Hi chaps,
I was wondering if anyone knows of or has used any software that simulates the effect of port resistance on the frequency response of a given system.
We all know the frequency response of a ported sub changes with power input due to port friction becoming significant at higher air velocities, but is there any way of simulating this?
I'm hoping to better simulate the comparison between tapped horns (and similar) with ported boxes - Hornresp seems to think the ported box will win, but having read around on here, I suspect the port problem might swing it the other way.
TIA,
Chris
I was wondering if anyone knows of or has used any software that simulates the effect of port resistance on the frequency response of a given system.
We all know the frequency response of a ported sub changes with power input due to port friction becoming significant at higher air velocities, but is there any way of simulating this?
I'm hoping to better simulate the comparison between tapped horns (and similar) with ported boxes - Hornresp seems to think the ported box will win, but having read around on here, I suspect the port problem might swing it the other way.
TIA,
Chris
They all provide this information in the form of port velocity, given in meters per second. You typically want the port velocity less than 14m/s.
Soundeasy's box simulator will do nonlinear driver simulations, as will LEAP. You could also roll your own in MathCAD, Mathematica or Matlab/Simulink if you are good with Math. The trouble is really in measuring the nonlinear parameters. To actually model port chuffing would take a full-on CFD package.
A high power impedance sweep would show a reduced port peak in a vented box and if you want to do port compression simulations with cheap programs, you could change Qp until it matched the high power sweep. The problem is that to get realistic driver parameters to enter you would have to measure T/S at higher powers as well (can of worms) or develop some model of how T/S changes with increased power (compliance changes a lot, as does Re - this affects Q and Fs, etc...)
In the end it would be WAY easier to just build both boxes and compare.
I wouldn't recommend Soundeasy, though. It is basically a collection of moderately connected tools by someone who knows a lot about calculating audio related things, but doesn't seem to have any intention of putting a finish or polish on it, or have it make any sense. He does write up some interesting findings in his manuals, and If I recall correctly from before I bought soundeasy there is a PDF on port compression at the Bodzio software site.
A high power impedance sweep would show a reduced port peak in a vented box and if you want to do port compression simulations with cheap programs, you could change Qp until it matched the high power sweep. The problem is that to get realistic driver parameters to enter you would have to measure T/S at higher powers as well (can of worms) or develop some model of how T/S changes with increased power (compliance changes a lot, as does Re - this affects Q and Fs, etc...)
In the end it would be WAY easier to just build both boxes and compare.
I wouldn't recommend Soundeasy, though. It is basically a collection of moderately connected tools by someone who knows a lot about calculating audio related things, but doesn't seem to have any intention of putting a finish or polish on it, or have it make any sense. He does write up some interesting findings in his manuals, and If I recall correctly from before I bought soundeasy there is a PDF on port compression at the Bodzio software site.
Last edited:
Possibly, but decent plywood isn't cheap over here. Since the cabinets will be quite large (for 15" drivers), that cost could get prohibitive.
So far, I have two cabinet designs. Running the driver close to Xmax, they have identical output at the low knee of the response, and the ported box wins by a few dB above.
The T-TQWT-thing I've simulated (Hornresp parameters below) has an outlet velocity peaking at 14m/s.
The ported box nudges 18.5m/s.
S1=600
S2=1300
S3=600
S4=600
S5=600
L12=20
L23=60
L34=68
L45=12
TH=1
Ported box is 95L, port area of 500cm^2, port length 70cm.
Driver is Eminence Kilomax 15, and each cabinet is being driven to ~12mm peak excursion, which I know is pretty much the limit of what these drivers will do. Ported box needs 70v to do this, other cabinet needs 63v.
I'll start a new thread on the specific designs - don't want to stray too far OT.
Thanks, all.
Chris
With a small box (95l) you will want to stick with a reflex design. What kind of extension are you trying to get out of this project.
Around about 40Hz. Having put a load of tracks through Audacity, that'll cover most music. The few tracks that'd be missed need mid-low 30s, which isn't going to happen with this driver in this cabinet size.
The ported cabinet comes out to ~130L overall - the port's not small. The other cabinet is a similar volume.
Chris
130L is still considered a compact box. If you don't need upper bass efficiency of a horn I would go ahead with the reflex design, if size and weight are not an issue the lower distortion and extra output from a horn are the way to go.
Have another look at the Hornresp parameters - this isn't a conventional horn by any stretch of the imagination. S1 is small to give a diagonal at one edge of the cabinet, perfect for wheels. The size of S1 made very little difference to the frequency response, so I figured wheels would be a nice touch. I'll attach a Sketchup image later to illustrate the fold I've got figured out.
