Member
Joined 2003
I think the confusion is that Rs is now located in filter tab, where it used to be Rg up at the source signal, so it's not clear that Rs is an always active value regardless of other filter selections. Perhaps some separation of the two, return of Rg to the top of the window when filters are not used, and next to Ls is DCR of the inductor, so when Ls is checked, total series resistance is Rg + DCR of Ls.
OK, understood so let's turn it like that : to MY confusion, I thought that Rs was the resistive part induced by the wire to the self. I understand now that it is not and completely independant from the check box !
Happy New Year!
Great work on the VituixCAD program. I have successfully made a few simple open baffle calculations with it and it is a wonderful design tool.
I recently found the Auxiliary / Box Volume Tab in the program and was pleased to see it graphed box Resonance frequencies. So I tried entering my favorite test case, a shallow half cube shaped box with the woofer centered on the front to see if all the resonance peaks lined up together at the same frequencies. Unfortunately they did not. Instead it showed the first resonance frequency for the side walls, top and bottom at half the frequency the should be. This is a common problem that I first noticed in some obsolete box software when I actually built a cube shaped box and measured the modes frequencies. For a cube with a driver centered on one side the resonance frequencies for the front to back mode is half that of the side to side and top to bottom mode. To get the highest possible first resonance in a box you build a half cube box and place the woofer centered on one of the square sides. It's all about the distance from the cone to each wall and which box modes the driver excites.
I know, I know, 90% of you think I'm full of it. As a noob I may not have configured the calculation as I intended, but I can't find my error. To verify my assertion, you can build your own half cube shaped box, or a cube shaped box, mount a speaker centered on one of the square sides and measure the resonance frequencies yourself. This is in my opinion the optimal geometry for a speaker box as it give the highest possible low frequency mode frequency for the box. So it is often possible to set the crossover lowpass frequency below this lowest box resonance frequency and have a resonance free pass band. Also damping material works better at higher frequencies so moving those mode frequencies up, makes it possible to damp them. Of course a half sphere rear enclosure is slightly better, but some what more difficult to build.
For quick comparison I entered the same geometry into the easily accessible Hornresp program. It can be configured to calculate the sealed box response without a horn. It calculates the correct mode frequencies for the half cube geometry box. See the plot below. Hornresp doesn't plot the modes directly, but they are very apparent in the Acoustical Power plot and can also be seen in the impedance plot.
Figure 1. Auxiliary / Box volume window for VituixCAD showing modes for 12.5 liter half cube box.
Figure 2. Hornresp calculation showing modes at the correct frequencies for 12.5 liter half cube box.
As you can see above, there is no mode at 600 Hz in the Hornresp software plot.
To replicate the plot in Hornresp you simply enter the values for a sealed rear chamber as I did below.
Great work on the VituixCAD program. I have successfully made a few simple open baffle calculations with it and it is a wonderful design tool.
I recently found the Auxiliary / Box Volume Tab in the program and was pleased to see it graphed box Resonance frequencies. So I tried entering my favorite test case, a shallow half cube shaped box with the woofer centered on the front to see if all the resonance peaks lined up together at the same frequencies. Unfortunately they did not. Instead it showed the first resonance frequency for the side walls, top and bottom at half the frequency the should be. This is a common problem that I first noticed in some obsolete box software when I actually built a cube shaped box and measured the modes frequencies. For a cube with a driver centered on one side the resonance frequencies for the front to back mode is half that of the side to side and top to bottom mode. To get the highest possible first resonance in a box you build a half cube box and place the woofer centered on one of the square sides. It's all about the distance from the cone to each wall and which box modes the driver excites.
I know, I know, 90% of you think I'm full of it. As a noob I may not have configured the calculation as I intended, but I can't find my error. To verify my assertion, you can build your own half cube shaped box, or a cube shaped box, mount a speaker centered on one of the square sides and measure the resonance frequencies yourself. This is in my opinion the optimal geometry for a speaker box as it give the highest possible low frequency mode frequency for the box. So it is often possible to set the crossover lowpass frequency below this lowest box resonance frequency and have a resonance free pass band. Also damping material works better at higher frequencies so moving those mode frequencies up, makes it possible to damp them. Of course a half sphere rear enclosure is slightly better, but some what more difficult to build.
