New sub design? Constricted Transflex, simple build (series tuned 6th order)

[sabaspeed521]

To provide further explanation: ..... This message should finish providing the rest of the information regarding how HR can be used in an unusual fashion to sim boxes like these , in case anyone want to try this simulation......


TB46 ,
When using a throat chamber and port in this way you have to be sure to switch to "throat adapter" mode in HR and also be sure to use TH mode ... Basically there should be no centimeter spacing between S1 and S2 ... Using offset here just causes a loss in valuable path length (but not net volume) (Hornresponse's "TH" mode and "OD" mode differ in this aspect because in OD mode the throat chamber's port attaches to S1 and adding centimeters to the length of S1 can create a flare shape because S1 attempts to adapt the cm sq area of the end of the port with the sq cm area of S1) ....Sorry for any confusion there .... Ok , moving along.... In TH mode the end of Lpt/Apt connects to S2 ..... of course after S2 the standard horn/pipe segments form the latter portion of our path and S5 is our exit/mouth/terminus or whatever you choose to call it ...

Throat chamber is defined by Vtc/Atc

As HR sees things the driver has no offset meaning it is simmed as being plugged right into the end of the (in our case tube shaped) chamber with the driver facing the path (instead of firing into the side of the path at some user defined distance from the closed end) ....... Ironically in the real-world built version there is a slight offset but it's effect on response will be minimal ...

So yes , thats basically how that ML-Transflex box is arranged in Hornresponse , i know it is highly unorthodox , but it works ...


in version 2.1 I eliminated the port (Ap1 and Lpt) and used s3 as the constriction instead ....

In version 2.2 im keeping the big throat chamber, bringing back the port, losing the flares, and setting up the S1 through S4 segments as a straight fat pipe shape again (S5 can be deactivated in this case and S4 becomes the new mouth to simplify things) ....

I'd love to see a Alpine sws12 version of this . Sim first and then picture would be swell.

 

[Matthew Morgan J]

OH DAMN ! TECH UPDATE ! I just realized that the report i made above about the exit of the throat chamber's port behaving differently in relation to how it interacts with S1 in various modes should have been referred to as a BUG REPORT and not a normal function of Hornresponse .....
The interaction between Apt/Lpt and S1 is a little bit different in all modes , OD and TH are actually very similar, it is CH and ND modes that have something going awry....I think it has something to do with the throat chamber's port , it shows a strange flaring (in Wizard's "schematic" display) that shouldn't be there in both CH and ND modes ... I can open up Loudspeaker Wizard to make a change (or not) then check my FB by looking for the sharp dip in displacement , then click on "save", the Wizard's window closes, then i hit calculate and then check "diaphragm displacement" in the main HR window and i can see that the FB has shifted a full 5hz ....... I think i have a set of inputs that are totally strange for sure, but this shift shouldn't be happening...

Does someone out there want to help me confirm a bug in HR?

David Mcbean? are you out there man? I love your software dearly, and i just wanted you to know that i really appreciate you, and i have a bug for you.

 

[Matthew Morgan J]

I'd love to see a Alpine sws12 version of this . Sim first and then picture would be swell.

You betcha , i am pretty excited about that SWS-12 too , Its a dirt cheap driver which is widely available and not only that but according to sims it looks like it is closely competitive with that Special run LAB15 ! ! ..... Close enough at least , in the right alignments , but im going to tinker with HR some more to make sure i can confirm this .....

 

[sabaspeed521]

Down

 

[Matthew Morgan J]

down for a bug check? you arent afraid of bugs? got your can of RAID ready?

 

[Matthew Morgan J]

Ok , i was just goofing around and trying different combinations of random pipeS and ports in different modes when i spotted this.... Check for FB via the displacement graphs in both the wizard and main interface , let me know if you see the discrepancy .... I am using HR version 34.80

dont forget to add the Atc/Vtc/Lpt/Ap1

 

[Matthew Morgan J]

Ok , just confirmed something .... I just removed the throat chamber port ... So now it is just a throat chamber feeding into the other pipe segments with no constriction yet i am still getting the discrepancy ... So even though the display of the port is weird (when used) the FB discrepancy is still there either way ...

At least i know that it is not all of my squeezing and pinching of the path that is causing the prediction error ,and since this issue is much less pronounced in OD and TH mode this shouldn't have affected our designs

Deviation is only a few hertz in OD mode and i don't see any deviation at all in TH mode .....

 

[Matthew Morgan J]

Ahhh , here we go , after removing the big throat chamber i dont see any wizard/main deviation errors in any of the modes ... So i would have to think that it is an issue with the throat chamber ...

