I wanted to share a project and I hope this is the right place. Anyways here it is.
Below is this first and current iteration
These are 60x120 oblate spheroid horns that are about 17" wide. Most of the info I got was on this forum from patrickbateman. The taps are 3/4" 2 per driver and the chambers are 15mm deep plus frustums. The throat of the tweeter is about 2" deep which extended the tweeter response to around 750 if I remember correctly. The Soundstage is the largest I've heard which is exactly what I wanted for 2 channel movies and music. The main problem with these is that the way I attached the throat caused nasty vibes between the petg parts and I never got around to fixing it. I eq most of it out but certainly not ideal. Inevitably I knew I wanted bigger horns anyways which brings me to the next version
The main differences are the mid chambers are only 6mm deep and obviously the taps are oblong 3×.625. And the phase plug is based off of the new b&c drivers. Oh and its 27.5" (anyone wanna tell me how low this should load the mids?) Let me know what you think. Right now im wondering Are the frustrums a good idea for the taps or does this cause more group delay? Thanks in advance. 👍
Below is this first and current iteration
These are 60x120 oblate spheroid horns that are about 17" wide. Most of the info I got was on this forum from patrickbateman. The taps are 3/4" 2 per driver and the chambers are 15mm deep plus frustums. The throat of the tweeter is about 2" deep which extended the tweeter response to around 750 if I remember correctly. The Soundstage is the largest I've heard which is exactly what I wanted for 2 channel movies and music. The main problem with these is that the way I attached the throat caused nasty vibes between the petg parts and I never got around to fixing it. I eq most of it out but certainly not ideal. Inevitably I knew I wanted bigger horns anyways which brings me to the next version
The main differences are the mid chambers are only 6mm deep and obviously the taps are oblong 3×.625. And the phase plug is based off of the new b&c drivers. Oh and its 27.5" (anyone wanna tell me how low this should load the mids?) Let me know what you think. Right now im wondering Are the frustrums a good idea for the taps or does this cause more group delay? Thanks in advance. 👍
I want to hear what the originals sound like without the resonance before I commit to $200 in filament. I thought I'd take the opportunity to try a novel idea at making a phase plug. Essentially I'm trying to create a plane wave without an abrupt transition.
It's a 180* turn over 2" length.
It's a 180* turn over 2" length.
Last edited:
A larger horn will have pattern control to a lower frequency, but since the expansion is basically conical it does not have a specific cut off, or "loading" like an exponential horn would have.
The mids "loading" would be about the same as the 17" wide horn, don't expect it to go lower.
The enclosed volume between the cone and the cone filler creates an acoustic low pass band filter, the larger the enclosed volume, the lower in frequency the HP occurs.
Generally in a two-way design you want the mids to go as high as possible, so the frustrum volume should be as little as possible with the excursion required for the SPL desired.
Can't tell what the frustrums in your existing horns look like, so can't tell how they compare.
Looks like it has a lot of reflective area on the diaphragm side axial twist "X" which will cause multiple arrival times at the exit. You might want to file them to a point.
B&C's HLX™ phase plug used on the DH350's 36mm diaphragm has three twisted central channels to gain the length required to match it's outer channels. The exit 1" diameter including both channels is a reduction from the diaphragm size.
Your throat extender appears to have no outer channel:
It's axial twist will make it's path longer than it's 2" length, increasing the LF loading of the HF driver, but what it will do to it's phase response won't be clear without frequency and polar response testing.
Your 25mm tweeter only has .667 Xmax, and a 1400Hz Fs, so the increase in SPL down low may help..
Art
Last edited:
Thank you for the detailed response!
Exactly. I want as much flexibility in the crossover as possible. My concern has always been the massive spike in GD at the bandpass frequency. I would like to avoid it all together with steep filters. This is the one thing I can't understand about these designs is why 2 or 3 ms of group delay at crucial frequencies is acceptable. What's your take on this?
Also, which do you think is better frustums or straight ports?
You're right about this. The other model I created does. I just wanted to try this one since I had modeled it a while back and needed a new throat to fix the horns. The new ones I made still differ from the b&c because it doesn't taper down to a smaller throat but expands to the horn. Honestly my main objective was to create a plane wave with as little interference as possible. Filing it to points is a smart idea.
Exactly. I want as much flexibility in the crossover as possible. My concern has always been the massive spike in GD at the bandpass frequency. I would like to avoid it all together with steep filters. This is the one thing I can't understand about these designs is why 2 or 3 ms of group delay at crucial frequencies is acceptable. What's your take on this?
Also, which do you think is better frustums or straight ports?
That won't happen, the steeper a driver, or horn/driver, or crossover filter’s roll off, the more phase change per frequency, hence more GD.
Group delay is the negative derivative in milliseconds of the phase response.
It is a measure of the rate of change of phase with respect to frequency, and is positive when the phase slope is negative.
FIR filters are capable of flat phase response, though still require delay for time alignment of drivers with different acoustical points of origin.
My take is phase and frequency response should be smooth through the crossover region.
Danley's three-way SH-50 MEH has ~360 degrees of smooth phase shift through the decade between 100 and 10kHz, what would be expected from a pair of 2nd order acoustical crossover points.
