Open source Waveguides for CNC & 3D printing!

[augerpro]

Like this: https://www.amazon.com/Revell-Plast...002HZ3N0/ref=cm_cr_arp_d_product_top?ie=UTF8?

 

[YSDR]

I don't have the SS D2604, but now I've found that they are all the same shape:
ScanSpeak-Waveguides

 

[augerpro]

Good news, thanks!

 

[Sammel]

I compared an SLS nylon print vs my own PLA print and they measured identically. Neither has a smooth finish, but are different, the PLA has the layers like you have, the SLS nylon has no layers but has sort of a sandpaper feel. I'm currently investigating ABS smoothed with acetone vapor for my box ports since smoothness did impact the performance in my measurements. Might be smething to look at for waveguides.

BTW: make sure you use 4 layers for all sides. I can see some infill poking through in some areas.

Thanks for the info I needed.
Don't know how many layers they used

I had some great results with Revel Plasto and high grit sandpaper. I got perfectly smooth surface ready for spray painting.

Thanks for the tip.

kind regrds
Samuel Strågefors

 

[pelanj]

augerpro: Yes, exactly this one, it dries pretty fast, sands easily and makes a nice smooth surface. It is what the plastic kit builders use. A bit on the expensive side, I was told acrylic wood putty would work as well, but have not tried that yet.

 

[mainframe99]

Here is the Bliesma T25B on two waveguides with simple filtering applied, 0 and 40 degrees (which I find to be a quick and dirty proxy for power response or early reflections). Hopefully this is a cleaner display of the performance we're talking about.

Curious, what filtering is being applied for these waveguides? Is it just a low Q low shelf or a low pass?

 

[mainframe99]

Tweeter Managed to get pretty flat with just a lowshelf 0.5Q at 10khz and a peak down at 1.8khz


MW19TX-4, SB26ADC-AugerPro Waveguide, only a test box at this stage, sealed 20L, 5ms gate on horiz off axis and 4ms on the vertical. Cross 1600LR4. Still needed 0.08 delay on tweeter for some reason?


Open to critiquing , sharp eyes will notice the waveguide not quite level - routing out ellipse isn't easy - probaby explains the uneven vertical behavior

+/- 80 deg horizontal


- zoomed in 2db increments


-vertical +/-30 deg 2db increments

the haphazard garage setup, sorry no klippel NFS



Pretty happy with waveguide. Donation incoming to augerpro


Okay the 2db increments were on my screen but didnt come through on the captures but it gives you an idea

 

[bcodemz]

Curious, what filtering is being applied for these waveguides? Is it just a low Q low shelf or a low pass?

It should be neither when done properly. The waveguide actually provides a broad peak gain. So the proper filter to EQ out the waveguide gain is actually just a single peaking filter with the correct center frequency and Q.

You can see this clearly if you load the on axis measurement into something like VituixCAD. Then apply a Linkwitz Transform to flatten the tweeter to 10Hz or something really low. If a normal tweeter is Linkwitz Transformed to 10Hz, it should be practically flat from 10Hz to 5000Hz or so. Doing that with a waveguided tweeter will let the waveguide boost in the form of a peak become very apparent. Then just play until the response is completely flat. The resulting filter to make it flat is the filter to counteract the waveguide gain.

 

[bcodemz]

Tweeter Managed to get pretty flat with just a lowshelf 0.5Q at 10khz and a peak down at 1.8khz


MW19TX-4, SB26ADC-AugerPro Waveguide, only a test box at this stage, sealed 20L, 5ms gate on horiz off axis and 4ms on the vertical. Cross 1600LR4. Still needed 0.08 delay on tweeter for some reason?


Open to critiquing , sharp eyes will notice the waveguide not quite level - routing out ellipse isn't easy - probaby explains the uneven vertical behavior

+/- 80 deg horizontal

View attachment 940752
- zoomed in 2db increments

View attachment 940753
-vertical +/-30 deg 2db increments
View attachment 940754
the haphazard garage setup, sorry no klippel NFS
View attachment 940755


Pretty happy with waveguide. Donation incoming to augerpro


Okay the 2db increments were on my screen but didnt come through on the captures but it gives you an idea

Horizontals look very good. I'm curious, do you have measurements of just the tweeter's vertical response without the woofer?

 

[mainframe99]

Just because I've got everything set up, here they are. Now remember - these are on a very agricultural jig that isnt centred properly, theres some crazy diffraction effects from the top of the baffle, and in a garage gated at 4ms. These measurements aren't exactly what you'll get but only to give you an idea of the waveguide trends. IMO these measurements should be done in a larger baffle but no way am I routing another ellipse today. Also no voltages are recorded so I cannot vouch for the SPL accuracy of my UMIK-1



Above speaker;





and below

 

[mainframe99]

And after some EQ tweaking shes +/- 1dB from 200hz up (on tweeter axis, 1m, 5ms gating)

 

[bcodemz]

Very nice. Thanks for the measurements. It's looking great! The diffraction is sort of unavoidable due to the woofer and baffle, but the waveguide is already reducing its effect with limiting dispersion.

 

[augerpro]

Fantastic results, and looks good too! What waveguide is this, the 8"? When you say LR4 1800hz, that is the actual acoustic response? If so, did you need any delay to get that phase agreement?

 

[mainframe99]

6.5 inch, aesthetic choice same width as mw19. 0.12 delay on tweeter in that last measurement. 1600 LR4 on the woofer 1600 LR4 on the tweeter after EQing drivers flat out of band, mini dsp. Still testing.

No doubt the 8” waveguide would perform better.

 

[mainframe99]

After over a weeks listening to the speaker and tweaking xo, (i settled on 2.4khz) I can safely say every box speaker I make from now on will have one of these waveguides. Get it right and it is just mind blowing the improvements over a flush mounted tweeter.

 

[shining31]

Rhaaaa, that's a good news And great work mainframe99
I'm working on my mw16tx + sb26adc/5.25"wg, I can't wait to listen to them !


Can't you guys tell me what are the hole diameters for threaded rod and the screws ?

I dont't have my notes with me and like to order some stuff ...


I remember screws M5 and threaded rod M4 ? Am I right ? I can't find it on this thread...

 

[augerpro]

After over a weeks listening to the speaker and tweaking xo, (i settled on 2.4khz) I can safely say every box speaker I make from now on will have one of these waveguides. Get it right and it is just mind blowing the improvements over a flush mounted tweeter.

Awesome! I once did a comparison where I used a Vifa NE19T with and without a simple waveguide. I would just flip back and forth in real time. At first I didn't have a preference, both had pros and cons. But the more I listened the more I just enjoyed the waveguided one more. The "liveliness" of the bare tweeter just became hashy. The waveguided one was calmer, clearer, and focused at me in some way that is hard to describe. I likely won't go back either. BTW, what drove you to raise the crossover?

Rhaaaa, that's a good news And great work mainframe99
I'm working on my mw16tx + sb26adc/5.25"wg, I can't wait to listen to them !


Can't you guys tell me what are the hole diameters for threaded rod and the screws ?

I dont't have my notes with me and like to order some stuff ...


I remember screws M5 and threaded rod M4 ? Am I right ? I can't find it on this thread...

I'm not at home now so I can't check, but the holes for the tweeter clamp are the right size to tap for one of those sizes, M4 probably. The mounting flange holes are big enough not to touch a #8 screw.

 

[mainframe99]

Not sure about the smaller waveguides - augerpro could confirm, but I tapped the hole directly (slowly) with an M4 tap, and M4 zinc bolt straight into it. My bolt was a little long so a nut was added. Just dont over tighten, not much torque is required to hold the tweeter firm. This is a SLS Nylon print mind you so not sure about other plastics!



I later added bluetac all around the tweeter to seal, as mentioned earlier in this thread.


Augerpro - after playing around in vituixCAD I got got better DI (and sound!) with a crossover at 2.4, I think it might be something to do with the Textremes wild off axis behavior. 1.6 worked perfectly though, no distortion or graininess of any kind from that tweeter.

 

[shining31]

Thx you !
@Augerpro, as usual, each time I have the opportunity to talk to you, thanks you for your work. It's inspiring, and show that serious work can be achieved with a sharing and DIY mindset !



@mainframe, an aluminium plate ! Obviously !! And thx you for your measurements and the logbook

 

[augerpro]

Thanks for the kind words!

BTW, please please please use foam under the clamp and just snug the tweeter down. A rigid clamp like that could put extremely high forces on the diaphragm and this stuff is delicate.