 

[Matthew Morgan J]

In both CH mode and ND mode i made the throat chamber even longer and shortened the other segments, Wizard/main deviation error is now almost 10hz !! ack ! thats bad .....

However, changing the aspect ratio of the throat chamber to a more boxy shape reduces the error but it is still there even at an Atc of 2000 sq cm ..... FB on main is still a few hz lower than in Wizard..

 

[David McBean]

The interaction between Apt/Lpt and S1 is a little bit different in all modes, OD and TH are actually very similar, it is CH and ND modes that have something going awry....I think it has something to do with the throat chamber's port, it shows a strange flaring (in Wizard's "schematic" display) that shouldn't be there in both CH and ND modes.

Hi MMJ,

For Nd and CH configurations, Ap1 and Lpt specify a conical throat adaptor between the throat chamber and the horn throat S1. For OD and TH configurations, Ap1 and Lpt specify a cylindrical port opening between the throat chamber and the side entry point into the horn at S2.

Kind regards,

David

 

[David McBean]

Check for FB via the displacement graphs in both the wizard and main interface , let me know if you see the discrepancy.

Hi MMJ,

I don't see any discrepancy.

Kind regards,

David

 

[Vix]

Hi guys,

I have read 45 pages of this thread in one day, and this sub is worth trying. But, even though bigger is better, I was thinking about the smaller one: say you have some smallish bookshelf speakers that run nice down to 80-100 hz (3-4" drivers). Now say you want to build simple, cheap and good sounding subs (two). Would it be possible to build this sub with some good 8" drivers (or even 6.5")?
The whole point is to build a combo that would over perform a Bose-like sat/bandpass-sub system.

 

[xrk971]

Hi guys,

I have read 45 pages of this thread in one day, and this sub is worth trying. But, even though bigger is better, I was thinking about the smaller one: say you have some smallish bookshelf speakers that run nice down to 80-100 hz (3-4" drivers). Now say you want to build simple, cheap and good sounding subs (two). Would it be possible to build this sub with some good 8" drivers (or even 6.5")?
The whole point is to build a combo that would over perform a Bose-like sat/bandpass-sub system.

Yes, build a mini Karlsonator. It is the same alignment but meant for fullrange speakers. A 0.4x scale mini K will give 80Hz bass extension.

http://www.diyaudio.com/forums/full-range/239338-mini-karlsonator-0-53x-dual-tc9fds.html

 

[Vix]

Hi xrk!

Sorry maybe I was not clear. Mini Karlsonator is fine for the top (80 hz and up) but now say you want to build a pair of subs for them, like the ones described in this thread. Using 10" drivers or more would probably be too much, that's why I thought for 8" or even 6.5" ones.
Basically, I am asking about a scaled-down design of the subs described here.

Thanks,

 

[Matthew Morgan J]

Hi MMJ,

For Nd and CH configurations, Ap1 and Lpt specify a conical throat adaptor between the throat chamber and the horn throat S1. For OD and TH configurations, Ap1 and Lpt specify a cylindrical port opening between the throat chamber and the side entry point into the horn at S2.

Kind regards,

David

Thank you David ,
I now do see what you are saying ... ND and CH can allow Lpt/Apt to be flared , tapered or straight based upon the realationship between the sq cm area of Apt and S1 ... and it does show faithfully in the schematic .... The different ways in which Lpt/Apt connect to either S1 or S2 also makes sense now, thanks for confirming and clarifying that .. ..

I still have that FB discrepancy error though when comparing wizard to main ..... It seems to only be noticeable with large throat chambers with pipe-like aspect ratios, most noticeable when the Atc is closer to the driver's SD in the particular sim i am working with ...
I had just tried HR on a different computer to see if i can replicate it on other machines with different specs and a different CPU , using the same version of HR and i didn't get the discrepancy at all , no error (with all of the same inputs) , but then something caught my attention when comparing the sims on the two different machines:

The install that suffers from the discrepancy error would always pop up a window when i clicked on the calculate button asking "DO YOU WANT TO MASK THROAT CHAMBER RESONANCES?" , and the other install that doesn't show the deviations wouldn't ask about that and didn't show any such pop up ......... So(when displaying inputs) i went to "tools" at the top left of HR then "options" and then i chose the "PROMPT BEFORE CALCULATION" under throat chamber and rear chamber resonances ("not masked" was where it was previously set) ...

Now i can see the discrepancy error on both installs ! I will post some screenshots of what i am seeing ...

 

[sabaspeed521]

Hi xrk!

