Yes probably but you can use a Linkwitz Transform to taylor it to something ( hopefully) usable for comparison?
Yeah, that can make the direct comparison difficult. I haven't thought this through enough I guess. Well, it's up to you, it's really not necessary.
Probably the enclosure volume is only going to affect the low-frequency alignment, I guess.
I think, above a few hundred Hz, the baffle-box dimensions might impact the response more than the low-frequency alignment.
Anyway, I will make an attempt to measure today if possible.. 🙂
I think, above a few hundred Hz, the baffle-box dimensions might impact the response more than the low-frequency alignment.
Anyway, I will make an attempt to measure today if possible.. 🙂
Just compare to your teardrop enclosure, in the region where you could operate the coax it will be close enough.
Thanks @fluid 🙂
Here is the comparison between the Sica Coax response in a closed teardrop-shaped enclosure and the current cardioid enclosure (Plot-1)
Please ignore the legends above. The dotted like is the closed box teardrop enclosure and the solid line is the cardioid one. Since the measurements of the driver in these two enclosures were made at different input voltage levels, I offset one of them so that they match around 1kHz & 2kHz.
Just for reference, here are the horizontal polars of the driver in cardioid enclosure (Plot-2)
I see some interesting things above... 😀
1) From Plot-1 It looks like the cardioid enclsoure has some frequency response gain on axis vs the monopole enclosure as @mabat had shown in simulations
2) Comparing plot-1 & plot-2 There are some resonances around 800Hz & 1100Hz. Atleast one of these could have caused the rattling sound I could hear while measuring the speaker I think
3) There is an on-axis dip, whose effect reduces at off axis angles around 1350Hz. I dont know what this is. Maybe diffraction?
But the diffraction response of the driver on this baffle looks like this:
4) From Plot-1. Beyond 3kHz, the response from the cardioid enclosure rolls off naturally like a low pass filter has been applied compared to the sealed enclosure. Is this because of the leaking ports causing some kind of acoustic low pass filter at these frequencies?
Here is the comparison between the Sica Coax response in a closed teardrop-shaped enclosure and the current cardioid enclosure (Plot-1)
Please ignore the legends above. The dotted like is the closed box teardrop enclosure and the solid line is the cardioid one. Since the measurements of the driver in these two enclosures were made at different input voltage levels, I offset one of them so that they match around 1kHz & 2kHz.
Just for reference, here are the horizontal polars of the driver in cardioid enclosure (Plot-2)
I see some interesting things above... 😀
1) From Plot-1 It looks like the cardioid enclsoure has some frequency response gain on axis vs the monopole enclosure as @mabat had shown in simulations
2) Comparing plot-1 & plot-2 There are some resonances around 800Hz & 1100Hz. Atleast one of these could have caused the rattling sound I could hear while measuring the speaker I think
3) There is an on-axis dip, whose effect reduces at off axis angles around 1350Hz. I dont know what this is. Maybe diffraction?
But the diffraction response of the driver on this baffle looks like this:
4) From Plot-1. Beyond 3kHz, the response from the cardioid enclosure rolls off naturally like a low pass filter has been applied compared to the sealed enclosure. Is this because of the leaking ports causing some kind of acoustic low pass filter at these frequencies?
Too many questions 🙂
I believe that the resonances around 1 kHz are caused by the flimsy board construction and should disappear with a more solid one.
It's only a pitty that the voltages were not the same. But I already like the "cardioid" results.
As for the difference at HF, I don't understand that.
I believe that the resonances around 1 kHz are caused by the flimsy board construction and should disappear with a more solid one.
It's only a pitty that the voltages were not the same. But I already like the "cardioid" results.
As for the difference at HF, I don't understand that.
The same pattern that gives the extra 6dB rolloff at the low end is present at the high end in the simulation, so it makes sense to me. Look where the predicted null frequency begins
The sim diff to measued in HR is something i have seen before. In my understanding this is due to the cone shape. The sim appears to use a flat plate. Not sure though.
It's not at all clear to me how could radiation from the port affect the on-axis response so that is 10 dB lower at 10 kHz compared to the monopole, given the directivity of the diaphragm at those frequencies and the amount of damping in the rear chamber. That's a very different situation compared to that on LF, where there's almost no directivity and very little effective damping.
Last edited:
Here's a simulation with and without the opening (i.e. a cardioid vs closed), at 0, 90 and 180 deg:
green = open, blue = closed
Obviously, at HF the on-axis responses should converge. There must be some other effect than the enclosure.
(To simulate this, i.e. to remove the opening temporarily, simply rename the "Port" section in the Ath script, e.g. to "_Port"...)
green = open, blue = closed
Obviously, at HF the on-axis responses should converge. There must be some other effect than the enclosure.
(To simulate this, i.e. to remove the opening temporarily, simply rename the "Port" section in the Ath script, e.g. to "_Port"...)
I think I might have pasted the wrong figure earlier for the teardrop box and caused all this confusion. Sorry for that..
Initially, when I measured the teardrop box, I used a Dayton EMM-6 Mic which I later found as having issues at frequencies above 4kHz or so.
All my current measurements are taken using a Sonarworks soundID mic.
I had recalibrated the Dayton mic based on my Sonarworks mic & another minidsp mic later. I had missed comparing the cardioid box response with this recalibrated measurement.
Anyway here is the comparison between teardrop vs cardioid box for this driver.
Please ignore the light blue dotted line.
The dark blue dotted line is the teardrop enclosure farfield merged with nearfield measurement. So it is only as good as how accurate the merging was, roughly below 300-400Hz.
The brown solid curve is the cardioid enclosure response (Please note that this is only a 5.1ms gated measurement. So curve below roughly 200Hz is just VituixCAD interpolation)
Now the higher frequencies in both cardioid box & teardrop box match to a good extent.. 😀
Initially, when I measured the teardrop box, I used a Dayton EMM-6 Mic which I later found as having issues at frequencies above 4kHz or so.
All my current measurements are taken using a Sonarworks soundID mic.
I had recalibrated the Dayton mic based on my Sonarworks mic & another minidsp mic later. I had missed comparing the cardioid box response with this recalibrated measurement.
Anyway here is the comparison between teardrop vs cardioid box for this driver.
Please ignore the light blue dotted line.
The dark blue dotted line is the teardrop enclosure farfield merged with nearfield measurement. So it is only as good as how accurate the merging was, roughly below 300-400Hz.
The brown solid curve is the cardioid enclosure response (Please note that this is only a 5.1ms gated measurement. So curve below roughly 200Hz is just VituixCAD interpolation)
Now the higher frequencies in both cardioid box & teardrop box match to a good extent.. 😀
Perfect! Now it makes sense.
I like how the dipole configuration essentially eliminates any baffle-step issues through the midrange, making the response almost a perfect band-pass.
I like how the dipole configuration essentially eliminates any baffle-step issues through the midrange, making the response almost a perfect band-pass.
Last edited:
Now, I am looking at making this into a full speaker. The coax driver will work in the leaky box from somewhere above 200Hz & woofer/mid-woofer handling the range of frequencies below that. Since I want to make this speaker as compact as possible, especially in the depth dimension (the prototype enclosure was just 17cm in depth), the maximum volume that I will be able to give the driver will be about 6ish litres.
I have the Sica 5H5.5CP mid-woofer drivers with me, which have been used in these builds:
https://www.donhighend.de/?page_id=6812
https://www.der-akustische-untergrund.de/speakers/intermediate-projects/harris-mk-ii/
and the commercial offering from Sigberg Audio (It uses the same combination of coax + mid-woofer that I have):
https://www.sigbergaudio.no/products/sigberg-audio-sbs-1-aktive-hoyttalere
I tried a closed box alignment for the woofer with 6L volume.
It looks like this:
I even tried a bass reflex alignment with same box volume:
The SPL picture looks nicer for bass reflex.
