| fjmi |
hi!
what's a really good software for simulating these subwoofers?
i tried allready winISD but i want to compare the results with proffessional software.
thanks very mush
fjmi |
|
|
| mike.e |
consider 'subsim'
winisd IS accurate. youl end up paying money for 'real' programs which do essentially the same thing but have slight bonus features. |
|
|
| Ron E |
| versuch 'mal BPBoxmodel vom Robert Bullock. |
|
|
| fjmi |
i thougt winISD is accurate, but to proof it, i tried the older betaversion vs the new alpha und the results were very different.
half of the volume und much more efficency -> i think understanable why i do not trust winISD temporary ;)
-> therefore iam searching for a proffessional programm to proof it.
i will post the resulsts of bpboxmodel und subsim. |
|
|
| Ron E |
| quote: | Originally posted by fjmi
i will post the resulsts of bpboxmodel und subsim. |
WinISD is questionable in many areas, you are correct to question it.
In BPBoxmodel it is instructive to switch between real and idealized port behavior ;) You will then wonder if it is possible to make a 6th order bandpass without prototyping it first..... |
|
|
| Rademakers |
| quote: | | WinISD is questionable in many areas |
What area's exactly?
Mvg Johan |
|
|
| Ron E |
Last I tried it the excursion, port velocity and group delay plots were wrong on the pro alpha product.
The Beta product vented box and bandpass group delay was way wrong and there were issues with the "port mach" calculations and errors with port sizing. There were other little errors as well that I have noticed in sims people have made public.
I wrote my own software (very much like Unibox but before unibox) long ago so I don't play with these freeware simulators much anymore - Even the best of them seem to have a cumbersome GUI that I have no patience for.
BTW Bullock's Boxmodel products indicate RMS (not peak) excursion and velocity, FWIW. Although they are DOS, they are the only reliable benchmark (except for the Tranmission Line model) available for free, IMO. |
|
|
| Rademakers |
| quote: | | Last I tried it the excursion, port velocity and group delay plots were wrong on the pro alpha product. |
Personally I found the excursion to give a pretty good idea what power my designs can handle, considering it's simulating with a 3 dB sinewave that is.
Port velocity has been helpfull too, in finding the right portsize so far. I use WinISD Pro a6.
I didn't do any measurments on it, just a lot of testing. Tho I found simulair results with the transfer magnitude on other (paid) programs. I know it's not the most accurate software around but considering the price I found it to be pretty good so far
Can you tell more about the tests or measurments you did? Not that I'm doubting your results, I just find it interesting because this is the first (scientifically based) negative story I heard about WinISD Pro, while I know many satisfied users.
I'm actually kinda promoting it's use by offering an extended free guide (on the Help-file) for beginners :o Luckily it's usefull for other programs as well. :xeye:
But the truth matters more than my ego for that matter ;)
Mvg Johan |
|
|
| fjmi |
and what should i do if i do not really want to build a prototype that would be 23000 times modified?
i need a special designed subwoofer for cinema-application AND for a candidacy at a speacial school -
just the design is necesarry, but i want a good thing, too, the design exists allredy in my mind (and it is fine, maybe i will post photos if it is build sometime.....untill april)
-------
the rearchamber seems to bee like a normal bassreflexbox.
even in winISD die simulation ist nearly the same -> i wanted to experiment with the frontchamber, specifically with the voulme and lengh/profile of the tube.
and if my english is to baaaaaaaaaad :) please correct it!! |
|
|
| fjmi |
hi! i have a problem with this programm (under winxp) :bawling:
if someone would please be so kind an try it for me :rolleyes:
it should look about the orange curve
http://img136.exs.cx/img136/5072/spl9uu.png
but (!) the SPL should be changeable to a normal 'bandpasscurve' by making the tube longer -> instead of 70Hz 50hz tuning freq.
some corner marks i want to realize:
*volume not to big (70l ist about the maximum)
* no passive radiator
* 2 different frequenze - passages (frequenzgänge ;) ) by longer tube
*spl adapted to normalroom - contortion (about + 10db at 20Hz, ... and -*db at 70-90Hz -changes on postion of sub and dimensions of the room ... simulated, too :whazzat: )
and other infos (simulated with winisd)
frontchambervoume=20l
rear= 55l
tuningfreq.rear=27Hz
tuning.f. front= 70 / 50Hz
tubedimension: 7cm*10cm * ??
by the way whats the maximum airspeed in this tubes that would make NOT a noise? (m/s)
and here are the infos of loudspeaker, i think for the price a fine one :)
thanks and greets from austria :)
fjmi
p.s. for interested people (german-attainments would be good)
here the 'original'(origin,source,birth,..??) thread :
klickme |
|
|
| mike.e |
The wobbley response looks bad.
| quote: | | by the way whats the maximum airspeed in this tubes that would make NOT a noise? (m/s) |
keep below ~20m/s apparently.
