I have put a lot of work into calibration of my superslicer profile. I do a lot of negative forms in ABS for workholding strange SLM titanium parts in my CNC mill. Superslicer is the only slicer i know, where i can calibrate inside concave geometries size dependent, which is an absolute must for me. Highly recommend to try it!
I´ve gone through quite a few slicers already and there is no comparison to the profile i am using right now. The problem is specific to printing waveguides for me. Usually my parts are a lot smaller and i design them with printability in mind. A waveguide is just very specific in geometry and way bigger thant anything else i´ve done so far.
I don´t use ABS if i don´t need it (no coolant exposure in this case), because even with the best process ABS will warp a lot more.
I don´t use ABS if i don´t need it (no coolant exposure in this case), because even with the best process ABS will warp a lot more.
Quite honestly, I'd suggest to wait a while, as someone may measure some driver(s) they have at hand first. Hopefully I'll be able do to that soon as well. This is still so unexplored area (i.e. an extended throat) that it's difficult to give any recommendations at the moment.
Also, the cylindrical throats of the latest designs allow for an easy elongation of the throat, and I expect this to have a strong effect on frequency response. Which reminds me I still need to prepare STLs for the extension parts...
Also, the cylindrical throats of the latest designs allow for an easy elongation of the throat, and I expect this to have a strong effect on frequency response. Which reminds me I still need to prepare STLs for the extension parts...
going way back to post 12,725 here where the R-OSSE formulas are shown again, we have a ton of variables to be optimized and the example shapes have a lot of curvature in the middle compared to an OS that goes to near zero curvature by 5-10 throat radii axially (yes I see q=1 flattens the R-OSSE out). I'm not smart enough nor have enough computing power to optimize all those variables, though I have programmed them in to MS Excel. I note that an OS equation has zero extra variables - if you pick a driver radius and expansion angle, there is nothing more to pick.
So my question is, IF an OS has the lowest HOM, THEN why wouldn't we want to terminate our waveguide in the complementary OS like the picture below?
This leaves zero variables to optimize. A property of the OS is that at the throat (and termination of the complementary OS), the radius of curvature approaches a constant, and also that constant happens to be in the range of common round-over router bits. So by just joining a matching roundover radius to an OS+complemetaryOS shape, we get the re-entrant waveguide with zero variables to optimize and one that we would expect to have minimum HOM unless I am missing something (which I probable am, hence why I'm asking the question).
So my question is, IF an OS has the lowest HOM, THEN why wouldn't we want to terminate our waveguide in the complementary OS like the picture below?
This leaves zero variables to optimize. A property of the OS is that at the throat (and termination of the complementary OS), the radius of curvature approaches a constant, and also that constant happens to be in the range of common round-over router bits. So by just joining a matching roundover radius to an OS+complemetaryOS shape, we get the re-entrant waveguide with zero variables to optimize and one that we would expect to have minimum HOM unless I am missing something (which I probable am, hence why I'm asking the question).
I don't think it works this way. Throat of a horn is very different from mouth in terms of its dimension relative to wavelengths. What works at a throat won't work further down the horn.
- In an OS waveguide, the wavefronts converge to spherical very quickly (it can be slower or faster though - see the generalization by adding an expansion factor; page 4), but then the theory assumes it continues like that forever, with no further source of diffraction. That's where some other means must be used, and basically, you need a much larger radius at the mouth termination for it to be "invisible".
In fact, I tend to think about the OSSE and R-OSSE profiles as controlled-diffration horns. That's probably the most suitable term.
- In an OS waveguide, the wavefronts converge to spherical very quickly (it can be slower or faster though - see the generalization by adding an expansion factor; page 4), but then the theory assumes it continues like that forever, with no further source of diffraction. That's where some other means must be used, and basically, you need a much larger radius at the mouth termination for it to be "invisible".
In fact, I tend to think about the OSSE and R-OSSE profiles as controlled-diffration horns. That's probably the most suitable term.
Last edited:
Hello, is anyone available for a gig? I'd like to follow Art's recommendations: altering a multiple entry horn (SynTripP) to improve its dispersion and efficiency. I'm not familiar with horns and the tools involved so I'd rather have someone experienced doing it.
First test with open enclosure.
Pretty much the same problem as before.
Next test is ABS style with reduced cooling above 80mm.