The ported box has more upper bass output, the whatever-cabinet-this-is simulates almost ruler-flat from 45-110Hz. -3dB at 40Hz. The ported one has similar output at 40-50Hz, but considerably more output higher up, but is taking ~100w more power. Each could probably be eq'd to the other with similar power requirements, but I'd like to know how compromised the ported box will be at high power <60Hz.
Both size and weight are an issue: I'm a 10.5 stone one-man crew. I can man-handle (and stack, just) 2x15" two-way cabinets, though those cabs are of no interest to me. The primary use of my system is live music at pubs and function rooms, or recorded music playback for parties etc.
Tops are a pair of reasonably nice 12"+CD cabinets, fed from a PV2600 for 500w/ch.
The sub (currently a Tannoy T40 with Kilomax driver = huge 55Hz peak) is powered off a bridged QSC USA850.
I also have a DCX2496 for processing.
Chris
The ported box has more upper bass output, the whatever-cabinet-this-is simulates almost ruler-flat from 45-110Hz. -3dB at 40Hz. The ported one has similar output at 40-50Hz, but considerably more output higher up, but is taking ~100w more power. Each could probably be eq'd to the other with similar power requirements, but I'd like to know how compromised the ported box will be at high power <60Hz.
Both size and weight are an issue: I'm a 10.5 stone one-man crew. I can man-handle (and stack, just) 2x15" two-way cabinets, though those cabs are of no interest to me. The primary use of my system is live music at pubs and function rooms, or recorded music playback for parties etc.
Tops are a pair of reasonably nice 12"+CD cabinets, fed from a PV2600 for 500w/ch.
The sub (currently a Tannoy T40 with Kilomax driver = huge 55Hz peak) is powered off a bridged QSC USA850.
I also have a DCX2496 for processing.
Chris
18m/s is not all that high a velocity. You aren't gong to have major chuffing problems that you can hear over the masking of the rest of the frequency range. If you want insurance, just flare the entrance and exit of the port, which is a good idea anyway and will reduce compression.
As far as building test boxes goes, you could use particle board or even OSB for a test box, no need to use spendy plywood. I stand behind my statement that building test boxes will be an easier and surer answer than trying to develop a nonlinear simulation.
Whenever you ask questions like this, it is good to have context, what is the driver, what is the intended use, what are the designs you are comparing, etc....
Last edited:
Chris,
The Eminence Kilomax 15 has only 7.9 mm Xmax, simulations indicating excursion above that will not be very representative of the actual response, and distortion will increase quite rapidly above Xmax.
~12mm (24mm peak to peak) is probably more excursion than you will see regardless of power input in the real world, and unless you are quite tolerent of gross distortion, you probably won't want to use power levels that exceed Xmax by more than a few mm.
As an example, I recently tested some Dayton PA385S-8 15" drivers rated for 10mm Xmax in open air (no enclosure at all), worst case scenario for excursion. Using 76 volts, the maximum measured excursion was around 17.5mm (one way), while Hornresp predicted 26.97mm at 22 Hz and 46.46mm at 4 Hz.
At 54.6 volts, 16 Hz, Hornresp predicted 23.59mm one way excursion, the actual measurement was only 11.5mm.
Full results of that test here:
post # 4298
http://www.diyaudio.com/forums/subwoofers/119854-hornresp-430.html#post3877117
In actual tests measuring the B&C18SW115-4 used in a TH compared to BR (with the same low corner), at high power (relative to low power) the BR lost a little LF, while the TH lost some upper response.
I'd expect similar results with your "T-TQWT-thing" compared to BR.
Art
Hi Art, thanks for weighing in.
I tested the Kilomax driver free-air, and got around 12mm one-way travel when really pushing it. Its a fair point, though, that as excursion increases past Xmax, the suspension will start to put the brakes on, so I'll potentially see droops in the response where excursion's getting large.
I suppose, then, its not even fair to treat the simulations as peak SPL output as I'd previously thought.
I'll have a look over your findings over the next few days - never thought I'd actually be following in your footsteps.
Its 2.30am here, so I'm going to call it a night.
Chris
I tested the Kilomax driver free-air, and got around 12mm one-way travel when really pushing it. Its a fair point, though, that as excursion increases past Xmax, the suspension will start to put the brakes on, so I'll potentially see droops in the response where excursion's getting large.
I suppose, then, its not even fair to treat the simulations as peak SPL output as I'd previously thought.
I'll have a look over your findings over the next few days - never thought I'd actually be following in your footsteps.
Its 2.30am here, so I'm going to call it a night.
Chris
- Status
- Not open for further replies.