That would depend on a couple of factors. Since you will now have a "crossover" (even if it is a single cap) your system sum may change. When you make the cap very large then it is only blocking low frequencies.
Put the driver info into the modeler, add a capacitor, and see what happens. Try 22uf to start.
What impedance data are you using?
Have you put a mod offset on any of the drivers to allow for voice coil offset? This will affect relative phase response and therefore how well or not the drivers frequency responses blend
Did you use manufacturer frequency response curves? If so ideally you'd apply baffle step and diffraction to get a realistic in box response
Have you put a mod offset on any of the drivers to allow for voice coil offset? This will affect relative phase response and therefore how well or not the drivers frequency responses blend
Did you use manufacturer frequency response curves? If so ideally you'd apply baffle step and diffraction to get a realistic in box response
As an electrical engineer you understand specifications can't be exceeded by too much before problems occur.
An important specification for loudspeakers is Xmax, linear excursion- how much one-way movement the cone/diaphragm/membrane can move before the voice coil starts leaving the magnetic gap. Past Xmax, distortion goes up rapidly.
Most low frequency drivers list Xmax, unfortunately few high frequency drivers do. A HF driver like yours probably has only about 1mm Xmax.
SPL (sound pressure level) must drop at 12dB per octave to have similar excursion:
A 12dB per octave crossover is a minimum requirement when running tweeters at higher levels.
I am a mechanical engineer, I have spent my entire working life with engineers of various professions, but no offense, I do not see an engineering approach in this work. 🙄
Not everyone appreciates what "constant acceleration" means. Ideal direct-radiators (no horns) are constant-acceleration devices.
What this means, in practical terms, is excursion increases at a rate of 12 dB/octave as frequency is lowered. It's unusual for any tweeter to move more than a millimeter before gross distortion sets in, followed by destruction. That's why a highpass filter is required, and 12 dB/octave is about the minimum slope that does the job.
Conversely, woofers and midranges have ugly breakup regions starting anywhere between 1 and 3 kHz. By "breakup" I mean the cone is moving in several directions at once, and a problem like that cannot be solved by equalization. It can only be avoided by lowpass filtering the woofer or midrange.
Modern crossovers also equalize the drivers so they have ideal rolloff slopes that are not ragged and uneven. This requires direct measurement of the drivers in the intended cabinet, since the cabinet shape alters the response.
What this means, in practical terms, is excursion increases at a rate of 12 dB/octave as frequency is lowered. It's unusual for any tweeter to move more than a millimeter before gross distortion sets in, followed by destruction. That's why a highpass filter is required, and 12 dB/octave is about the minimum slope that does the job.
Conversely, woofers and midranges have ugly breakup regions starting anywhere between 1 and 3 kHz. By "breakup" I mean the cone is moving in several directions at once, and a problem like that cannot be solved by equalization. It can only be avoided by lowpass filtering the woofer or midrange.
Modern crossovers also equalize the drivers so they have ideal rolloff slopes that are not ragged and uneven. This requires direct measurement of the drivers in the intended cabinet, since the cabinet shape alters the response.
Forget the xover for now you're starting with a bit of a mishmash of drivers, your mid is a small wide range driver it really doesn't need such a large dia tweeter as it will handle up around 5Khz, something like HiVi T20-8 3/4" would be a more suitable candidate.
The 1 1/8th" Daytons are more designed for bigger mids like a 6 or 8" that "beam" at higher frequencies and maybe suffer cone break up and ringing like the Dayton RS series of drivers with larger metal cones.
Also the bass driver could be bigger to, making small cones do low sub work often leads to more distortion. Id be inclined to run a 10"sub actively up a few 100Hz to reinforce the wideband mid.
The 1 1/8th" Daytons are more designed for bigger mids like a 6 or 8" that "beam" at higher frequencies and maybe suffer cone break up and ringing like the Dayton RS series of drivers with larger metal cones.
Also the bass driver could be bigger to, making small cones do low sub work often leads to more distortion. Id be inclined to run a 10"sub actively up a few 100Hz to reinforce the wideband mid.
Most tweeters come with a back chamber already fitted, so that gives them a mechanical 12dB/oct high pass. Without a capacitor, the typical power rating might be 3-5W, check the datasheet.
With a 5" woofer, I would definitely skip the mid-range altogether. It doesn't really add much in terms of total performance. Lots of 6"+ woofers are designed to hand over to a tweeter directly @ ~2.5kHz.
With 2 or more sound sources playing the same signal, there can be an issue of interference and comb filtering, where the phases are misaligned. Even in an 'optimal' case, where the speakers are perfectly in phase, it's often only in one particular direction where this occurs, like "on axis". Meanwhile, the room still gets sprayed with a complex mix of interference patterns at other angles because one speaker driver is physically further away than the other. A steeper crossover, say 24dB rather than 12dB / octave, can reduce the overlap, so for instance @ 4kHz the woofer is significantly attenuated and doesn't clash very much with the tweeter's output.
