Question, what do you have to change when dual loading these? Besides the obvious thing which is a removable panel of some sort because one will be completely out of reach.
Unrelated but on the note of dual loading: how much power can you get from a 15 amp outlet (to the driver), assume class D amplifier like maybe a inuke or newer crowns.
1,000 watts per cab is nice and all but I don't know if you can actually get 2000 watts from a standard outlet, assuming your amp has that much current capacity.
What's nice about the SWS is it is dual 4 ohm, meaning it can be wired (dual cabs) either 4 ohms or 8 ohms (ignoring 1 ohm because of the obvious reasons). 8 ohms 1000 watts is hard to come by though so 4 ohms much more reasonable.
Unrelated but on the note of dual loading: how much power can you get from a 15 amp outlet (to the driver), assume class D amplifier like maybe a inuke or newer crowns.
1,000 watts per cab is nice and all but I don't know if you can actually get 2000 watts from a standard outlet, assuming your amp has that much current capacity.
What's nice about the SWS is it is dual 4 ohm, meaning it can be wired (dual cabs) either 4 ohms or 8 ohms (ignoring 1 ohm because of the obvious reasons). 8 ohms 1000 watts is hard to come by though so 4 ohms much more reasonable.
Last edited:
Just wondering how do you mount the two drivers in both these cases, and what does the one on the left look like (I've cad'ed out the straight pipe as I've shown in an image but the other I'm not sure what it looks like and/or the build complexity).
Thanks a bunch, those are some crazy good sims and they are better than mine on the compact end. If it's not too much I'd be interested in what a single and dual 12 sws 12d4 would look like size wise and output wise (in the same style of the one's you did).
Here are the sws 12 specs
15 mm xmax 500 watts rms
479 1.34 E-04 157.48 7
20.19 3.7 4.84
That's the same format as hornresp in the correct units.
Your welcome and no problem ...These sims are quick and easy , i should be able to crank out a few more for you and then i will get back tinkering and updating the ML cabs ..
The drivers in those traditional pipe sims i made for you are "spanned" along the line meaning they would be placed closely one above the other with S2 and S4 being the midpoint between drivers ... This works well enough with smaller drivers but larger drivers being spanned in this way require the S2 & S4 points to be moved further into the pipe which can take away from performance in some cases, so in that scenario it is better to make two smaller cabs (one for each driver) instead of trying to put both into one cab ...
Build complexity is still about as simple as it gets except for the fact that you may have to create an access panel for the upper driver in the dual driver cabs since that upper woof will be somewhat concealed..
Here are some more juicy details with a standardized pack space of 20 cubic feet (about half my 4runner) ~ 1 20 inch wide T60 . Cost estimate rough of course, haven't cut sheet anything, could use more than a gallon--not sure yet.
***Ignore the dual driver loaded one's after adjusting the l12 l45 to what they actually are it's no longer viable. single loaded cabs still are though.
. Cost estimate rough of course, haven't cut sheet anything, could use more than a gallon--not sure yet.***Ignore the dual driver loaded one's after adjusting the l12 l45 to what they actually are it's no longer viable. single loaded cabs still are though.
Attachments
Last edited:
That's what I was looking to find out -> drivers side by side vertically, that's what I had presumed, need to update sims to move L12 and L34 up a bit to accomodate the dual driver version.
Last edited:
I will attempt to answer this question with the best of my knowledge but im sure others will have their own experiences and perspectives to share..
You should be able to get quite a bit from a standard 15a outlet especially when using the modern Class D amps with switching power supplies .... The amp sections themselves should be at least 80% efficient (often higher) and the SMPS section is probably even more efficient than 80% , so conservatively you can estimate at least 60% of the circuit's power will be available for your subs continuously, but the fact that music is generally so dynamic works in our favor allowing momentary peaks to exceed the continuous rating for a fraction of a second when trying to reproduce something transient like a kick drum...
Ideally, even though these new amps are very efficient you will want multiple dedicated circuits ran to a DJ booth or amp rack if you plan on running some subs with large amps ... Even though some of these amps have a PFC feature in their supplies there is still no guarantee that everything will behave properly when there is voltage sagging taking place and to trip a breaker is certainly embarrassing and inconvenient..