Sorry maybe I was not clear. Mini Karlsonator is fine for the top (80 hz and up) but now say you want to build a pair of subs for them, like the ones described in this thread. Using 10" drivers or more would probably be too much, that's why I thought for 8" or even 6.5" ones.
Basically, I am asking about a scaled-down design of the subs described here.

Thanks,

It's possible, but may be hard to find a workable substitute for the swe10s4 if trying the original design, I've only found 1 substitute and it was the sws12d4 (and by association the sws10d4). This if for the constricted design. For a straight pipe design a wider array of drivers should work but it is a little taller for the same tune.

 

[Matthew Morgan J]

Hi MMJ,

I don't see any discrepancy.

Kind regards,

David

Ok , i just figured out what is happening here .... You only see the error when you choose "prompt before calculation" under tools / options , and then click on the calculate button and choose "no" when you see the popup.... When using that "prompt" setting you will have the deviation between main/wizard because one masks resonances and one doesn't .... If you use the other options under "tools" like "masked" or "not masked" then the error is gone!

It seems like the wizard is masking resonances by default without asking (only when in prompt mode), but of course when hitting calculate on the main interface it gives you the pop up window , and if you choose "no" then thats when we run into the problem ..... One result (wizard) is masked , and the other (main) is not ......

I suppose i will just switch to the "not masked" option under tools from now on to avoid this ...

 

[Matthew Morgan J]

Hi xrk!

Sorry maybe I was not clear. Mini Karlsonator is fine for the top (80 hz and up) but now say you want to build a pair of subs for them, like the ones described in this thread. Using 10" drivers or more would probably be too much, that's why I thought for 8" or even 6.5" ones.
Basically, I am asking about a scaled-down design of the subs described here.

Thanks,

Vix , what sort of extension are you looking for? this sort of simple folding + constriction cannot be made too short unfortunately if we want to maintain the quarter wave action ... It might be possible to find an 8 but the box would end up being sort of tall & skinny ... The Karlsonator fold (two folds basically) gets more path length out of the same cabinet height but since it has a generally expanding path it will ideally want driver with different parameters than what we have been working with here , and with the expansion the FB would actually end up being not too terribly different than what we get from a single fold constricted box at the same height .... Expansions require longer paths to maintain the same FB when compared to straight paths with no expansion ... Tapers (meaning gets smaller towards open end) allows the path to be shortened as well , similar to constriction (ports/vents/mass loading/mouth shading) schemes ..

Saba is correct about how difficult it is to try to find an alternative for the 10S4 , this is especially the case with the 40hz tuned small ML-transflex .... I have found a few drivers that almost come close but they are either not quite as flat in response or fall short on SPL or xmax , or are too expensive......

Although , if you want an 8 with a higher tune that would certainly change things ....

I could even try a double fold + constriction with no expansion to get a box significantly shorter in stature with added depth, but the simple way that i came up with to accomplish that introduces a significant offset on the driver placement , which will require a driver with a different set of parameters to avoid the downward tilt in response that would probably occur... Drivers that would normally have a rising response tilt (meaning they show an upward tilted response without any offset but everything else being the same in sim) would be called for which would TYPICALLY (but not always) exclude car audio subs .... I think there is an affordable Sony 8 sold at ApexJr that might work in such a thing , but it is 8ohm , and it doesn't list XMAX in the specs so it is hard to say what the max SPL would be ... Maybe one of the Mavin drivers or an 8" by Goldwood or GRS could be made to work but i would have to look at the options and try them in sim ...

As far as SPL is concerned: if you are talking about a miniature shelf system for use in a small room then the xmax and power handling requirements wont be too extreme really ...

 

[xrk971]

Hi xrk!

Sorry maybe I was not clear. Mini Karlsonator is fine for the top (80 hz and up) but now say you want to build a pair of subs for them, like the ones described in this thread. Using 10" drivers or more would probably be too much, that's why I thought for 8" or even 6.5" ones.
Basically, I am asking about a scaled-down design of the subs described here.

Thanks,

I will see what I can come up with. Maybe dual 6.5in drivers ? Is 40Hz low enough for you?

 

[tb46]

Mainly for xrk971. ;-)

Hi xrk971,

I noticed your Post #436: "I modeled the sketch in 426 as drawn...". So I modified your original AkAbak script from Post #167 to fit w/ a drawing I made of Matthew's enclosure from Post #426. I just enjoy looking through your AkAbak work, I'm nowhere near that level. Anyway, it's not a bad box, and using AkAbak it could possibly be improved. I changed the area of the constriction corner (Duct 'D45') a little, but maintained the 200cm^2 cross-sectional area. (To use the script change the .txt extension into .aks.)

Now "they" are telling me that I have to go to some grandchild's birthday party, so bye for now.

Regards,