If anyone wants to go on this journey with me, I've attached the waveguide files that will be used for these projects. I still need to finalize the baffle dimensions. I will probably go with 4 ohm woofers, but haven't decided on that yet. That's a major cash outlay for me atm.
bruh, VituixCAD.
Just kidding. You're gonna love it. I wonder whats happening at 500hz - thats not a cabinet resonance I bet, I recon its more something to do with the 'cabinet' being a waveguide thing that old mate patrick was on about. Speculation though, I don't actually know. Hope you can figure it out.
Wish I lived down the road from Erin, I'd be churning out DIY designs weekly I recon, and he'd always be drunk. (from beers for using his NFS)
Edit - just saw your reply - give me the FRDssss
Just kidding. You're gonna love it. I wonder whats happening at 500hz - thats not a cabinet resonance I bet, I recon its more something to do with the 'cabinet' being a waveguide thing that old mate patrick was on about. Speculation though, I don't actually know. Hope you can figure it out.
Wish I lived down the road from Erin, I'd be churning out DIY designs weekly I recon, and he'd always be drunk. (from beers for using his NFS)
Edit - just saw your reply - give me the FRDssss
Think I may have already but here it is. It started as a revel m105 copy with the aluminum SBA woofer, but I just couldn't shake a kind of glare in the midrange no matter what I did. I threw in the old trusty 15w8530, and I've lived happily ever after. I need to update the ASR thread I did on it, but I want to wait to finish the cabinets when it warmer here so I can also post some nicer pics.
If anyone wants to build it, PM me for correct XO parts first as I may have made some little tweaks after this. Also the old backwards tweeter phase check yields a -40db null.
I believe the glare is because no matter you do, you can’t suppress the resonances off axis
Disconnect the tweeter in Vitixcad2, and view the midwoofer only, down to 35dB (a very quiet room, at night), and see compare what hash you get up to 2 octaves past the XO point, all the way out to 90 degrees, for the SB15 vs SS 15W
# Paper cone vs metal cone data
# coming soon
Disconnect the tweeter in Vitixcad2, and view the midwoofer only, down to 35dB (a very quiet room, at night), and see compare what hash you get up to 2 octaves past the XO point, all the way out to 90 degrees, for the SB15 vs SS 15W
# Paper cone vs metal cone data
# coming soon
^ regarding this phenomenon I gotta ask what the crossover is like? would this hash/glare associated with metal/ceramic cones be reduced if driver distortion current was reduced at cone breakup, like Purifi tech paper shows? First PDF, low distortion crossover https://purifi-audio.com/tech/
yeah that method only works if the on axis and off axis breakups are right around the same freq.
you may be on to something there.
circled in red the possible problem? my XO was LR4 at 2.2khz, so the 10khz breakup stuff was way gone.
anyway, I'm raining on augerpros parade a little bit. Back to waveguides!
Thanks augerpro for the work, I will follow and learn !
Two questions:
1) You seem to confirm that the best overall power response is not obtained with the closest ctc but with something around 1.2 to 1.4 wavelength (already discussed in other thread). Without this constraint, do you still think elliptical design is better than circular one ?
2) I remember reading kimmosto saying that a smaller waveguide than the midwoofer should give better overall response in XO region. Of course it was not a rule because waveguide size is one parameter among others but in your experiment with mw16tx, I will be curious if it would be possible to compare the 5inch and 6inch version of the t25 WG ?
Two questions:
1) You seem to confirm that the best overall power response is not obtained with the closest ctc but with something around 1.2 to 1.4 wavelength (already discussed in other thread). Without this constraint, do you still think elliptical design is better than circular one ?
2) I remember reading kimmosto saying that a smaller waveguide than the midwoofer should give better overall response in XO region. Of course it was not a rule because waveguide size is one parameter among others but in your experiment with mw16tx, I will be curious if it would be possible to compare the 5inch and 6inch version of the t25 WG ?
It it at the same frequency (@7300hz):
I'll be using the series notch for both these and the Satori TX. Even on the active versions I'll make this an option. If this breakup were not controlled and allowed to be produced as non-linear distortion I could see that being an issue for sure. But there are solutions. I mean, Revel uses these and glare isn't one of the adjectives I've ever heard in reviews of their stuff.
I don't think it is the 3khz bump you point out. Satori TX has it. Many Scanspeaks do, in fact most of the beloved ones are worse. It's good you narrowed it down to the woofer vs the tweeter, I'm not sure what the cause of the issue was for you. How do the off axis plots look, including vertical? Or maybe there is something else and your ears just prefer the Revelator.
valpo> I started using the elliptical because I got better performance early on. Now I design both and post whichever performs best on my website. CTC was never the reason I started with the elliptical mouth. But ellipticals still allow flexibility of placement. For example, with MTM I was modeling the other day it appears they still require a close ctc. And for this current project, the ABEC sim was best with the waveguide closer to the top edge, I could not have done this with a circular waveguide. I also like the slightly more directive vertical response of the elliptical. But really it it just comes down to which work best for a given tweeter. I do have to go back and retry circular for a few tweeters on my website, but the T25B and SB26 I already have.
