Update of post #556 but using correct strouhal calculation method:
data from the RTA chuffing/noise/harmonic distorsion graphs, but arranged in a more comprehensive form:
the correct calculation method for strouhal number (credit to DavidMcBean) just rearranges the "strouhal number" limit of 1.
It is now clear that for useful music reproduction the strouhal number should remain a good safety margin above 1.
because the graphs above are mostly interpolated between the measurements at 0,5 - 1 - 2 - 4 - 8 V, I will provide a new measurement with smaller increments and additional subjective impressions.
relating the strouhal number to compression and distorsion
data from the RTA chuffing/noise/harmonic distorsion graphs, but arranged in a more comprehensive form:
- strouhal number, calculated for the port variants at respective input voltages (referring to port exit radius).
- compression ratio (dB), referred to the lowest input signal
- 3rd harmonic distorsion
the correct calculation method for strouhal number (credit to DavidMcBean) just rearranges the "strouhal number" limit of 1.
It is now clear that for useful music reproduction the strouhal number should remain a good safety margin above 1.
because the graphs above are mostly interpolated between the measurements at 0,5 - 1 - 2 - 4 - 8 V, I will provide a new measurement with smaller increments and additional subjective impressions.
I will see if I have time today to verify your tool with my existing ports. If anyone wants to use this port, attached is the F360 .f3d file. The only user change should be port length and port radius at the center of the port. Everything else should scale with those changes.
Will you be adding the Strouhal number to the spreadsheet? If this is the best criterion for judging for performance, it would be good to include this calculation.
Here is the comparison of this spreadsheet to 5 test ports I made. All ports are Dmin = 1.5". Some caveats...the measured tuning can shift within about +/- 0.7hz every time I make a measurement. I selected the value that came up most often in a run of 6 measurements. Also, I'm converting from imperial to metric when I use stv's worksheet, so there will be a bit of error there.
I also included aspect ratio of the length to Dmin. It is interesting that if I had a port with aspect ratio 3.25:1 there would probably be zero error between calculated an actual. Above and below that we see a systematic error with a pretty clear relationship to the actual measurement. Certainly this is very promising and usable for me!
Here is the comparison of this spreadsheet to 5 test ports I made. All ports are Dmin = 1.5". Some caveats...the measured tuning can shift within about +/- 0.7hz every time I make a measurement. I selected the value that came up most often in a run of 6 measurements. Also, I'm converting from imperial to metric when I use stv's worksheet, so there will be a bit of error there.
I also included aspect ratio of the length to Dmin. It is interesting that if I had a port with aspect ratio 3.25:1 there would probably be zero error between calculated an actual. Above and below that we see a systematic error with a pretty clear relationship to the actual measurement. Certainly this is very promising and usable for me!
Last edited:
If you really want to turn it up to 11, attached is the .f3d for my port with 1 resonator to damp the pipe resonance. 2 resonators damped the pipe resonance more but really suffered in chuffing and distortion. 1 resonator is a decent tradeoff. I couldn't detect any issue in my very short listening session, but Erin didn't hear any issues either when I sent my prototype to him. Again, user inputs are length (d3 in sketch 1) and port radius at the center of the port (d2 in sketch 1). I couldn't make the same tweak as the version I posted above because one of my operations depends on this port radius value, and for whatever reason, operations cannot depend on a driven dimension in F360. So after using stv's worksheet to design the optimum port, use half of Dmin to enter this value in d2. The third user input is related to the diameter of the caps that cover the resonator. Enter the value from d151 in sketch 8, into d74 in sketch 5. I use super glue to attach these and fill the resonator with wool batting.
That's very generous of you, thanks!
I'll try it out, very curious about the result.
Thanks for the tests! Yes, that looks pretty usable to me too!
I'll be glad to make an imperial unit version of my spreadsheet, just give me some days.
I might eventually include a guideline for strouhal number simulation via hornresp (yes, hornresp can output a strouhal number graph since the last update!!).
Strouhal number seems to be a good indicator for port behaviour.
One reason for widespread airspeed rules of thumb could be a tradition of common tuning frequencies for vented speakers - maybe 40-50 Hz?
I'm still not sure about what limit for strouhal numbert to choose. 1 seems a bit too low for high quality music listening.
By the way, once the spreadsheet is grown up please feel free to use it for your "monster box construction collection".
I've wanted to share this port design for awhile but since I could not figure out the tuning, I kind of put it on the back burner. If your spreadsheet works out and can calculate a Strouhal number, I'd be happy to host that and my .f3d files on my website so people can start making these ports.
When do you think you might be able to do your own tuning verification? It seems to me the systematic error is a good thing because you can correct for that by finding an empirical correction factor.
When do you think you might be able to do your own tuning verification? It seems to me the systematic error is a good thing because you can correct for that by finding an empirical correction factor.
I will test
1) ports with more "extreme" dimension ratios within the next weeks
2) a higher power 12" woofer and some corresponding bigger ports with different tunings. This will hopefully happen until end of june.
I suppose after tests number 1) and considering your tests above that supposedly are using a small driver (is that correct?) we can safely say the calculator gives usable results for small speakers.
For bigger (sub-) woofers I'd like to do test number 2).
The green line on the attached chart shows peak particle velocity vs frequency for a Strouhal number of 1, given port outlet area Ap.
Would this feature be worth adding to Hornresp? It would be switched on or off by pressing the L (for limit) key.
Attachments
Absolutely!
I am still working on whether the strouhal number limit of 1 is really useful for high quality music reproduction or if it should stay even higher.
But the number one limit can be a great guideline that needs to be respected with some safety distance.
What we do know for sure, confirmed in the papers, is that going below a strouhal number of 1 does degrade efficiency and increase noise.
I am about to make a chart with all measured vs calculated ports, to see if I get similar deviations!
Could you check once more the aspect ratio - I cannot replicate your results:
First example Lact=3.75 Dmin=1.5
3.75/1.5=2.5
You stated 2.17 however.
Am I getting something wrong here?
Thanks again!
Doh! My original notes used the length minus 1/2 of the termination roundovers as you normally do when calculating port length with roundover terminations. So the length used for aspect ratio is actual length minus 0.5". I'm not sure which method you want to use but I would think subtracting these termination lengths would be more accurate, at least for shorter ports. Unless it is somehow accounted for already in your calc?
Perfect, I just wanted to know the calculation method for comparability!
I have a similar "error" for big length/diameter ratios.
I'll test some more ...