Last edited:
Clearly but this is the real world and the real world says it has to fit on one breaker :-/. Kidding of course. If I ever get into a predicament like this I would take outlets that are on different breakers and run extension cords as needed. 1500 watts per breaker means 4 subs at most per 2 breakers assuming headroom.
Are you using Akabak now? How do you know the particle velocity is lower in the constriction in the new design?
When Saba originally said he didn't hear chuffing you took that as proof positive there were no issues, now that he hears turbulence at moderate volume it's time to build and measure before moving forward. Measurements are necessary, that much is clear, this has been my position from the beginning.
Yes, of course ports are commonly used and they can be effective if sized properly. Constrictions like this example on the other hand are rarely if ever used, with the exception of Brian's tapped horn where he used it in an existing design to fix a problem.
First of all, I analyzed Weltersys's ports and his were moderately undersized. Your Alpine box constriction was grossly undersized. There's a difference here, and your original Lab design was quite a bit more undersized.
Second, the sine sweep info posted here did show SIGNIFICANT compression when 20 and 160 hz were compared. http://www.diyaudio.com/forums/subw...using-dayton-pa385-8-drivers.html#post3898842
Over 2 db of port compression even after power compression is subtracted, as I pointed out in post 264.
Third, as I pointed out Weltersys's Shoehorn sub is a bad example to study port compression issues. Here's a picture of a more common ported box showing only port output with the Shoehorn port output in light grey.
An externally hosted image should be here but it was not working when we last tested it.
A more conventional ported box will have a much more pronounced port output near Fb and much less contribution high in freqeuncy. In the shoehorn the level at 100 is only 6 db down from 40 hz, in the other design there's over 20 db of difference between port output at tuning and at 100 hz. THIS is the kind of design you need to look at to study port compression issues.
There are a few reasons why the Shoehorn exhibits only minor compression, you can't use one very unique example to say that compression isn't an issue.
I also showed there's no reason to use a constriction as a design feature in a new design, you can get the same response at the same spl over the same bandwidth without using a constriction.
Your position seems to be that the software is wrong (even though you don't use the appropriate software), sometimes in some designs port compression just does not occur for some unknown reason and the official understanding of how port compression works is in question. I'm going to have to continue to disagree.
Last edited:
I'm fairly sure he hasn't posted his "new design" and as such its hardly fair to judge what may or may not be the port velocity seeing as I haven't even seen the new lab 15 design he's working on. We'll revisit the port topic the next time I consider building a ported or otherwise majorly constricted sub. Until then can we let it rest? We've both heard your words of wisdom.
For my current "pending" design I'm looking at 15 m/s in hornresp (seeing as the mouth is one big chamber of equal cross sectional volume I find it safe to assume this velocity as being accurate). How can I find out if there will be any "port compression" for this 7.5x13 (cross section internal) inch rectangular folded prism pipe shape.
For grins I simmed out a 6 foot tall version of the 15 inch loaded alpine sub that is 132 dB average corner loaded 1m output full blast which is exactly it's displacement limit . Usable to 87 Hz and only costing 2 sheets of ply, a 110$ driver, and the essentials (2-3tubes PL, mounting hardware). Think of it as a simpler to build THTLP (BFM land) in output
. Usable to 87 Hz and only costing 2 sheets of ply, a 110$ driver, and the essentials (2-3tubes PL, mounting hardware). Think of it as a simpler to build THTLP (BFM land) in output
No Sir, not Akabak, not yet anyway ... I was given a tip (by someone here on DIYaudio) for use with Hornresponse in order to calculate based on worst case scenario, it just requires a little math ... It can be done by comparing the sq cm of the mouth area to the constriction area, figure out the ratio and multiply accordingly .... I checked the results against what you and others were saying the m/s rate should be at the squeezed point of the path and the figures matched up...
The reason i say it is "worst case scenario" is because in a classic quarter wave resonator the velocity is highest at the open end of the pipe (mouth) yet the constricted point in these designs is not at the end of the path (especially so with newer revisions of these boxes where the controversial squeeze has been moved further up inside of the path)
Yes , indeed ! I agree , I am really looking forward to seeing more measurements now that the additional leaks are now sealed ...