That looks awesome! Now I remember seeing your post in ASR.
That is one expensive woofer thou...
What's the issue around 700Hz?
I find 500Hz issues way to significant for just a baffle resonance.
If you wanna do a passive crossover (which I would fine a bit of a compromise), go for 8 ohm.
Otherwise 4 ohm. Passive crossovers at 4 ohm are doable, but it's always a lot of stretching and pulling.
700hz issue? No issue there that I can see. I have a couple reasons for thinking the 500hz wiggle is panel resonance (all the panels, not just the baffle). It shows up in two different woofers, one with a neo magnet and the other with a ferrite, so probably not frame/magnet vibration. There is no port, so these are not port resonances. Now the ABEC model does show some bunching up of responses there, but not very strongly. I think the diffraction model suggests there might still be some bunching up there - see any other speaker on Erin's website - but I just don't think it is all explainable by diffraction. So that leaves the box itself, and in fact if you look at similar sized models on Erin's website, they almost all show an issue in the 400-1000hz region, except the ones that do some box panel damping. That aligns with my own testing of box materials. And really, it's not that bad relative to most of the speakers he's measured, it just stands out because the rest of the response is so good compared to those other speakers.
What do you mean by "stretching and pulling"?
Last edited:
The 5th graph has some issue going on on 700Hz.
edit: never mind, that's a different speaker LOL
It's just a lot of energy for just a little baffle vibration.
For some addition bracing in would probably move it.
I would like to see some impedance measurements as well as a burst decay waterfall.
They give a lot more insight about resonances, vibrations and interference issues.
What I mean with "stretching and pulling", is that it's tricky to not get below 3ohm with a passive filter.
To unfortunate that only data from 1kHz is shown for the tweeters.
My own testing shows panel resonances from typical build methods are almost never "just a little baffle vibration." They are very large in the 400-1000hz region, with peaks that can be as loud as the driver itself. Bracing is no guarantee to remove them either, see my testing: https://www.somasonus.net/box-construction-methods.
You likely won't see these issues in an impedance sweep, they have nothing to do with the air inside the box.
EDIT: well, nothing to do with standing waves in the air cavity. Simple sound transmission through the air of course puts energy into the panels.
You likely won't see these issues in an impedance sweep, they have nothing to do with the air inside the box.
EDIT: well, nothing to do with standing waves in the air cavity. Simple sound transmission through the air of course puts energy into the panels.
Last edited:
True, but than you already assume it's baffle vibrations.
An impedance sweep as well burst decay (waterfall in periods), will rule these things out.
Without doing these measurements it's a guessing game.
500Hz is around 34cm / 13.5 inch internal resonance.
Depending on how the damping material is being used and what kind, this can create issues.
edit; also from those test done at your website, we cannot see what we are exactly looking at.
Just some SPL measurements aren't enough to draw conclusions on such vibrations.
Ideally that's done with either an accelerometer, or a velocity sensor, or a laser measuring displacement.
Last edited:
I’ve been corresponding with augerpro showing him my data of 3 different midwoofers in the same box, 1. with the bunching up at 500Hz (no gasket)-pictured below and 2 other midwoofers without the bunching up (2. thick foam gasket under frame and 3. corrugated foam gasket)
And you’re right b_force the zoomed impedance trace was instructive.
In my case it was 3 different woofers mounted differently in the same box. So it’s the woofer/baffle interface that made the difference.
I’m 100% confident Augerpro have these sorted/eliminated by the time the design is sent to the contract manufacturer/ DIYer. He was the one who showed me how with his Monster Construction Shootout thread/website. 😀
And you’re right b_force the zoomed impedance trace was instructive.
In my case it was 3 different woofers mounted differently in the same box. So it’s the woofer/baffle interface that made the difference.
I’m 100% confident Augerpro have these sorted/eliminated by the time the design is sent to the contract manufacturer/ DIYer. He was the one who showed me how with his Monster Construction Shootout thread/website. 😀
Last edited:
If by "assume" you mean "logically deduced by eliminating causes which are not present", then yeah I guess I'm "assuming".
Regarding measuring panel radiation, I have no idea what an accelerometer squiggle sounds like. I do know what a certain frequency response sounds like. So those are the measurements I ran. If you can tell me how an accelerometer squiggle sounds, more power to you, but I suspect most other humans hear how I do. How are frequency response measurements not the sort of info telling you how the box radiation sounds?
Last edited:
@mainframe99 I've been making waveguides for the Scanspeak ring radiator, and making a Scanspeak speaker was something I was thinking about. Revelator was top of my list, but I don't see the coated ones like you have anymore.