My personal experience is I can not hear a difference between flat (0 degree) phase response, and 720 degrees of smoothly changing phase response as is introduced by two 4th order crossovers in a 3-way system.
Whichever results in the smoothest phase and frequency response in the acoustical crossover region is "better".
That said, a very high aspect ratio cylindrical or pyramidal frustrum would acoustically behave the same as a straight port.
Whatever shape used, rounding the hard edge reduces diffraction (whether audible or not..) and makes the hole easier to paint 😉 .
Art
I think you can incorporate time alignment into fir filters. I always align ir to start and do a simple minphase inversion in rew. With fir I think you can use brick wall limits built into the fir itself for crossovers. Atleast this is what i plan to try when I upgrade my minidsp to the new Scarlett interface being release this month.
So i ran some measurements with the new throat adaptor and this is the response (everything i measure is at LP, honestly its all i care about) btw i have zero room treatments other than the rug.
this is the tweete,r which i currently cross at 1550hz. The helix phase plug actually extended my highs which where rolling off at like 13k without a proper phase plug.
I did quick eq and crossover on the system and these are the measurements i got at LP. Nasty 3rd harmonic at 250hz. Still getting vibrations from the mids in the horn. its improved drastically but still an issue.
idk much about phase but the crossovers are in the middle of the slopes at 50hz 150hz and 1550hz.
I'm new at most of this stuff so I couldn't tell you if these measurements are very good or not but i thought i would share what i got. I think sorting out the vibration is going to be the main goal for now even if i can't hear it after eq.
this is the tweete,r which i currently cross at 1550hz. The helix phase plug actually extended my highs which where rolling off at like 13k without a proper phase plug.
I did quick eq and crossover on the system and these are the measurements i got at LP. Nasty 3rd harmonic at 250hz. Still getting vibrations from the mids in the horn. its improved drastically but still an issue.
idk much about phase but the crossovers are in the middle of the slopes at 50hz 150hz and 1550hz.
I'm new at most of this stuff so I couldn't tell you if these measurements are very good or not but i thought i would share what i got. I think sorting out the vibration is going to be the main goal for now even if i can't hear it after eq.
Last edited:
You are measuring the room response, so hard to tell what the enclosure is doing.
25 dB drop from 2kHz to 10kHz. That's why compression drivers are used for high frequency 😉
If the REW delay timing were set correctly, you should only see ~360 degrees of wrap for each 24dB of crossover filter slope used.
I'll look into this. The only timing reference I use is the acoustical one that sounds at either end of the sweep.
Funny I guess I didn't really think about how it's just a room measurement. I'm gonna do some research on proper measurements. I think I saw something on audioexpress I might check out. You have any references?
The REW help manual itself is a helpful reference.
There are a number of YouTube videos on various subjects regarding measurements.
SynAudCon (Synergistic Audio Concepts) has a library of articles regarding all aspects of sound, 11 pages of titles under Tests and Measurements alone.
https://www.prosoundtraining.com/category/synaudcon-library/
There are a number of YouTube videos on various subjects regarding measurements.
SynAudCon (Synergistic Audio Concepts) has a library of articles regarding all aspects of sound, 11 pages of titles under Tests and Measurements alone.
https://www.prosoundtraining.com/category/synaudcon-library/
Ive been having second thoughts about printing such a large version of this for two reasons. 1 i cant seem to figure out what is the resonance in the horns. Ive gotten rid of most of it and the rest i had to completely highpass. And 2 if making a larger horn doesnt do anything to horn load the midrange frequencies then its not much use to me. Maybe it would provide a larger sense of scale.
As far as the resonance goes ive tried to fill the horn with foam and that didnt do anything. I printed this horn with 1.2mm nozzle 4 layers thick and used the lightning infill. The horn is extremely solid imo and heavy. The last thing i can think to try is put a tpu gasket between the driver and either side of the enclosure because thats the only other part i can think of. The next thing im thinking is the driver choice. The silver flutes have low qts but they are extremely compliant and i wonder if this is the cause. But This is happening at around 200-300hz so idk. I put this on hold and worked on my next upgrade which was the lackluster bass/midbass.
Here are pix i snapped quickly before work this morning at 2am.
I must say this is a drastic improvement. It sounds like live drums in my room!
As far as the resonance goes ive tried to fill the horn with foam and that didnt do anything. I printed this horn with 1.2mm nozzle 4 layers thick and used the lightning infill. The horn is extremely solid imo and heavy. The last thing i can think to try is put a tpu gasket between the driver and either side of the enclosure because thats the only other part i can think of. The next thing im thinking is the driver choice. The silver flutes have low qts but they are extremely compliant and i wonder if this is the cause. But This is happening at around 200-300hz so idk. I put this on hold and worked on my next upgrade which was the lackluster bass/midbass.
Here are pix i snapped quickly before work this morning at 2am.
I must say this is a drastic improvement. It sounds like live drums in my room!