For quick comparison I entered the same geometry into the easily accessible Hornresp program. It can be configured to calculate the sealed box response without a horn. It calculates the correct mode frequencies for the half cube geometry box. See the plot below. Hornresp doesn't plot the modes directly, but they are very apparent in the Acoustical Power plot and can also be seen in the impedance plot.
Figure 1. Auxiliary / Box volume window for VituixCAD showing modes for 12.5 liter half cube box.
Figure 2. Hornresp calculation showing modes at the correct frequencies for 12.5 liter half cube box.
As you can see above, there is no mode at 600 Hz in the Hornresp software plot.
To replicate the plot in Hornresp you simply enter the values for a sealed rear chamber as I did below.
Auxiliary calculates just "fundamental" modes with dimensions. It does not know where the driver and "internal mic/ear" are located so it cannot calculate actual pressure response at any location inside the box. Designer has possibility to attenuate resonances by locating driver(s) and inner end of vent(s) to position where the (lowest/) strongest resonances are attenuated enough. For example driver to the middle point of the longest dimension. Unfortunately VituixCAD does not provide any help with that. Locating driver is usually easy, but vent requires both simulation of internal response at the inner end and multiple experiments and tuning. I usually simulate initial location with REW's room simulation which has adjustable source and mic location. Dimensions x10 -> frequencies x10. The lowest pipe resonance could drop more than 15 dB compared to intuitive/beautiful location.
Fair enough. My point was assuming the most obvious location of the driver centered on the front face of the box, based on where width, height and depth are measured and making a nearfield measurement with a microphone outside the box directly in front of the driver, you will get the response calculated by Hornresp and you won't see significant output at the approximately 590 Hz mode in the measurement because the driver can't effectively pump that mode. I'm not sure where a driver could be located on a box face that would excite the modes as shown in the VituixCAD plot.
The tool just presents a result that sort of perpetuates the myth that a cube enclosure will have resonant frequencies emitted from the speaker all lined up at the same frequencies, while the half cube shape does not. Obviously as you state, as the driver location is not explicitly described, it is up to the user to figure out what those plots mean in actual use.
Again, great tool and it is much appreciated.
Last edited:
Few speakers / concepts where woofer generates the lowest resonance.
Gradient Evidence; 3-way with down firing woofer. Lower XO point is 200 Hz and the lowest resonance ~230 Hz. On the safe side, but not much.
Many Jordan jx92 designs. Must be thousands of these out there
Side or front firing woofer(s) very close to floor. White in the middle has the lowest XO point at 400 Hz which is more than octave above the lowest mode.
Putting a woofer at the end of a long pipe shaped cabinet isn't something I've done since my first two builds in the late 1980's for that reason. I have since separated the shape of the driver enclosures from being dictated by the wide front baffle that I prefer in most of my designs, following the lead of Arnie Nudell and recently Grimm Audio with their nice half cube shaped woofer box laying on it's back on the floor facing up.
I'm not sure I would call 30Hz margin "safe" as that would require one of those " infinite slope " crossovers with a notch carefully tuned to 230Hz. In this popular tower design the tool correctly identifies two of the three box modes that get excited. Every tool has it's limitations. As always engineering involves calculation, construction and then verification testing.
Damping of long slim box is not a problem. Damping material is located to velocity maximum i.e. in the middle. Problem could be keeping Helmholtz resonator alive at LF so there can't be too heavy filling. Usually internal resonances are not as significant as external so primary target for box shape and driver location is to provide best possibilities for smooth in-room response at possible speaker and listener locations. To be accurate and flexible without requirement for EQ or bass level control. Woofer close to floor (such as Grimm) is just one compromise. It has quite good changes to sound the worst possible bass ever - just like Gradient Evidence or 1.1-1.3.
A long and high cabinet is only problematic in a sense that the frequency of the internal standing wave will be quite low as well.
For a cabinet that is 1 meter long internally, we are talking about f=344/(2*L) = 172Hz without end-correction.
Which is a little hard to dampen by most damping materials, even if you stack them up.
Not to mention that it often conflicts with the positioning of parts (speaker, BR port etc).
Sometimes a extra angled board in the middle will help = creating a ML TL
For a cabinet that is 1 meter long internally, we are talking about f=344/(2*L) = 172Hz without end-correction.
Which is a little hard to dampen by most damping materials, even if you stack them up.
Not to mention that it often conflicts with the positioning of parts (speaker, BR port etc).
