it's wrong when I already know it's wrong?
But it isn't required. It's already clear just by looking at it that you messed it up terribly.
I know what exponential is and i avoid it because you can't fold it efficiently unless the rear chamber is very large. I prefer lower T from the Salmon family of hyperbolic/exponential flares. There's no real reason to use an exponential flare unless you are trying to get the most possible bandwidth. And in your case your example horn doesn't have particularly wide bandwidth. Even if it did, once you simulated properly as the offset driver horn that it is, the reflection off the closed end and the huge resulting notch in response would limit your high knee to 3 or 400 hz.
Do you have anything new to add or are you just going to keep telling me to scale the dimensions from the picture? I see you've given up trying to prove undersized horns behave like full size horns which is nice. If you don't have anything new to add I think you can drop it now.
Just a couple of words about xmax ...
I would be vary wary of comparing xmax between products from different manufacturers. Sometimes they are comparable, sometimes not even close.
For 18 Sound xmax is calculated like this (as documented on at least some of their driver spec sheets) -
(9)Linear Mat. Xmax is calculated as; (Hvc-Hg)/2 + Hg/4 where Hvc is the coil depth and Hg is gap depth.
Note that this is a slightly more complicated formula than the simple (Hvc-Hg)/2 that you usually see. And the + Hg/4 part of the formula will give a higher number than the simple formula that does not include the + Hg/4 part of the formula.
B&C is a bit different, they say this about xmax -
There's a few interesting things to note in this quote.
1. They say xmax is typically defined as the 10 percent distortion point as per AES2-1984 BUT B&C ignores this standard completely.
2. Instead they spec xmax based on linear mathematical xmax, presumably using some formula based on coil length and gap depth although I could not find the formula they use in a quick search.
3. They spec Xvar based on Bl or CMS dropping to 50 percent of the at rest value, which is going to equate to distortion a lot higher than 10 percent, in fact a lot higher than 20 percent (in Klippel testing Bl = 70 percent is generally equivalent to 20 percent distortion and is the standard for subwoofer testing), and as far as I can tell no one else in the industry uses any similar method to spec usable excursion.
4. B&C states that measuring distortion can have ambiguous results which I suppose can be true, but I think it's better than the ambiguity inherent in calculating a linear mathematical xmax based on the driver physical measurements.
So I'm not sure why they spec xmax and xvar the way they do, since linear mathematical xmax can be ambiguous, Xvar is a very high distortion "usable excursion" figure, and there is the AES2-1984 standard which B&C totally ignores which is not only readily available but also fast becoming the new standard.
So what in the bloody hell is going on with xmax specifications?
I did find this following quoted post on a different forum, not sure how accurate it is, but it paints a picture that I've been talking about for awhile - all companies want the highest xmax figures possible so they are figuring out subtle ways to cheat. The original formula (Hvc-Hg)/2, calculated the point at which the entire gap is no longer full of coil. If you exceed this distance the coil starts to increasingly leave the gap. There's some pics here to explain it for those that don't know.
http://www.speakerwizard.co.uk/driver-ts-parameters-xmax/
When you start adding to the original formula to get a higher xmax number, the specified number is going to be past the gap and theoretically into some very high distortion numbers. (But not necessarily, as linear mathematical xmax can be ambiguous and may not accurately describe distortion performance.)
It would seem that different companies are just making up their own linear mathematical xmax formulas. They take the standard (Hvc-Hg)/2 formula and add a few mm based on an extra part added to the formula. And there seems to be no standard, everybody is just doing their own thing.
So what formula is B&C using? Let's find out.
For the 18sw115 the gap depth is 14 mm and the coil winding height is 34 mm. They claim linear mathematical xmax is 14 mm.
(Hvc-Hg)/2 = (34-14)/2 = 10 mm - so they are not using the standard formula
(Hvc-Hg)/2+Hg/3 = (34-14)/2+(14/3) = 14.7 mm, so they are not using this version of the forumula
(Hvc-Hg)/2+Hg/4 = (34-14)/2+(14/4) = 13.5 mm, so they are not using this version of the formula unless they are rounding 13.5 mm up to 14 mm
I would have to do the math on a bunch of different B&C drivers to figure out what formula they are actually using to spec xmax, but from this example alone it seems like they MIGHT be using something close to (Hvc-Hg)/2+Hg/4, which is what 18 Sound is using.
So B&C and 18 Sound linear mathematical xmax might be somewhat comparable. But linear mathematical xmax gives somewhat ambiguous results that may or may not coincide with actual distortion testing, why not just use AES2-1984?
Then we've got the new crop of super drivers with 30+ mm claimed xmax, they have their own completely different standard. Since it's super easy to get a nice Bl curve and VERY hard to get nice Le or Cms curves at 30+ mm excursion, these companies base xmax on the Bl curve alone and completely ignore everything else. Sometimes it's not even a measured Bl curve because they don't actually bother to get the driver tested, just a simulated Bl curve from the driver design software.
We desperately need some standard. I would have thought maybe AES2-1984 would be it but clearly manufacturers are resisting it. This makes no sense to me, I've seen Klippels of B&C 18tbw100 and they look fantastic. Why not just use Klippel results instead of the ambiguous linear mathematical xmax and the made up xvar? Why is 18 Sound using a pimped up version of the standard (Hvc-Hg)/2?
Last edited:
18 Sound and RCF use the same one. Faital Pro uses the most generous I've seen. Beyma is between those. B&C looks like they might be the same as Beyma.
Eminence uses klippel but the VoiceCoil test came up short of claimed xmax on a Kapplite 3015LF.
BMS are straight shooters and seem to match up very closely in independent tests.
So yes a crapshoot all around. The best bet is independent testing from a 3rd party. From subjective comments on the web it does appear that Faital Pro drivers lock up quicker than comparable drivers from other brands. Databass has good comments for BMS, B&C, and 18 Sound.
Eminence uses klippel but the VoiceCoil test came up short of claimed xmax on a Kapplite 3015LF.
BMS are straight shooters and seem to match up very closely in independent tests.
So yes a crapshoot all around. The best bet is independent testing from a 3rd party. From subjective comments on the web it does appear that Faital Pro drivers lock up quicker than comparable drivers from other brands. Databass has good comments for BMS, B&C, and 18 Sound.
B&C uses the same formula as Beyma.
BMS 18n862
(50-12)/2+(12/3.5)= 22.43
What's interesting about B&C's xvar value is that it's actually less than xmax in some of their drivers. The suspension is ignored in geometric calculations.
Klippel seems to be the best overall bet but will be at the mercy of manufacturing consistency of speaker soft parts.
BMS 18n862
(50-12)/2+(12/3.5)= 22.43
What's interesting about B&C's xvar value is that it's actually less than xmax in some of their drivers. The suspension is ignored in geometric calculations.
Klippel seems to be the best overall bet but will be at the mercy of manufacturing consistency of speaker soft parts.
Last edited:
You don't make that decision on xmax. It's all about the motor (lower qt) and increased thermal powerhandling in the 115 series. I do in fact use the sw100 in numerous designs, and the 115/9601 in less because I focus on front loaded horns. It's the same reason the 21sw152 is the winner in my book over the 21nlw9601, more thermal chops.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.