On the back is not an option considering that speakers will be near the wall, but side is a good option I guess. I can put it axactly in middle and with no interference with the mid room. I will try to draw it asap.
Thanks for the links, they will be usefull. The port draw I attached before is 6,6 cm but it is one of 2. I should have another one on the other side of the woofer.
In case of one tube on the side I can make it bigger around 9,4 cm (constant diameter). I wil try to use your optimizer.
Very interesting. Thanks a lot
@stv
thanks to your suggestion I was able to calculate this
and then design this
Using this tube on the side the speaker should be something like this
What do you think?
thanks to your suggestion I was able to calculate this
and then design this
Using this tube on the side the speaker should be something like this
What do you think?
Last edited:
Great!
So now I have some minor questions:
1. as input to your optimizer in the max port output SPL I used the value retrieved from the vituixcad simulation
is this correct?
2. I have in mind to use ABS filament and to use epox resin to make it smooth. anything against this approach?
3. the side position has any difference (from a sound point of view) rispect the front position of the tube?
4. Using the side position I could use a PVC tube. Which are the main difference between a straight tube and a horn tube?
Earlier I suggested reducing the volume for the mid. You really, really don't need a bucket where a bottle is enough. Better even. Optimize the volume combined with the high pass to be used, so you get a nice crossover slope. As an example: aim for enclosure size so that Qtc equals 1,0 and apply appropriate series network (capacitor) with Qf = 0,5. Bonus: almost no cavity resonances in the mid enclosure that should worry you.
As for ports: never, never put them in corners of enclosures. Somewhere in the middle of a panel reduces quite some resonance issues that might appear. Back panel really is a very good position, even when you think a wall is near: it's not, except if you put them mounted flat on the wall or built in a closet. Just 4 or 5 cm from the wall and you'll have no issues at all. In fact you can judge the quality of a design partly by the port placement 😀
As for ports: never, never put them in corners of enclosures. Somewhere in the middle of a panel reduces quite some resonance issues that might appear. Back panel really is a very good position, even when you think a wall is near: it's not, except if you put them mounted flat on the wall or built in a closet. Just 4 or 5 cm from the wall and you'll have no issues at all. In fact you can judge the quality of a design partly by the port placement 😀
The simulation shows only 85 dB output from the port (red graph). And you should use the maximum SPL, not the simulated 2,83 V output, probably more (95-100 dB?).
The suggested port may get quite big, because of the low tuning. You could allow some compression and define a lower max SPL.
Seems like a good method!
You could use the raw printed port to verify the tuning and coat it if everything fits.
It has mostly positive effects:
less diffraction for the baffle mounted drivers, eventual port or enclosure noises are less audible. Because of the low frequencies exiting the port the direction has a very low (no) influence.
If you use a tube port with roundovers inside and outside and you chose the same exit diameter the tube port will be longer for similar tuning.
Without roundovers the tube will excite chuffing noises and compress already at low levels.
Same for edges in the port, so probably better avoid sewage pipe joints.
You can use the optimizer tool to find the required diameter (Dext) for a port tube, allowing ~1 dB of compression at SPLmax. Just disregard Dmin and length in this case, for the tube port.
Last edited:
Reducing the volume, there is a bump at 150Hz that then complicates the crossover filter
ok, got it. I will check also back port.
Ohhh, this changes a lot the situation.
in this way the port length is too much for my 3d printer. So I have to consider again a PVC tube with a diameter of 10 cm. But now I know that I can put it on the side or on the back. So the dimension is not an issue anymore.
As alternative, I should print the tube in 2 pieces cutting it in the middle.
You'll have that bump in about every situation, unless you're going to correct the impedance with a parallel notch. Look at the sims in your other topic. They too have the bump, electrically and acoustically. The trick is this: use the combination of the electrical and mechanical highpass to create a decent highpass slope. And I mentioned other advantages of a smaller mid cabinet. I wouldn't ignore them. Even the circumstance that a lot of other designers use (too) big mid enclosures doesn't imply it's the right approach. Hope I don't step on some toes 🤐.
Decrease the Qa to 5-10 (full stuffed box) and consider the baffle step loss (brown line).
That's the difference in my sim with the same filter (dash-dotted line is with a 10L box).
Last edited:
And you gained 5,5 liter. But with 3,5 liter and 470uF in series you get this:
Black solid line: almost perfect 3d order slope. And it can handle a lot of power, almost 20Veff (dotted line) full range. Cone movement with max power stays in reasonable areas (red line).
Which goes to prove the 17cm SB is more of a perfect candidate for a satellite-sub system than for a 3-way. Like some of us said already I think.
Black solid line: almost perfect 3d order slope. And it can handle a lot of power, almost 20Veff (dotted line) full range. Cone movement with max power stays in reasonable areas (red line).
Which goes to prove the 17cm SB is more of a perfect candidate for a satellite-sub system than for a 3-way. Like some of us said already I think.
Ok, I was able to simulate something very similar with a volume for mid of 2,5L.
470uF in series of what? do you mean an electrolytic cap?
In series with the SB17. And yes, the normal/wise choice would be an electrolytic capacitor. As I often think, people claiming they can hear electrolytic opposed to film will have a hard time delivering proof on that. One has to account for higher ESR, but that’s easy in a midrange bandpass filter.
I was caculating the volume of the tube to subtract it to the cabinet volume and I found a strange thing for me.
Using the vituixcad simulator or winisd, for a strainght tube I have a diameter of 9,2 and a length of 22,6. This means
Vt = 4,6*4,6*3,14*22,6 = 1,5L
If I use the tube port with roundovers, fusion calculate a volume of 2,936L (around the double). Is this normal?
This port is quite small and will compress and make chuffing noises at high SPL because of turbulent flow at both port ends.
The optimizer port should behave much better, but it's also bigger.
For a similar behaviour as the optimized port you could try a tube port with Dext of 14,1 cm as diameter. Such a port will be even bigger (longer, around 50 cm) for correct tuning.
I don't recall if you already tried this, but if you place even a large resistor in parallel with the mid, it may make the mid high-pass, and low-pass easier. For example, if your mid has an impedance peak of 30 ohms, and you parallel another 30 ohm resistor with it, it becomes 15 ohms at the peak. If you added a 10 ohm instead it would be a little under 10 ohms.
That indeed is the cheap way to flatten the impedance. But add a cheap coil and a cheap cap and you can tune impedance flat without burning power away on a parallel resistor.