I don't know how they made WinISD, but I'm aware that it may be possible that using it to simulate designs far enough outside the norm, it could calculate things wrong, if normally insignificant parameters become significant but aren't included in whatever formulas it uses. Kind of like how I can't believe port dimensions are not considered in the SPL graph.
I'm playing around with what happens if weight is added to the driver, and wanted to try to find how the resonance frequency changes, to better understand what's going on. It doesn't always make sense, and I recently found something that really didn't make sense.
Using the impedance graph to judge the resonance frequency in a closed box, adding drivers raises the resonance according to the peak impedance, but if weight is added to the drivers, then adding drivers decreases the resonance. Well, not exactly. Checking it again, I was able to find conditions such that adding weight to 1 driver lowered FS, 2 drivers increased it slightly, then 3 or more drivers decreased it.
It's a faulty formula, right?
I'm playing around with what happens if weight is added to the driver, and wanted to try to find how the resonance frequency changes, to better understand what's going on. It doesn't always make sense, and I recently found something that really didn't make sense.
Using the impedance graph to judge the resonance frequency in a closed box, adding drivers raises the resonance according to the peak impedance, but if weight is added to the drivers, then adding drivers decreases the resonance. Well, not exactly. Checking it again, I was able to find conditions such that adding weight to 1 driver lowered FS, 2 drivers increased it slightly, then 3 or more drivers decreased it.
It's a faulty formula, right?
It certainly seems that way, unless you are inadvertently mixing up the combinations of additional weight and number of drivers.
Increasing Mmd decreases the resonance frequency.
Increasing the number of drivers (connected in series or parallel) increases the resonance frequency.
In the following examples, notice how the system with a single driver and Mmd equal to 20 grams has the same resonance frequency as the system with two drivers in parallel and Mmd of each driver equal to 40 grams.
Single driver with Mmd equal to 20 grams:
Single driver with Mmd equal to 40 grams:
Two drivers in parallel with Mmd of each driver equal to 20 grams:
Two drivers in parallel with Mmd of each driver equal to 40 grams:
Part of what led me to find this error was noticing that the graphs didn't make sense, so I was experimenting to try to understand what was going on. In part, I didn't know if the added weight was per cone or total per isobaric pair, so I did similarly as you above, and found similar relationships between weight and number of drivers, but that's when I also noticed the graph could go both ways.
1-8 drivers all with 0g and 25g
1-8 drivers all with 0g and 25g
I have now done the same tests as you, using the Dayton Audio MX15-22 driver. In Hornresp, with 25 grams added the resonance frequency continues to increase going from 1 to 8 drivers. There is no "doubling back" half way through.
(The 1 driver resonance frequency is a bit different to yours because the enclosure sizes we have used for our tests are different).
1 driver with 25 grams added:
4 drivers each with 25 grams added:
8 drivers each with 25 grams added: