Progress, on monday i will receive the pertinax plates, and additional epoxy resin.
The extension of the baffle to hold the mid and tweeter are ready to be cut to size:
The drawings on the cld plate are to check all dimensions.
The with existing baffle integration is by means of a stepped seam. Both panels are 2 layers with a cld layer in between. From the existing panel a part of top layer will be cut away and from the extension a part of the bottom layer. The whole will be glued with a special epoxy which has a similar young's modulus .
This will be done after the cavity is reduced in volume by filling it with pieces of pertinax and epoxy glue.
By the end of the week this should be ready ;-)
Then the final layer of pertinax will be applied with a cld layer.
And then the shaping of the baffle by standing;-)
The part that will be 3D printed are the rings or inserts for the woofers and midrange. This will be printed with a flexible material , TPU probably.
For the tweeter i am contemplating about a flex fixture, with the rear more tidgidly by means of some damping putty.
The extension of the baffle to hold the mid and tweeter are ready to be cut to size:
The drawings on the cld plate are to check all dimensions.
The with existing baffle integration is by means of a stepped seam. Both panels are 2 layers with a cld layer in between. From the existing panel a part of top layer will be cut away and from the extension a part of the bottom layer. The whole will be glued with a special epoxy which has a similar young's modulus .
This will be done after the cavity is reduced in volume by filling it with pieces of pertinax and epoxy glue.
By the end of the week this should be ready ;-)
Then the final layer of pertinax will be applied with a cld layer.
And then the shaping of the baffle by standing;-)
The part that will be 3D printed are the rings or inserts for the woofers and midrange. This will be printed with a flexible material , TPU probably.
For the tweeter i am contemplating about a flex fixture, with the rear more tidgidly by means of some damping putty.
Last edited:
Before starting to cut material - the ohoh moment - double checking every dimension of interest. As with the last 25cm or so of baffle extension every mm off is an issue ;-)
Including the "waterlines" for the curvature to be grinded/sanded. Will take some effort.
(found not significant errors ;-))
And also marking the holes to drill for the woofers to mount. Could not resist putting up a small show, even my wife liked it ;-)
Ah, hard to wait for it being finished.
Including the "waterlines" for the curvature to be grinded/sanded. Will take some effort.
(found not significant errors ;-))
And also marking the holes to drill for the woofers to mount. Could not resist putting up a small show, even my wife liked it ;-)
Ah, hard to wait for it being finished.
Study of effect rounding.
As both @hifijim and @lrisbo indicated to focus on the rounding a bit more.
So i did make the curve more round so to speak. And then did simulations to get a feel for the effects.
Approach:
-Take the top portion of enclosure (290mm)
-Simulate in VCAD
-Simulate in AKABAK
-Compare portion vs full height of enclosure
-Summarise
Summary:
Of course well known, rounding of baffle edges has a significant effect. And both VCAD and AKABAK show similar results.
My enclosures are some 1086 mm high (baffle,), Using a portion, in my case top part of some 290mm introduces an bottom edge which has its effects as well, which cannot be neglected.
In my GAYA2 case the room to optimize is limited, a 30mm rounding is not possible on the top edge. So i made a spline curve based surface.
This turns out to be not bad at all!
In the next day i will run an AKABAK with full size enclosure , but for now projecting the VCAD sim from 290mm to 1080mm, i believe the tweeter response will be smooth enough, on both on and off-axis.
What remains:
The aspect i cannot sim yet is the use of a waveguide for the tweeter, as the choice for tweeter is final yet.
As soon as made i will sim that situation as well. In the mean time i proceed with preparing the enclusore, enough work to do ;-)
Top portion of Enclosure:
VCAD-Enclosure (width of baffle is used), both edge of 0mm and 30mm, responses exported for loading in VCAD:
Rhino3D , prepared for AKABAK
Meshed, and in AKABAK:
Also a 0mm edge model:
VCAD sim:
VCAD 0mm edge:
VCAD30mm edge (Note: means all 4 edges of baffle)
AKABAK sim:
Because i concocted a LEM model (loosely based on Bliesma t34A) it includes a drop off al lower frequencies. so i also ran this also with an infinite baffle, to get a baseline curve. Also the element size is limited to about 6.9 kHz. But the the focus is on the respone portion below.
Hard Edge:
Curved baffle:
While there are quite some similarities beteen the sims in VCAD and in AKABAK, the question that remains is the effect of just using a portion of the enclosure. So in VCAD a full height sim is done:
VCAD-Enclosure:
VCAD 0mm:
VACAD 30mm edges:
What remains to do is a AKABAK sime with full enclosure. And a sim with a waveguided tweeter.
As both @hifijim and @lrisbo indicated to focus on the rounding a bit more.
So i did make the curve more round so to speak. And then did simulations to get a feel for the effects.
Approach:
-Take the top portion of enclosure (290mm)
-Simulate in VCAD
-Simulate in AKABAK
-Compare portion vs full height of enclosure
-Summarise
Summary:
Of course well known, rounding of baffle edges has a significant effect. And both VCAD and AKABAK show similar results.
My enclosures are some 1086 mm high (baffle,), Using a portion, in my case top part of some 290mm introduces an bottom edge which has its effects as well, which cannot be neglected.
In my GAYA2 case the room to optimize is limited, a 30mm rounding is not possible on the top edge. So i made a spline curve based surface.
This turns out to be not bad at all!
In the next day i will run an AKABAK with full size enclosure , but for now projecting the VCAD sim from 290mm to 1080mm, i believe the tweeter response will be smooth enough, on both on and off-axis.
What remains:
The aspect i cannot sim yet is the use of a waveguide for the tweeter, as the choice for tweeter is final yet.
As soon as made i will sim that situation as well. In the mean time i proceed with preparing the enclusore, enough work to do ;-)
Top portion of Enclosure:
VCAD-Enclosure (width of baffle is used), both edge of 0mm and 30mm, responses exported for loading in VCAD:
Rhino3D , prepared for AKABAK
Meshed, and in AKABAK:
Also a 0mm edge model:
VCAD sim:
VCAD 0mm edge:
VCAD30mm edge (Note: means all 4 edges of baffle)
AKABAK sim:
Because i concocted a LEM model (loosely based on Bliesma t34A) it includes a drop off al lower frequencies. so i also ran this also with an infinite baffle, to get a baseline curve. Also the element size is limited to about 6.9 kHz. But the the focus is on the respone portion below.
Hard Edge:
Curved baffle:
While there are quite some similarities beteen the sims in VCAD and in AKABAK, the question that remains is the effect of just using a portion of the enclosure. So in VCAD a full height sim is done:
VCAD-Enclosure:
VCAD 0mm:
VACAD 30mm edges:
What remains to do is a AKABAK sime with full enclosure. And a sim with a waveguided tweeter.
The shape of the enclosure will act as a large chamfer in the vertical direction, the main place where any obvious diffraction will come from is the corner crease. That is hard to avoid unless the enclosure behind is altered to have a round over on that crease. A waveguide will help that to some extent depending on the size and coverage angle. It will be hard to get much of a worthwhile improvement from what you have already when the main enclosure design is fixed.
Is what i am estimating as well, given tests with a waveguide on my accuton c30-6-024 in the "old" version of my GAYA2 boxes;-)
Will do some sims with waveguide and c30-6-024, need to model a lem and bem model first for this accuton ;-).
(And replace the sloppy backchamber of the real thing) :-0
Brandon ( @augerpro ) helped me some time ago in getting a decent functioning waveguide for the accuton.
Now i am getting able to sim in akabak so i sim this.
That waveguide was with elliptic mouth, i can now work on a circulair mouth.
When Purifi brings their tweeter to the market, i will replace the waveguided accutons with those.
Will do some sims with waveguide and c30-6-024, need to model a lem and bem model first for this accuton ;-).
(And replace the sloppy backchamber of the real thing) :-0
Brandon ( @augerpro ) helped me some time ago in getting a decent functioning waveguide for the accuton.
Now i am getting able to sim in akabak so i sim this.
That waveguide was with elliptic mouth, i can now work on a circulair mouth.
When Purifi brings their tweeter to the market, i will replace the waveguided accutons with those.
while AKABAK3 is using 24 cores to sim possible tweeter configs, i continue to modify my enclosures with focus on mid/tweet part:
The back side is so dead and inert, it forms a perfect "ground" for supporting the magnet of the PTT4.0M08-NAC04 unit ;-)
And the extension of existing baffle:
Final checks before starting to sand the top curved shape, and the removal of a part of the existing baffle, for the stepped joint.
Once these extensions are glued and shaped, the routing for mid and tweet can take place.
And the drilling of the holes for mounting the bass units.
The pertinax i used is so strong i can make threaded holes. I ordered a special drill holder for exact and perpendicular drilling, should arrive monday or tuesday.
While inert and ridgid mass is good to deal with resonances, my cabinets will gain some 15 kg each. Will be a challenge to lift them for measuring :-0
The back side is so dead and inert, it forms a perfect "ground" for supporting the magnet of the PTT4.0M08-NAC04 unit ;-)
And the extension of existing baffle:
Final checks before starting to sand the top curved shape, and the removal of a part of the existing baffle, for the stepped joint.
Once these extensions are glued and shaped, the routing for mid and tweet can take place.
And the drilling of the holes for mounting the bass units.
The pertinax i used is so strong i can make threaded holes. I ordered a special drill holder for exact and perpendicular drilling, should arrive monday or tuesday.
While inert and ridgid mass is good to deal with resonances, my cabinets will gain some 15 kg each. Will be a challenge to lift them for measuring :-0
Made a checklist of jdi's before mounting the drivers. Should reach that goal later this week:
Also the top layer of the baffle shape is printed, and just checked for dimensionally correctness.
I decided to print this part, not to lose time.
I figured out how to shape quickly(using my router with a cylindical bit as a hi-perf dremel) when using pertinax plates (hartpapier (hartgewebe) ) instead.
But given the dust i need to do this in my backyard, but the weather lately was too bad.
So if needed as a( extensive )mod lateron this can still be done, but for now this is the plan.
Also the top layer of the baffle shape is printed, and just checked for dimensionally correctness.
I decided to print this part, not to lose time.
I figured out how to shape quickly(using my router with a cylindical bit as a hi-perf dremel) when using pertinax plates (hartpapier (hartgewebe) ) instead.
But given the dust i need to do this in my backyard, but the weather lately was too bad.
So if needed as a( extensive )mod lateron this can still be done, but for now this is the plan.