It's not that simple, it also and largely is related to the cone depth, shallow cones beam less than deep cones for the same diameter. So your way of determining it is just not true. But it is true that cone diameter is a factor.
I'm thinking about a waveguid for the tweeter to match directivity, that is a good point.
It's that i hear the crossover where the drivers are not perfectly matched in phase. You can try to phase align but to do this over the whole region where they cross is almost impossible. And the lack of crossover in that passband is for me less a compromise than the mostly not perfectly flat response or beaming of good fullrange drivers like those from Mark Audio (as they do it less than most).
But what i try to do with this project is to fix that with a 3 way with crossover out of that passband. I did multiway speakers, and i even have an other 3 way in the planning that is more conventional on crossover points. But for this crosover in the passband is not an option at all, the point of this project is not to do that. It's a challenge that will take probally a lot of time and trial and error to get there, but it's what i want to do.
Alright I feel you. Do you have possibility to go out and listen some Danley multiple entry horn products? There might be a night club, fellow hobbyist or pro audio store with them. I've never heard one, but this is the design that I would check out with the design goals you have. Multiway point source with control on the beaming.
I mean, the more you think about why the problems exist you are trying to avoid it is always due to wavelengths and physical size of the drivers. Phase mismatch to some angles because of multiple drivers being different size and location. Multiple entry horn like Danley products conveniently offers the most sophisticated solution I've come across, a rare occasion where the physical dimensions don't prevent overlap and the result is full range point source multiway speaker. By the design, dispersion is quite narrow but other compromises seem to be more in the cost and size than audio quality compared to coaxial speakers for example where there seem to be some problems around the xo, both drivers are a bit of compromise due to the other being there. Single fullrange driver is of course the simplest thing but it can't compete with multiway speaker on any other aspects than being point source and a lot cheaper. MEH is point source and enjoys any benefits multiway system might bring, but the cost is high. My definition of cost includes money, time, complexity, size, weight, opinions of the family etc.
Since all kinds of compromises are inevitable we always want to push the trade-offs from audio quality to cost to reach the ultimate maxima audio performance don't we?😀 My two favorite subjects on single post, trade-offs and physical dimensions 😀 nothing more to say, gotta go sleeping.
Anyway, will follow what solution you come up with.
I mean, the more you think about why the problems exist you are trying to avoid it is always due to wavelengths and physical size of the drivers. Phase mismatch to some angles because of multiple drivers being different size and location. Multiple entry horn like Danley products conveniently offers the most sophisticated solution I've come across, a rare occasion where the physical dimensions don't prevent overlap and the result is full range point source multiway speaker. By the design, dispersion is quite narrow but other compromises seem to be more in the cost and size than audio quality compared to coaxial speakers for example where there seem to be some problems around the xo, both drivers are a bit of compromise due to the other being there. Single fullrange driver is of course the simplest thing but it can't compete with multiway speaker on any other aspects than being point source and a lot cheaper. MEH is point source and enjoys any benefits multiway system might bring, but the cost is high. My definition of cost includes money, time, complexity, size, weight, opinions of the family etc.
Since all kinds of compromises are inevitable we always want to push the trade-offs from audio quality to cost to reach the ultimate maxima audio performance don't we?😀 My two favorite subjects on single post, trade-offs and physical dimensions 😀 nothing more to say, gotta go sleeping.
Anyway, will follow what solution you come up with.
Last edited:
Thought a bit more on the bed. If I had similar goal in mind I would start with two different size full range drivers, big and small, and tried crossover where they still are combining as point source, less than 1/ 4 wl spacing. I think it would sound great, be pretty simple, cheap and easy overall. You'll find out if the sizes are right to achieve your goals. Add tweeter afterwards if needed, and a sub.
I had 5" fullrange speakers for many years and loved it. Nowadays not gonna happen, no opportunity to have single listening spot. Loving big multiway speaker currently. Yeah mono 😀 alright back to sleep
I had 5" fullrange speakers for many years and loved it. Nowadays not gonna happen, no opportunity to have single listening spot. Loving big multiway speaker currently. Yeah mono 😀 alright back to sleep
Last edited:
Nope.
It's almost exclusively related to diameter (and by diameter I generally mean from the top of the surround (surround mid-point) from one side to the opposite side, NOT from frame edge to edge.)
What you can alter is the overall dispersion pattern, or how it beams. A common trick here is "spreading-out" the dispersion of a driver with a rising high freq. response with a phase-plug (or a dust cap that acts more like a phase plug).
..and a semi-exception to this is a decoupled dustcap or whizzer (..and the whizzer variety often also makes use of a phase plug - though by the time you extend out to 45 degrees or more the response typically drops like a stone at higher freq.s.)
Last edited:
The TC7FD04 is a small efficient cone (considering it's bandwidth).
Subtract 10.5 db to get its 1 meter result.
Subtract 10.5 db to get its 1 meter result.