I just noticed the other day that the "Karlsonator" has an abrupt pinch in the path about 2/3 way down the path (from the closed end) and i did read much of the discussion about that cab but i don't recall seeing anyone complain about any chuffing or compression problems with that design ... It is possible that there isn't any such problems with that design despite the constriction ....
Brian has never said anything about chuffing or noticeable compression in his design either, although im not sure how much testing he did for the latter ...
I find it really interesting that his DFD vent filled in the midbass dip so incredibly well, this phenomenon should be studied further ...
If i recall correctly i think he used a 3:1 ratio squeeze which is what i am approaching now with the new revisions of the ML-Transflex cabs ...
Anyway , my point being is that it is not so unheard of because other people have tried similar things and have had no complaints.....
Im sure you know that real world results are worth paying close attention to because there are factors that come into play that may not have been taken into account with all of our simulations and theory ...
Not sure which Alpine box you are referring to but "grossly" you say? Thats a bit much ... The early 40hz version was actually not all that bad, but like i said it is something i have improved upon .... When it comes to the Lab15 (special) version I have to agree that the first version definitely had an undersized port and i appreciate you pointing that out to me, it has been improved upon now as well ...
20hz?!? Really?!?! Ok come on now JAG , now you are just grasping or trolling with this point that you are trying to make here .... To be fair that box has an FB of 39hz and it is obviously vented so you would never want to use it as low as 20hz, it's output down there would be insignificant and with no loading the cone would just flap around ... Your argument is invalid , and i mean that in a nice way, really
.... I do enjoy a good debate but lets keep it reasonable ..Why? After all not every box that employs a squeeze/vent/port will be conventional .... Thats the really cool thing about DIY, we can design and build some boxes that are highly unconventional and may very well lead to innovation if all goes well ...
BINGO! In other words you are saying: "I would rather not use this example because it is unorthodox and ports/vents do not behave in a classic textbook manner when used with unusual alignments" ......
So then you have to come to terms with the possibility that ports/vents may not suffer from as much compression or chuffing in certain alignments (which would also imply they could potentially suffer even more in some other alignments) .....
In the Weltersys example those ports/vents didn't appear to have any issues, in your own words: "only minor compression" so from that i would have to draw the conclusion that those ports were NOT undersized for his alignment after all, at least not at that power level .....
Could it be that those same ports are undersized for some alignments , marginal for some alignments and fully acceptable for others? ...
It is really not unreasonable to suggest this is a possibility when so many other qualities or parameters vary from design to design, it shouldn't be a surprise that port behavior may vary too when used in different designs.. .
I will try to keep an eye out for evidence of this ..
Sheesh man ... Is that really your impression of me?
When some real world results don't line up with sim of course i am going to ask questions, or make statements, does that mean that i am an ABSOLUTE disbeliever in your software in particular? No , not at all ... ...... Is it really a software problem? I don't know.... I posted the suggestion out to our discussion to see if anyone could offer up anything in response, real-world evidence for or against etc (not just theory).... HOWEVER, i did do a fun experiment by creating a simple test box with the same driver in various different softwares with the same ports and got numerous different velocity results so what i can safely say is: Not every software can be right if they disagree .... They obviously use different models to calculate their particle velocity results , is it horribly unreasonable of me to suggest that they may not all be accurate in every situation ??? Oh, but your software is right?
.... Every time ... In every design imaginable ........ Maybe it is, and maybe it isn't , im no expert but i am glad that you have so much faith and i am sorry if i have offended your sensibilities ...Again , what i do know is that your software produced higher figures (as does HR with the right calculations) so i may go with that as a guideline, like i mentioned earlier: a "worst case scenario"...
I do NOT believe that "sometimes in some designs port compression just does not occur for some unknown reason" , BUT i am open the idea that there could be some variance between designs if i see more evidence suggesting it ...
............................................................................
