BTW I also have a couple designs for the Bleisma T25B on tap. The throat is very small so i'm hopeful we can get at least SB26ADC-level results, and hopefully even a bit better in the top octave.
2nd BTW: I have a couple variants for the typical Scanspeak tweeter. If someone would send me a tweeter to use, I'll send those designs to the printer also. Two would be similar to the one mbrennwa is using, and a third is large, 6.5" I think.
Yes, that Bliesma T25 has an inverted surround, so maybe you can overlap it a bit with the throat, to get a smaller diameter.
Jeroen
Jeroen
Much worse without a WG... Because it's very much wider in despersion.
SB 26ADC traced from manufacturer, simulation Without WG - Album on Imgur
An externally hosted image should be here but it was not working when we last tested it.
SB 26ADC traced from manufacturer, simulation Without WG - Album on Imgur
If you can show me these measurements for the 6,5" I can trace them again and see the difference. 🙂
See my reply above for the tweeter without WG
Do you have to remove the grill, or does that remain fixed in place? I'm very curious about these drivers.
Also, I've got a pair of SS6600s that I could loan you for testing. Feel free to PM.
I still have to get it printed. Probably gonna be a couple months. Still I like your results. One design criteria for the 5" waveguides was that they would have similar dispersion to a 12-15cm woofer at 2.5khz. That frequency would be ideal for the crossover, but I did wonder how much room there was to go lower in the real world. Your results show we can go lower, though your crossover is probably a bit low if we are going for perfectly smooth directivity (according to your sim anyway). Revel uses a similar low crossover on the M105 and that is the same tweeter, similarish waveguide, and 15cm woofer, and their results are pretty good.
Yes grill removed. It's pretty easy to do, I think ergo over at htguide did a walkthrough with the T34.
I'll PM you when I get a better idea when I'll have the waveguides printed.
Okay then! For my project I have actually printed the 6 inch version, so it will be a bit better I hope. But my baffle will be ready in a few months as well, so it's nice to be able to simulatie it beeforehand.
Maybe in the meantime I will design a mtm with 5 inchers, using Horbach Keele crossovers. See where that leads me...
The best of the waveguides so far are on there. I'm going to put some wavelet plots up when I get a chance.
A beryllium dome without a grille is not a great idea in terms of safety if you ask me. The beryllium dome tends to break if you touch it, and the stuff seems to be toxic. Maybe you could integrate a grille or similar protection like with the SB beryllium domes into the waveguide?
The images are visible to me on Firefox and on Chrome Android. Imgur might be blocked by a plugin in your browser.
I don't know why, but when I try to post the images here directly they can't be loaded or something. I don't know why my pictures don't show directly... Could it be relatied with my permissions?
Did you try to embed the pictures in the text? That does not work well for many users here. Just attach the images using the "paperclip" icon in the toolbar.
Linking to external graphics is not a great idea, because the links may (will) change at some point, or the files may disappear, and then the thread breaks because the information is not accessible anymore.
Alright, I think I got it!
Once again. Simulation using Vituixcad:
- Of-axis response via the Diffraction tool with SB17NBAC-4
- SPL traced SB26ADC in 5" WG
Versus a simulation of SB26ADC traced from manufacturer, thus without WG:
Once again. Simulation using Vituixcad:
- Of-axis response via the Diffraction tool with SB17NBAC-4
- SPL traced SB26ADC in 5" WG
Versus a simulation of SB26ADC traced from manufacturer, thus without WG:
Last edited:
Big improvement over no waveguide! If you are bored, try pushing that crossover point up until that DI curve is smooth as can be. I'm curious if you have to all the way to 2.5khz, or if a lower frequency can work as well.