My subwooferi s below 12m/s |
|
|
| simon5 |
Some people accept 34m/s (10% speed of sound) at full power since you're not running it at full power all the time...
Higher end products aim for 17m/s (5% speed of sound) or lower... |
|
|
| fjmi |
below 20m/s ????
that means i maybe can use a tube dimension of 3*2cm?!? (=16m/s) -> 6cm² vs. a SD of ~360cm² !!
if i understand correctly 'wobbley response' is the same like group delay (?), it get's down to 23ms at 26Hz if i use a 30l case :)
but then i do not know how it stands about the spl.
--> what should i do in your opinion? :bawling:
(e.g. what software should i use, for bpbox i had to change my windows or anything else) |
|
|
| fjmi |
hi folks!
the programm allready runs :) (was the false path of my dos 'window')
and the results seem to be better than these i awaits :smash:
but i have no idea what i should enter in
rloss (qas,qls,qps) and in fbox
currently (standart)valuas are:
qas=28
qls=14
qps=28
fbox=14
what is my front'chamber' pvent or mvent? (i chose pvent)
and whats up with parallel/cascade (both options bring good results)
and WBS/WBF
i want to simulate a bad (the badest) possible spl, not a good if ... and that and .... is optimal
---
current system:
35l/16l
27/65 (45) Hz
i guess that the delay is okay(below 30ms), but can i see some values somewhere?
the spl has at ~25hz a little knoll, but i want to damp both chambers a littel -> is that a good idea?
what should be the max. lengh of the tubes? mine are under 50cm but air velocity could be higher (just under 5%)
and last but not least, does it change the lengh if i put a little bit of foam in the tubes?
i think it would slow down the efficency, but that would be no problem.
thanks for you answers and your efforts! :)
fjmi
p.s. whats to best time to write with you?
the post of simon5 was made at - Today, at 04:17 AM (dont know if this time is the same at your location) |
|
|
| simon5 |
This woofer seems to be really hard to tune in a 6th order bandpass box.
There is no alternative for your project? Something like Peerless drivers or something?
I guess your first idea was quite good...
I think I would go with:
55 liters tuned to 27 Hz rear and 20 liters tuned to 55 Hz front.
But then you'll have to make some compromises about vent noise because port length is long in a small box. |
|
|
| fjmi |
why should i use peerless drivers?
it is possible but i found this ravemaster driver and it looked good.
if this bandbass simulation software is to some extent correct, there is no noise even at 200W inputpower. the tubelengh is ok, and under 50cm.
the only thing i do not like at the 55l casing was the bad response of ~35ms
in 35l it is quite good with about 20ms. |
|
|
| simon5 |
50 cm long in a small box like that is not that easy to do.
Peerless XLS line have 12.5 mm of excursion versus 5.4 mm for this one, that means more than twice the excursion.
If you really want to be linear down to 20 Hz, you'll need more than this driver.
Also, you can't use more than 70W RMS, or you will overdrive this driver. |
|
|
| fjmi |
the 50cm tubes are not a problem in this case.
maybee peerless are better but tehy cost not 55€, much more if i remember correctly and the fs is mostly at ~30Hz vs 25
i do not want a linear frequency-stride down to 20Hz, quite the converse, i do not want it ;)
first i thougt the same, more than 80W are not possible because of the 5,4 mm, but in a smaller case there seem to be 200W possible (down to ~23Hz)
p.s. this subwoofer is for normal home-cinema and some 'partymusic'
and i payed my attention to the room gain (10-20Hz up to +20db!!! depends on the listener position) |
|
|
| mike.e |
Make the transfer function,have a nice flat response. Dont worry about GD theres nothing you can do to improve it once your chosen F3 and box alignment is chosen.
Ports 1metre long I would start worrying about Tline resonances. Note that tiny box + high power + low tune = really really long ports.
Try making a flat response to ~35hz atleast,you dont want to exceed driver excursion at modest SPL
Personally I enjoy SPL of 110-120dB SPL on peaks. |
|
|
| fjmi |
| the rear chamber has ~45cm long tubes, 'airspeed' is ~7% @ 200W -- 180 is the max. because of drive excursion, group delay below 30ms, i can't see any big problems :xeye: |
|
|
|