But I am unable to decide on which one to proceed with for this woofer. I have a general aversion to bass reflex just because I had seen the midrange leakage through a few ports once, both commercial & DIY. (Maybe they were just bad implementations)
In general, for a closer to wall spacing and a desktop-type application, is a closed box alignment better? (Maybe sometime in the future, I may have a compact subwoofer for this setup)
In the meanwhile, I drew a baffle (box picture in the making) and it looks like this 😀
(Top part holds the coax & bottom holds the midwoofer)
I have the Sica 5H5.5CP mid-woofer drivers with me, which have been used in these builds:
https://www.donhighend.de/?page_id=6812
https://www.der-akustische-untergrund.de/speakers/intermediate-projects/harris-mk-ii/
and the commercial offering from Sigberg Audio (It uses the same combination of coax + mid-woofer that I have):
https://www.sigbergaudio.no/products/sigberg-audio-sbs-1-aktive-hoyttalere
I tried a closed box alignment for the woofer with 6L volume.
It looks like this:
I even tried a bass reflex alignment with same box volume:
The SPL picture looks nicer for bass reflex.
But I am unable to decide on which one to proceed with for this woofer. I have a general aversion to bass reflex just because I had seen the midrange leakage through a few ports once, both commercial & DIY. (Maybe they were just bad implementations)
In general, for a closer to wall spacing and a desktop-type application, is a closed box alignment better? (Maybe sometime in the future, I may have a compact subwoofer for this setup)
In the meanwhile, I drew a baffle (box picture in the making) and it looks like this 😀
(Top part holds the coax & bottom holds the midwoofer)
Most small sized commercial speakers will feature a port because of the need to be small. The same goal you have in mind for your implementation. Neumann does an exceptional job at suppressing port resonance in their designs, they use a similar approach as kef do, a partially lossy/reactive port sidewall, but on this board I have seen one of the best implementations by @Kwesi:
1) https://www.diyaudio.com/community/...sb-acoustics-sb15nbac30-4.391310/post-7149349
2) https://www.diyaudio.com/community/...sb-acoustics-sb15nbac30-4.391310/post-7160288
In short: its manageable.
1) https://www.diyaudio.com/community/...sb-acoustics-sb15nbac30-4.391310/post-7149349
2) https://www.diyaudio.com/community/...sb-acoustics-sb15nbac30-4.391310/post-7160288
In short: its manageable.
That's a really great implementation by Kwesi - but it's still 20dB under main level ... that's not a lot. And 95% of reflex ports are significant worse.
So no - I'm not a fan of reflex when it goes into midrange. But when you don't excite these frequencies you are fine.
I still think it's a pitty to not build this as fullrange speaker. Even with the reflex version you will miss a lot, at leat 60Hz needs to be heared. When you listen at very low levels you can do closed and put some EQ in front - but then I would chose a long stroke chassis.
So no - I'm not a fan of reflex when it goes into midrange. But when you don't excite these frequencies you are fine.
I still think it's a pitty to not build this as fullrange speaker. Even with the reflex version you will miss a lot, at leat 60Hz needs to be heared. When you listen at very low levels you can do closed and put some EQ in front - but then I would chose a long stroke chassis.
@IamJF: If you know of any other (5-6 inches) longer stroke, reasonable cost woofer for this kind of small box (6-7L volume), could you please suggest some..?
I have been planning to use the Sica only because I have it in hand, and I saw it being used in the Sigberg audio speaker & some other DIY projects.
I have been planning to use the Sica only because I have it in hand, and I saw it being used in the Sigberg audio speaker & some other DIY projects.
That's to small for my usual projects, sorry. I have Seas ER18RNX here from a project which changed over time which can cope with small volume and has good sensitivity, medium Xmax and very low THD. But also doesn't get low and would need a good EQ boost.
In 8" there is SB23MFCL45-8 which sits on my table and waits for a test enclosure.
But as small is beautiful in these days ;-) there should be plenty of 5-6" chassis with good Xmax - maybe someone can give a tip?
In 8" there is SB23MFCL45-8 which sits on my table and waits for a test enclosure.
But as small is beautiful in these days ;-) there should be plenty of 5-6" chassis with good Xmax - maybe someone can give a tip?
Dayton E150HE sounds perfect if not too expensive. Great Klippel results https://audioxpress.com/article/tes...e-subwoofer-from-the-dayton-audio-epique-line