So i removed the rear chambers for the midrange speakers and that eliminated whatever buzzing i heard. Im not entirely sure what caused it. Its either the fact that they were very small enclosures for these high vas woofers or the wires were rattling inside. Its such a tight fit in there i couldnt really figure a way to secure them. Now the only distortion i hear is the ports themselves. Theyre 3/4 inch ports 2 per driver. I currently have them crossed at 150 which is where i like them to be since i like alot of boost below that point anyways. Now i think the original design has merit in that the port area is much larger.
I agree with everything that Art said.
A couple of things I'd add:
There are only two obvious benefits to the HLX phase plug, as far as I can see:
1) That phase plug allows B&C to make their compression driver smaller. This is really great for a company where a lot of their compression drivers wind up in arrays; every company selling arrays is trying to figure out how to make them smaller, lighter and louder. Years ago, EAW used a custom version of a Celestion compression driver for their "Anya" array. Celestion shaved about two centimeters off the height. That doesn't sound like a lot, but when you're trying to get LOUD, being able to pack the tweeters tighter improves your power handling (of the array overall), the output and the high frequency polars.
2) There's probably a profit motive too; I have some of the tiny compression drivers from BMS, Celestion and B&C. The BMS are beefy and expensive. The B&C is very nice, but there's lots of plastic in the design and you can tell by looking a them, that the BMS is cheaper to manufacture. I'm not saying that one is better than the other, just that reducing the size can help increase margins and that probably plays a factor here.
Creating a plane wave is fairly straightforward; just equalize the pathlengths in the phase plug. I made an article about that here: https://www.diyaudio.com/community/threads/diy-compression-drivers.317125/
There is no "hard and fast" rule that compression drivers have to produce a plane wave. The shape of the wavefront at the exit of a compression driver should simply match the angle of the entrance to the waveguide. IE, if you have one of those old school compression drivers from the 70s or 80s, with a very narrow exit angle, and you bolt them onto a waveguide with a wide entrance at the throat, you're going to have problems.
If you're just having fun building phase plugs, that's great. I did that, it's educational.
If you're trying to build the best speaker possible, and depth of the speaker isn't an issue, I don't see how the HLX phase plug contributes to that.
After putting dome tweeters on waveguides and building DIY compression drivers out of dome tweeters, I've mostly just gone back to plain ol' conventional compression drivers. They just work.
A couple of things I'd add:
There are only two obvious benefits to the HLX phase plug, as far as I can see:
1) That phase plug allows B&C to make their compression driver smaller. This is really great for a company where a lot of their compression drivers wind up in arrays; every company selling arrays is trying to figure out how to make them smaller, lighter and louder. Years ago, EAW used a custom version of a Celestion compression driver for their "Anya" array. Celestion shaved about two centimeters off the height. That doesn't sound like a lot, but when you're trying to get LOUD, being able to pack the tweeters tighter improves your power handling (of the array overall), the output and the high frequency polars.
2) There's probably a profit motive too; I have some of the tiny compression drivers from BMS, Celestion and B&C. The BMS are beefy and expensive. The B&C is very nice, but there's lots of plastic in the design and you can tell by looking a them, that the BMS is cheaper to manufacture. I'm not saying that one is better than the other, just that reducing the size can help increase margins and that probably plays a factor here.
Creating a plane wave is fairly straightforward; just equalize the pathlengths in the phase plug. I made an article about that here: https://www.diyaudio.com/community/threads/diy-compression-drivers.317125/
There is no "hard and fast" rule that compression drivers have to produce a plane wave. The shape of the wavefront at the exit of a compression driver should simply match the angle of the entrance to the waveguide. IE, if you have one of those old school compression drivers from the 70s or 80s, with a very narrow exit angle, and you bolt them onto a waveguide with a wide entrance at the throat, you're going to have problems.
If you're just having fun building phase plugs, that's great. I did that, it's educational.
If you're trying to build the best speaker possible, and depth of the speaker isn't an issue, I don't see how the HLX phase plug contributes to that.
After putting dome tweeters on waveguides and building DIY compression drivers out of dome tweeters, I've mostly just gone back to plain ol' conventional compression drivers. They just work.
Thanks for the input. The idea behind spiral throat i made helps by reducing the narrowest pathlength down the 2 inch deep throat so it extends the response in both directions. For $13 Im pretty happy with it. Btw I remember reading about your 12$ newark mids making some port noise in your metlako design? Irc you thought this was due to the high q of the driver. Did you ever resolve the issue?
Im not surprised. They give alot of output and bandwidth. My stereo is just never loud enough lol. I want visceral. I just bought a pair of eminence nsd2005-16 open box for $90 shipped on ebay. These are my first compression drivers. Havent measured them yet but supposedly has an fs of 580hz. Ive got a metamaterial enclosure on the printbed right now. Ive got the day off so i might post something on it later.
This is my first attempt. I just used omni calculater to find 580hz length then divided by 4 and got 6 inches. Its very long and idk how it will affect it yet. Also idk how it will differ with a horn attached. Honestly i think i should have printed an open back retainer first and went from there but im adhd. Squirrel! Lol
factory rear chamber, no horn
6" long metamaterial chamber, no horn
let me know what you think. the only drawback i see is the physical size of it.
6" long metamaterial chamber, no horn
let me know what you think. the only drawback i see is the physical size of it.