Nothing really important to add., but I do want to show some basic calculations. Progress has been a lot slower since school started again. I wanted to finish prototyping the outer shell during the winter break but that was extremely naive of me in hindsight. 10 days in Utah did not help on top of that, either.
Like I said last time, I do have a few printed parts made already, mainly to test the tolerance. An issue that greatly tests my patience, however, is the time it takes to print some of the parts. The center hub that will be housing the floating tweeter case takes well over 15 hours to print. And that number is printed at a layer height of 0.3mm, with less than 10% line support, 10% grid infill, and 3 wall layers. None of these are even remotely the right thing to do. Here are some reasons why:
If the nozzle is 0.4mm, the biggest height it should print while already sacrificing a great deal of quality and accuracy would be 0.2mm. Line support cannot even support itself since it is literally a line, which means it tends to collapse as the height increases. I feel that the grid infill should be at least 20-30%, but I can let it go since the real strength comes from wall layers, which should have been at least 5.
If the focus was on print quality, the time it would take to print the center hub would no doubt be over a day.
A video I came across on Youtube gave me some more insights on how to not only strengthen the speaker enclosure, but also lower the resonant frequency:
In this video, he uses a mixture of plaster of paris and PVA as the infill of the printed parts, which added a significant amount of weight to the enclosure and in turn lowered the resonance. I wonder if sawdust can be a safer alternative to plaster of paris.
Another internal debate is on the drivers placement. The "push-pull" woofers currently have tiny enclosures. One way to improve is to hollow out the enclosure wall that separates the two woofers, but in doing so I will have to give up on the other one. The woofers have different Xmax, different impedance, and different rms power level. If the combination of dynamat and polyfill can indeed make the enclosure sound bigger, I would rather keep both of the drivers.
Its enclosure, if the wall were to be solid, would have a size of approximately 0.775m^3. The highest omni frequency it could produce would be limited by the enclosure rather than the driver itself:
R = (3 * 0.775 / (4 * pi))^(1 / 3) = 0.57
fomni = 345 / (2 * pi * 0.6) = 97Hz
At around 100Hz, the sound would have been omni-directional anyway. I have yet to decide on the crossovers, but I might let the two opposing woofers handle frequencies below 100Hz.
I intend on connecting the pieces mainly with bolts, maybe hex screws, since the ability of breaking it down feels important to me. But man, it would be such a breeze if I just glued everything together.
Finally, the cost. A single DSP amp board with 4x30w outputs is already $60, and I expect to be using two. I will have to connect some drivers(8ohm, 15w) via parallel to the same output. The impedance will be halved given the equation:
Rtot = (R1 * R2) / (R1 + R2), R1 = 8ohm, R2 = 8ohm
Rtot = 4ohm
Total power would be:
Ptot = P1 + P2 = 30w
This means that the woofer drivers will be completely safe with the amp's power output. On the other hand, the 4 tweeters(4ohm, 15w) will have to have their own outputs for two reasons: each pair would be in a different frequency range, and the impedance would be too low. In order to make sure that they will not blow up under high volume, since the power output is literally double the power level of the tweeter, I expect to lower the gain of the output in SigmaStudio. Some calculations here:
Pamp = 30w, Ptweet = 15w, r = 4ohm
vamp^2 = Pamp * r = 30 * 4 = 120
vamp = 10.95v
vtweet^2 = Ptweet * r = 15 * 4 = 60
vtweet = 7.74v
Voltage_gain decrease = -20 * log10(10.95v / 7.74v) = -3dBv
v_rms = 10^(-3 / 20) * 1v = 0.71v
The output voltage needs to be 0.71v lower so that the amp is essentially pushing 15w to the tweeter.
Written 1/31/2024
Yorumlar