Attachments
Last edited:
You want 88dB, so that severely limits your options. You are almost in prosound output territory for a single mid.
TB W4-1757S has WIDE dispersion due to flat geometry.
SS Discovery 15W has a phase plug and higher output.
I wonder if the SEAS U16 would give you enough output. Great mids!
Wolf
TB W4-1757S has WIDE dispersion due to flat geometry.
SS Discovery 15W has a phase plug and higher output.
I wonder if the SEAS U16 would give you enough output. Great mids!
Wolf
Nope.
The shape of the cone, the surround, the nature of the drivers chaotic region behaviour all play significant roles in dispersion.
The flat-pistom model is VERY limited.
dave
(dave)
Go back and look through Hi Fi Compass's & Tim's measurements..
A distributed mode driver can also be an exception.
Again, though what you are looking for is where it BEGINS to narrow in dispersion, NOT the overall result or "character" of that dispersion.
Go back and look through Hi Fi Compass's & Tim's measurements..
A distributed mode driver can also be an exception.
Again, though what you are looking for is where it BEGINS to narrow in dispersion, NOT the overall result or "character" of that dispersion.
Last edited:
Yeah, the Tectonic stuff does have good off axis response, but they aren't sensitive enough for the OP's application.
Wolf
Wolf
It's off-axis is excellent despite starting to "beam" at about 2.8 kHz (..though the graph doesn't do a great job of showing spl differences, it also isn't showing the typical narrow dip around 2 kHz that the driver has..):
Attachments
Last edited:
Those two are among the recommendations I would give. I guess it’s worth mentioning, if you know someone with a set of Bowers & Wilkins 800 line speakers they could order a set of replacement midranges for you. They were about $200-325 USD each depending on the model within the 800 line.
My thoughts are if you want hifi drivers:
SB acoustics 4" models. Maybe the PAC, NAC or CAC.
Or if you want pro drivers:
Faitalpro 3Fe22 or 4fe32.
Oon
SB acoustics 4" models. Maybe the PAC, NAC or CAC.
Or if you want pro drivers:
Faitalpro 3Fe22 or 4fe32.
Oon
On a slightly different note, if you are using the 2 × 8" to provide the higher sensitivity for bass, it might be worthwhile to model the baffle response. The higher sensitivity of the woofer will become the baffle step compensation.
If using the 4" SBA, the have a drastic drop of about 6-10 db at 4-5 kHz. I would target that as a crossover point. Maybe second order filter starting at about 5kHz will mean the filter kicks in a bit out of the crossover range so less impact on the signal at the crossover range.
This project by Troels might be something to look for in inspiration.
SBAcoustics-3WC
Another consideration is have considered co- axial?
Oon
If using the 4" SBA, the have a drastic drop of about 6-10 db at 4-5 kHz. I would target that as a crossover point. Maybe second order filter starting at about 5kHz will mean the filter kicks in a bit out of the crossover range so less impact on the signal at the crossover range.
This project by Troels might be something to look for in inspiration.
SBAcoustics-3WC
Another consideration is have considered co- axial?
Oon
Last edited:
Apparently in some people's use of this driver, the dip is not present (Moosespeaker's Exponents). Some that have had it, the dip disappeared off-axis (Curt's Mavericks).
In my measurements of the various models of TB flats' impedances, there is typically an energy storage between 1.2k and 3.2k that shows as a 3-4 ohm magnitude higher than the nominal impedance.
It really depends on the batch I guess.
Wolf
Tectonic adverts transducers that widen the dispersion shorter wavelenghts than the physical dimensions, worth to check out what it is all about. Home - Tectonic | Tectonic
About sensitivites, for mid someting like 86dB nominal sensitivity sould be ok. Woofers will need baffle loss compensation, like oon_the_kid said, and tweeter is always easy to tame down. My Avalanche project has 2x8" 4R in series, but dsp and multi-amping make gain control piece of cake. Dedicated midranges typically have several dB more sensitivity than fullranges/midwoofer versions of same series, at the expence of Xmax
Nominal sensitivity is often misleading, eg. Tangband tells sensitivity at peak of response chart. Same with many PA mids.
For really high sensitivity design, drivers must be PA stuff. Then bass extension of woofers might be a challenge for hifi, unless subwoofers are used anyway. But a well designed 3-way doesn't need subwoofer for music!
Nominal sensitivity is often misleading, eg. Tangband tells sensitivity at peak of response chart. Same with many PA mids.
For really high sensitivity design, drivers must be PA stuff. Then bass extension of woofers might be a challenge for hifi, unless subwoofers are used anyway. But a well designed 3-way doesn't need subwoofer for music!
That is also a good tip, that and some other SB 4" drivers show good behaviour to about 4 to 6kHz. I'll study those more. And many of those are relative cheap to buy down here.
Coaxial has other issues in the crossover range, and that is mostly also in the frequency range i don't want a crossover. I did look at it, but it's not a solution for my case.