Other considerations: overlap and comb filtering is not necessarily all that bad. If well done, the speaker amplitude sums cleanly on-axis, while the off-axis response has reduced energy due to randomised power summing.
On the 3rd hand, steeper filters tend to kill break-up modes and reduce unwanted sound coming from speakers operating outisde of their preferred band. But you may be putting the cart before the horse, so to speak, because multiway speakers tend to sacrifice "out of band" sound quality based on assumptions that filters will be deployed.
The last thing that jumps out at me is the giant rubber surround of the 2" mid driver. This is likely bad, as the mechanical properties of those things are very questionable when they flex. A 2-way CHN-110 + CHN-50(p?) combo would be my bet for a multiway design with zero or ultra-minimal filtering.
With a 5" woofer, I would definitely skip the mid-range altogether. It doesn't really add much in terms of total performance. Lots of 6"+ woofers are designed to hand over to a tweeter directly @ ~2.5kHz.
With 2 or more sound sources playing the same signal, there can be an issue of interference and comb filtering, where the phases are misaligned. Even in an 'optimal' case, where the speakers are perfectly in phase, it's often only in one particular direction where this occurs, like "on axis". Meanwhile, the room still gets sprayed with a complex mix of interference patterns at other angles because one speaker driver is physically further away than the other. A steeper crossover, say 24dB rather than 12dB / octave, can reduce the overlap, so for instance @ 4kHz the woofer is significantly attenuated and doesn't clash very much with the tweeter's output.
Other considerations: overlap and comb filtering is not necessarily all that bad. If well done, the speaker amplitude sums cleanly on-axis, while the off-axis response has reduced energy due to randomised power summing.
On the 3rd hand, steeper filters tend to kill break-up modes and reduce unwanted sound coming from speakers operating outisde of their preferred band. But you may be putting the cart before the horse, so to speak, because multiway speakers tend to sacrifice "out of band" sound quality based on assumptions that filters will be deployed.
The last thing that jumps out at me is the giant rubber surround of the 2" mid driver. This is likely bad, as the mechanical properties of those things are very questionable when they flex. A 2-way CHN-110 + CHN-50(p?) combo would be my bet for a multiway design with zero or ultra-minimal filtering.
Last edited:
I see you point. I will look into getting a larger mid/sub woofer for my low rather than run out that little guy too far.
I have little knowledge on how to balance out highs. I see I will need a 2nd order Xover minimum.
Question: I see some of these high go all the way up to 40K. Is this something to seek out? I see the sims do not range out that high so their behavior on that end will be a mystery to me
An alternative to a protection cap is a pad/series resistor and coil to ground, it has the advantage of being effective at Fs.
Just out of curiosity, why did you want to get away without crossover and not pick full range drivers (they of course need some EQ too…).
40Khz no human can hear it so no need for it, at best its 20Khz when you're young... I'm in the 50's with hereditary and industrial hearing loss, lucky get above 12Khz and that's drowned out by tinnitus anyway.
Some say we need over 20Khz it adds to the detail I call adds to the snake oil like 5Hp stickers on a car
Some say we need over 20Khz it adds to the detail I call adds to the snake oil like 5Hp stickers on a car
In image you put here:
https://www.diyaudio.com/community/attachments/xsim-3-way-dual-woofer-png.1243244/
The speaker symbols appear with an exclamation mark, which means that their specific files were not loaded and the simulation was made with the models idealized for them and not with the real models.
I think I don't have to explain what the consequences are. The simulation result is completely wrong. **** in, **** out.
https://www.diyaudio.com/community/attachments/xsim-3-way-dual-woofer-png.1243244/
The speaker symbols appear with an exclamation mark, which means that their specific files were not loaded and the simulation was made with the models idealized for them and not with the real models.
I think I don't have to explain what the consequences are. The simulation result is completely wrong. **** in, **** out.
The full range I've looked at do not seem to be actually full range. They would likely be find for classical, rock, or jazz. I listen to that but I also listen to a lot of modern electric pop. That type of music needs the dynamics lower in the range. Something a full range is unable to play
I cna hear down to 21 Hz, I have tested this. Sorry to hear that your hearing is so poor
Possibly, I like to feel it. The sub in my car gets super low. Its mounted to the chassis of the vehicle so you can feel the music through your body. It is quite the experience. I don't need to hear it to experience it.
Even people who are totally deaf can easily feel sound at 21 Hz.
Sorry to see that you don't understand SI prefixes. You may want to study those, they get used pretty regularly in the engineering fields.