I'm building a 3 way speaker with Satori drivers and for fun, I've been simulating different baffle shapes in VituixCad with the goal of smooth frequency response. I ended up coming up with this baffle shape which seems to give a smoothish response.
My thought is that if the baffle dimension is constantly changing radially with the correct change function, you should be able to get a smooth response. I just made the change function linear because I wasn't sure what the correct one is although I'd guess exponential or logarithmic would be the right one. I made the longest dimension the crossover frequency.
I was also surprised that even on a rectangular baffle if the speaker placement is right, you can pretty well smooth the response.
Any thoughts on this? I haven't listened or measured it yet.
Are there any other shapes I should look into or experiment with?
My thought is that if the baffle dimension is constantly changing radially with the correct change function, you should be able to get a smooth response. I just made the change function linear because I wasn't sure what the correct one is although I'd guess exponential or logarithmic would be the right one. I made the longest dimension the crossover frequency.
I was also surprised that even on a rectangular baffle if the speaker placement is right, you can pretty well smooth the response.
Any thoughts on this? I haven't listened or measured it yet.
Are there any other shapes I should look into or experiment with?
Last edited:
2" - 3" radius quarter round machining on Left / Top / Right baffle edges is used on:
Kii-3
Vapor cabinets
Magico cabinets
B&W 802
=====
Side woofer(s) are used to allow modest width front baffles when large radius / bevel cuts are used to reduce diffraction.
The truncated pyramid bevel cuts as used by Avalon have proven low distortion advantages and artistic value.
Kii-3
Vapor cabinets
Magico cabinets
B&W 802
=====
Side woofer(s) are used to allow modest width front baffles when large radius / bevel cuts are used to reduce diffraction.
The truncated pyramid bevel cuts as used by Avalon have proven low distortion advantages and artistic value.
Attachments
Wait! You started with the baffle -so projection of the waves forward only; now considering the enclosure, you put into the "equation" the front wave, the backwave, the front wave travelling in space and returning to the enclosure-and then back again.
Larger radius or elliptical rounding are probably the best, and moreover it looks good. If you add waveguide it will be even better. For example, 170mm WG and edge rounding 30mm, 220mm wide baffle.
See some of my projects:
PKAudio: DIY loudspeaker projects
See some of my projects:
PKAudio: DIY loudspeaker projects
Petr,
That's some really impressive builds you've got there, both physically and electrically. Will definitely bookmark your website and follow your progress on the ones still in development. Just wish the whole site would have English translation.
Since you've played with all three, what's your subjective perception of the NE123, MR16P and the 12MU vis-à-vis each other? Also of the WO24P since you chose to sell it.
If I understand what you're saying correctly, I didn't factor in frequency response of the driver itself, the effect of the enclosure on the low end rolloff, the off axis response and reflections? That is true, I'm still learning a lot and taking it piece by piece. I am a relative noob at all this.
That appears to correspond with the graphs above for configurations J and L.
The side woofer is an interesting idea as a way to increase baffle response for lower frequency drivers by sacrificing depth for front baffle area. I'll have to look at off axis response of the midbass driver.
I have CNC'd Tweeter holders that are based on (a). Subjectively, the sound improvement is beyond only reduced ripples.
I'm examining 100mm high tooling for my shaper to cut shapes (a) and (j) in front and side panels.
I have yet to hear a waveguide or horn that did not subjectively degrade the treble with added noise floor or phasey characteristic - Baffle in front of a tweeter? Pida, curious as to what your experience is with waveguides as you said "it will be even better". Can PM if that's best.
I'm examining 100mm high tooling for my shaper to cut shapes (a) and (j) in front and side panels.
I have yet to hear a waveguide or horn that did not subjectively degrade the treble with added noise floor or phasey characteristic - Baffle in front of a tweeter? Pida, curious as to what your experience is with waveguides as you said "it will be even better". Can PM if that's best.
Nope, I was just saying that the speculation upon the emission of the speaker based on the geometry of the baffle ( and geometry of the cone itself) should focus only on emission & propagation of the ( front) wave.
Once you take the whole ( speaker+box+room= a box in a box) you include also the backwave and its effect, when enclosed, to the speaker ( movement).
The words that you used are very unkind
What's that terminology ? Axis? Off-axis !? Frequency response ? AArrrgh
My working and listening experience includes only hi-fi dome and ring radiator tweeters in WG 170mm outer diameter and 30mm depth or smaller. I prefer waveguided tweeters, they provide both good measured performance and the sound I like.
"it will be even better" referred to diffraction mitigation. WG + rounded edges (>30mm roundover, or the quarter of ellipsis) provided the best result in that regard. The waveguides I used are quite shallow, so they are still sensitive to edges of cabinet.
"it will be even better" referred to diffraction mitigation. WG + rounded edges (>30mm roundover, or the quarter of ellipsis) provided the best result in that regard. The waveguides I used are quite shallow, so they are still sensitive to edges of cabinet.
I wonder if anyone has tried somethings similar for OB, a bit simpler than flower petalsI'm building a 3 way speaker with Satori drivers and for fun, I've been simulating different baffle shapes in VituixCad with the goal of smooth frequency response. I ended up coming up with this baffle shape which seems to give a smoothish response.
My thought is that if the baffle dimension is constantly changing radially with the correct change function, you should be able to get a smooth response. I just made the change function linear because I wasn't sure what the correct one is although I'd guess exponential or logarithmic would be the right one. I made the longest dimension the crossover frequency.
I was also surprised that even on a rectangular baffle if the speaker placement is right, you can pretty well smooth the response.
Any thoughts on this? I haven't listened or measured it yet.
Are there any other shapes I should look into or experiment with?
Hi Scott,
I tried it in OB and like it a lot, but it wasn´t simple... at least not for me. Some experimentation lead me to the petals, but for sure easier to realize form could be found as well. A lot of the smoothness comes from the fact that the flower picks up the angle of the membrane with the least possible irritation - that has made quite a difference.
All the best
Mattes
I tried it in OB and like it a lot, but it wasn´t simple... at least not for me. Some experimentation lead me to the petals, but for sure easier to realize form could be found as well. A lot of the smoothness comes from the fact that the flower picks up the angle of the membrane with the least possible irritation - that has made quite a difference.
All the best
Mattes
Attachments
Hi Scott,
sorry, I couldn´t see the spiral picture, I have a pretty bad satellite internet connection here (in the Greenland Sea...). Obviously the spiral is easier to do, but as I have some trouble (nothing scientific, just feelings...) with asymmetrical polars, I went with the petals.
Rudolf and others claim correctly that the round shape is the worst case for OB, but the petals are a way to break up the single resonance AND to maintain symmetrical polars.
Sorry for the confusion with the spiral...
All the best
Mattes
sorry, I couldn´t see the spiral picture, I have a pretty bad satellite internet connection here (in the Greenland Sea...). Obviously the spiral is easier to do, but as I have some trouble (nothing scientific, just feelings...) with asymmetrical polars, I went with the petals.
Rudolf and others claim correctly that the round shape is the worst case for OB, but the petals are a way to break up the single resonance AND to maintain symmetrical polars.
Sorry for the confusion with the spiral...
All the best
Mattes
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.