Last edited:
Now that you mention Karlsonator, I do think that GregB's excellent design is indeed a ML Transflex 6th order bandpass alignment. It is used to extend bass of a fullrange driver rather than as a sub, but looking at its topology, it is exactly the same with the benefit of Karlson aperture on the exit. I have a model of the Karlsonator in AkAbak and never bothered to check the velocities in the constriction because we never have an issue with reports of chuffing. I do know that there are some extreme velocities at the small end of the K aperture - the air jets out of that on a kick drum like you wouldn't believe. But, since it has such a small characteristic length, the Reynolds number is low and the flow is still probably laminar. Never noticed any chuffing or jet noise on the front K aperture but can feel it. It would blow a candle out 4 inches away. I will go back into Karlsonator script and check velocities at constrictor. Main point is that the this alignment is not new but has been done before, and quite successfully for bass augmentation in a fullrange. The mini Karlsonator based on this design is one of the most successful speakers I have posted based on feedback from the many builds to date.
http://www.diyaudio.com/forums/full-range/239338-mini-karlsonator-0-53x-dual-tc9fds.html
In a similar way, I think this sub woofer thread will see many happy builds too. One thing you may notice with this alignment is that it is very flexible. With the proper scaling of the overall volume (read, depth and width of channel) and constrictor, you can get almost any driver (within) reason to have a flat response. It is very Qts tolerant - doesn't have to have a low Qts driver, and in fact works better with moderate Qts drivers as most MLTL's do. The higher Qts allows an fb to be set below the fs. Much tougher to do with a low Qts driver.
Having said this, it might be interesting to see how scaling the Karlsonator model to match the drivers mentioned here fare. This was the subject of the MOAK sub thread here:
http://www.diyaudio.com/forums/subwoofers/254948-moak-mother-all-karlsonators-aka-magnificent-monster.html
Last edited:
And I’m sure from his and other pioneer’s papers that he was concerned about all design trade-offs that affected an audio system’s total distortion, which would include keeping a BLH’s audible distortion as low as practical. Back then, the signal chain was the weakest link in accurate reproduction and now thanks to advancing technology it’s down to just the speaker system.
Don’t recall ever seeing any specific research comparing stepped Vs tapered expansions other than numerous measured responses of different horns of similar LF cutoff showing it, not to mention the fact that it’s enough for it to be clearly heard.
The closest research to it I’ve seen posted has been the impedance mismatching of a simple folded pipe/horn Vs one with the appropriate corner reflectors that Bob Brines did early in his learning curve, which clearly showed the HF roll off with no measurable difference down in the speaker’s desired pass-band. Unfortunately, I can’t find it on his site now, but I imagine he’ll send it to you if you ask.
Again, what compression ratio Olson used is irrelevant beyond the fact that it was [and still is by many] considered to be one of the best performing BLHs ever built. Again, what’s relevant is the modern math I used based on T/S is so well proven as to be a given.
The software only proves that it’s a constriction doing exactly what it was meant to do! Once you make it big enough to not be a true constriction, then one must come up with another way [or a driver with different specs] to accomplish it.
Note too that neither HR nor AkAbak appear to take any acoustical damping into consideration, so it seems reasonable to me that since they show exaggerated peaks/dips in the response that is either much muted in reality or flat not there strongly implies that any particle velocity, etc., calculations are just as ‘distorted’ and only good for comparison’s sake.
GM
I went over the math for this a few pages back. It has to do with calculating the Reynolds number in the channel based on characteristic length and velocity. A number under 2300 means laminar and a number >100,000 is highly turbulent. 10's of thousands is mildly turbulent. Mildly means large scale vortices (cm's to meters in scale) and lazzy flapping flows. High turbulence means homogenized tiny little swirls under 1mm in scale. The Re number and the smoothness factor of the walls also affect when turbulence onset happens. The 2300 number assumes smooth walls. A rough wall can trip this at lower Re number.There are tables that show correlation of wall roughness vs Re and where onset of turbulence is. This has been extensively studied in fluid mechanics for a long time - nothing new here. Some more background on Re number and turbulence can be easily found with a search. Here is an example:
https://www.princeton.edu/~asmits/Bicycle_web/transition.html
https://www.princeton.edu/~asmits/Bicycle_web/transition.html
For woofers, ~4-5 mm, others mostly < 2 mm with comparably low electrical power handing, hence partly the need for very efficient drivers.