Sometimes a extra angled board in the middle will help = creating a ML TL
@kimmosto it is just an incredible piece of work you have done with this SW. Thank you so much, I have used it for a while and I almost never find myself reaching for anything else. However, from time to time, there are functions I would love to see. Some of them are probably very easy to implement. I will probably think of many more as time goes, but here are a few. I hope it is ok that I share some suggestions here:
I have so many other ideas, I probably need to stop here for now.
- The polar plot is limited to +/- 90 degrees (unless I have missed something). 180 degrees would be awesome.
- I spent half a day looking for the frequency slider for the polar map. It is a bit hidden, and the manual did not give me any clues there.
- When I pushed the 90 vs 45 degree buttons while being in the polar map, the polar map changed, but it is not clear to my why this happens. May this be a bug?
- The power scale is limited to 5kW. I have found myself in the need of more on many occasions.
- You previously had a function where one could use the transfer function directly from the filter program in box calculations. I understand the separate filter dialog is useful, but I do miss the old version.
- If we had the possibility to put the box simulation in the the filter program, we could have the filter program use the simulated impedance, frequency response phase etc in filter simulations.
- The passive radiator dialog is limited to 10 units. I made a huge mistake when using 12 PRs. I doubled the parameters for the PR and put in 6 instead. But I did not compensate Qms correctly. If we had the possibility to use more PRs, it would be very helpful. Another useful function that would affect the same situation in a good way is if Rms and Qms are both shown, and automatically cross calculated. It would also be nice if we could lock the paramters that are given by the manufacturer. I often experience that the Vap is altered by some other parameter (not Cmp or Sd).
- In the driver dialog, the possibility to lock parameters before activating the cross calc function would be an excellent function. For example, plotting in Eighteensound drivers can be a P in the A due to their limited datasheets.
- For power simulations, having the posibility to show voltage and current curves would be an excellent addition. As you already have the complex impedance and power curves, I believe these are datasets already in the back end of the SW.
- In the driver properties dialog, we can adjust X Y and Z directions. It is a bit hard to figure out what is what. X, Y and Z is intuitive enough, but if + if closer or farther from the listener is not as intuitive in my opinion. Some markings could be useful.
I have so many other ideas, I probably need to stop here for now.
Member
Joined 2003
All directivity / polar plots are capable and intended for 180 deg of information. Make sure "half space" is not checked in options. Requirement is that you load 180 deg of information to the driver frequency responses.
Not sure what you're referring to, please provide screenshot example.
Here are my settings:
The menu:
Am I missing something here?
+/- 45 degrees selected (looks like it shows full 360):
+/- 90 degrees selected:
This looks very much like a bug. I feel pretty sure there have been a 360 degree option at some point, but I am not 100% sure.
You can check 90 degrees or check 45 degrees...OR uncheck them both and the default is 180 degrees.
Member
Joined 2003
What A4E said. 45deg plot of polar chart may be a bug, but I'm not sure what the point would be to reduce the view down like that. On other views, reduced to 90 of 45 allows for zoomed or reduced clutter of multiple traces. To be honest, I don't think many people even use polar chart, since polar map, line chart, waterfall are much better presentations of the same information.
2.0.111.1 (2024-02-10)
Main- Show ±45 deg in context menu of directivity chart enabled for Polar chart.
- Generator's default maximum voltage increased to 283 V.
- Maximum voltage increased to 283 V.
- Maximum power to 8 Ohms increased to 10 kW.
- Maximum power to 4 Ohms increased to 20 kW.
- Maximum signal voltage increased to 283 V.
- Maximum signal power increased to 40 kW.
- Added current (A) trace to Power chart.
- Maximum number of passive radiators increased to 24.
Frequency scrollbar for Polar chart has been visible in user manual as an image and text since the beginning (several years). Text is now moved a bit higher. I'm aware that scrollbar is not very visible especially in Windows 11 with default theme.
Coordinate system is explained in every measurement instructions (4) and user manual. Delay/phase reaction is one indicator which shows "polarity" of Z. Of course it requires understanding how phase works when distance/delay changes. Comparing e.g. buffer's delay and distance helps in that.
As DcibeL already mentioned, Polar chart is probably the least useful and used directivity graph. 'Show +/-45 deg' was added 2.5 years later than 'Show +/-90 deg' (having at least some significance in the world). So missing support was not noticed by anyone in almost two years.