progress report.... 🙂
Empty test boxes finished ...large roundovers sculpted from XPS....
Polyester fibers ready for cutting and stuffing...this is 25 mm thick, 250 gram/m2 polyester, bought in textile shop where it is 30 times cheaper in comparison with audio shops. 😀
Exactly one m2 would go into 1 cubic foot if I would cut 30 x 30 cm pieces....this would almost be equivavalent to MJK's 0.5 lb/ft3 which he uses most of the time in his projects.....I will try this density and try to expand it manually to half of density. Idea is to use as less as possible, but I don't have any experience so experiments are necessary...
DATS impedance sweeps of empty boxes.... Tuning is at 32 Hz, which is 1.5 Hz lower than according to Hornresp simulation. But real box has fold and terminus is not into the free air but close to the box wall so effectively length is a little bit longer. 3/4 WL resonance should be around 135 Hz, but offset of 56 cm is good so no resonance wrinkle. Response measurement of terminus will show if this resonance is fully supressed. 5/4 and 7/4 WL resonances are clearly visible exactly where they should be at 230 and 320 Hz. It is nice to see correlation with simulation, it gives me confidence that I'm on the right track...Also i'm happy that both boxes are virtually the same.
There are some small parasitic resonances around 200 Hz (internal heght of the box resonance?) in both boxes and of course resonances of 18W drivers around 800 Hz, like in datasheet. It can be seen that this area is composed of one wide and two narrow wrinkles. What I like very much, with drivers in free air or inside the box, is that there is nothing above 800 Hz. Revelators are really drivers with no breakups at the end of their passband...this should help to make simple crossover to the tweeter.
Boxes are not glued, just screwed and sealed with sealing foam, so I expect that some of these minor resonances should go away in final box...
Box 1:
Box 2:
Empty test boxes finished ...large roundovers sculpted from XPS....
Polyester fibers ready for cutting and stuffing...this is 25 mm thick, 250 gram/m2 polyester, bought in textile shop where it is 30 times cheaper in comparison with audio shops. 😀
Exactly one m2 would go into 1 cubic foot if I would cut 30 x 30 cm pieces....this would almost be equivavalent to MJK's 0.5 lb/ft3 which he uses most of the time in his projects.....I will try this density and try to expand it manually to half of density. Idea is to use as less as possible, but I don't have any experience so experiments are necessary...
DATS impedance sweeps of empty boxes.... Tuning is at 32 Hz, which is 1.5 Hz lower than according to Hornresp simulation. But real box has fold and terminus is not into the free air but close to the box wall so effectively length is a little bit longer. 3/4 WL resonance should be around 135 Hz, but offset of 56 cm is good so no resonance wrinkle. Response measurement of terminus will show if this resonance is fully supressed. 5/4 and 7/4 WL resonances are clearly visible exactly where they should be at 230 and 320 Hz. It is nice to see correlation with simulation, it gives me confidence that I'm on the right track...Also i'm happy that both boxes are virtually the same.
There are some small parasitic resonances around 200 Hz (internal heght of the box resonance?) in both boxes and of course resonances of 18W drivers around 800 Hz, like in datasheet. It can be seen that this area is composed of one wide and two narrow wrinkles. What I like very much, with drivers in free air or inside the box, is that there is nothing above 800 Hz. Revelators are really drivers with no breakups at the end of their passband...this should help to make simple crossover to the tweeter.
Boxes are not glued, just screwed and sealed with sealing foam, so I expect that some of these minor resonances should go away in final box...
Box 1:
Box 2:
Revelators measured with DATS after some break in period, streched a little bit with 20 Hz signal before measurements. In vertical position, properly clamped, Vas measured with added mass glued to the back of the cone...I didn't want to leave any marks on the front side...9 dBu or 2,82 V used for sweeps.
Qes higher, VAS lower as expected. Fs around 33 Hz, not 28 Hz like in datasheet. They should really change datasheet because no way that these will ever come to 0.39 Qes and 0,36 Qts or 58 L of VAS...
Qes higher, VAS lower as expected. Fs around 33 Hz, not 28 Hz like in datasheet. They should really change datasheet because no way that these will ever come to 0.39 Qes and 0,36 Qts or 58 L of VAS...
Last edited:
Pair produced in 2023 is closer to specs and has lower Fs, pair from 2022 is a little bit more off. Hope that with additional playing suspension will loose more so they come closer. I made also sweeps with -2 dBu and 4 dBu. With lower levels Fs goes 1-2 Hz higher. I think that 2,82 V is more representative. Also when you do measurement immediately after 20 Hz test signal values come closer to the specs...
Last edited:
Oh, and thanks to pkaudio for his comment on sealing the tweeters.
They are not OK. Insted of 500 Hz Fs bump I have some kind of vented alignment with two impedance peaks... bass reflex tweeties ...😍
They are not OK. Insted of 500 Hz Fs bump I have some kind of vented alignment with two impedance peaks... bass reflex tweeties ...😍
I am glad my advice was helpful. 6600 + WG is good combo. 6600 can sound too hissing which requires additional parts in crossover to mitigate this, and WG is the most correct way to mitigate this.
So, it seems that rounded shapes are not necessary after all ... but I would not change my rounded design. 🙂
Audibility of speaker baffle edge diffraction
I don't know is it OK to share these kind of links, they are open source I guess, but coming from other sides. If something is wrong, mods can erase....
Audibility of speaker baffle edge diffraction
I don't know is it OK to share these kind of links, they are open source I guess, but coming from other sides. If something is wrong, mods can erase....
Progress report... some-sounds-in-the-room....first mockup crossover, cap on the tweeter starting to cutting tweeter response above 10 kHz and baffle step coil plus electrolytic cap on both woofers, for now working as a normal 2 way, not 2.5 way...I don't have all necessary values and must buy additional elements, but not before full scale measurements...
So, empty TL has just a little bit honky, wooden barrel sound, which goes away when you add a little bit of damping. First 40 cm from closed end is fully stuffed and behind woofers just loosely. It has enormous bass energy, I will need more damping...At the same time bass is very clear and textured, SB ceramic bookshelf can"t come even close although it is tuned deep. Both speakers "are going down to 40 Hz" but that"s all. Dynamics, jump factor and tangible feeling is on the side of Scan Speaks of course. Tunnel is holding cones and they don"t travel a lot even at high levels....
Revelators have a lot od resolution and are listenable even without crossover...not all units can do that. Sound is alive and relaxed, now I just have to do crossover which will not kill that.... 🙂
So, empty TL has just a little bit honky, wooden barrel sound, which goes away when you add a little bit of damping. First 40 cm from closed end is fully stuffed and behind woofers just loosely. It has enormous bass energy, I will need more damping...At the same time bass is very clear and textured, SB ceramic bookshelf can"t come even close although it is tuned deep. Both speakers "are going down to 40 Hz" but that"s all. Dynamics, jump factor and tangible feeling is on the side of Scan Speaks of course. Tunnel is holding cones and they don"t travel a lot even at high levels....
Revelators have a lot od resolution and are listenable even without crossover...not all units can do that. Sound is alive and relaxed, now I just have to do crossover which will not kill that.... 🙂