Nope, still irrelevant to this discussion! If a 40 mm Xmax driver alignment required a constriction IN a pipe/horn, it would still boil down to choosing one based on its T/S specs and if its compression ratio is too high at higher power, then either it will need to be power limited and/or a different driver or alignment will be required.
OK, so? I’ve been a HE [prosound] ‘hornie’ since age 8, so you’re ‘preaching to the choir’ with me, but I also know that one only need design/build to the needs of the app and the smaller, less efficient it becomes, the more trade-offs required. If MMJ/whoever encountered any ‘chuffing’, then that’s what adding damping is for and if it caused a bit of compression it would more than likely be mostly thermal power distortion before the constriction could affect it much, hence my ‘much ado about nothing’ attitude.
Yeah, when the TH threads started, I 'campaigned like a Southern Diplomat’ for a separate prosound forum because I knew that discussions like this would cause way too much confusion for ‘newbies’ and casual DIYers in a subject already quite complex.
Yes, he did and ideally MMJ’s design isn’t the most acoustically efficient, an inverse tapered TQWT driving the rear chamber is for the reason you state, same as it’s the superior alignment Vs a BR for the drivers in question, hence my repeated caveat, but it doesn’t negate his design for the reason you claim.
There’s lots of ‘absurdities’ on audio forums, not the least of which is folk’s spouting absolutes of questionable accuracy and/or what they should be using/building based on their personal preferences rather responding to the OP’s Qs and considering the needs of their apps.
In retrospect, I guess I should have right off suggested MMJ scrap his design for an inverse tapered TQWT driven 6th order as I wanted to and short circuited the vast majority of his thread, which in a different way was what you were trying to do, but that’s not what these forums are for AFAIK.
GM
In Brian's case it was a novel and inventive way to fix a problem. I don't think it's a good idea to use the concept as a design feature in a new design.
The design Saba built had 30 m/s velocity (26 m/s with hpf in place) and Flare It said the safety zone for that port size was 9 m/s. Weltersys's number were 31 m/s with an 18 m/s safety limit. That's a big difference.
20hz?!? Really?!?! Ok come on now JAG , now you are just grasping or trolling with this point that you are trying to make here .... To be fair that box has an FB of 39hz and it is obviously vented so you would never want to use it as low as 20hz, it's output down there would be insignificant and with no loading the cone would just flap around ... Your argument is invalid , and i mean that in a nice way, really
BINGO! In other words you are saying: "I would rather not use this example because it is unorthodox and ports/vents do not behave in a classic textbook manner when used with unusual alignments" ......
I explained twice now why Weltersys's example is not a good one to study for port compression. The reason is not that it doesn't compress, his tuning is high and his port output is very blunt and broadband so it's very hard see the compression effects. You HAVE TO go outside the passband a bit to see that there is actually significant compression happening because inside the passband it's happening everywhere, not just a Fb. The theory and/or physics of the compression phenomenon does not break down in the Shoehorn, it's just hard to pick out the effect because of the unusually flattened and broadband output of the port.
Yes, I've said several times now that his ports were only marginally undersized, that's why there's only a bit less than 2 db of compression inside the passband.
At one point or another you did actually say each of those things and you seem to maintain that there are real world examples that don't adhere to the laws of physics. Small ducts introduce losses, there's no way around that.
I didn't notice any audible chuffing for as long as I had POC2 in operation, and I have used it at high levels from time to time. I still have the box around so I could do some more extensive tests, but it that might require a bit of work to get done.
The DFD was created totally by empirical means. Its effect was confirmed by an Akabak model afterwards however. A search on the forum for the thread in which it was discussed should turn it up.
I still don't know why it worked in POC2. An attempt to duplicate the process on my TP build failed miserably.
The only absolute I spouted was the need for measurements. This will be the third time I'm saying my analysis could be wrong but until proven wrong I'm sticking to my story because I'm just telling you what the software says.
The only guy that has built one is reporting turbulent noises at moderate volume now, so unless that's considered acceptable it looks like I have been considering the needs of their apps all along.
I've posted a lot here because several people seem to be trying to argue that compression isn't an issue and a couple have suggested it is not well understood. I haven't told anyone what they should be building, and my personal preferences are much larger horns. The only thing I've done here is suggested there is an avoidable issue and measurements should be done.