I've read here on the general knowledge of "half a wavelength away" at crossover and "the closer the better"
However...
When looking at the Herron Audio speaker at an audio show (they had a prototype) I asked Keith, the designer. The M/T spacing was a LOT more than the minimum possible. It was significantly spaced! So I asked.
The answer was a "specifically calculated distance". This makes me wonder if there is a formula or a guideline? And that distance may not be the minimum possible but depends on the crossover slopes.
Any thoughts on this?
However...
When looking at the Herron Audio speaker at an audio show (they had a prototype) I asked Keith, the designer. The M/T spacing was a LOT more than the minimum possible. It was significantly spaced! So I asked.
The answer was a "specifically calculated distance". This makes me wonder if there is a formula or a guideline? And that distance may not be the minimum possible but depends on the crossover slopes.
Any thoughts on this?
Go download a copy of Boxsim from the Visaton website. You can quickly mock up the front baffle for that speaker you saw. Place two drivers on there, quickly input the crossover frequency and slope and see what happens when you move the drivers around. The off axis vertical response will change. Farther the drivers are apart, the smaller angle will be required to get the same change. "calculated" likely means they looked good like that to the marketing department. Software | Visaton
One of the strangest decision in speaker building is the big distance between the tweeter and mid(woofer). I think it has a good reason why it looks ugly (because it's bad, haha).
I only see one possible benefit of the large distance, which is the lesser diffraction effect caused by other driver(s) near the tweeter (this is less of a problem with controlled directivity tweeters). But usually that diffraction effect is not a problem anyway imo.
I only see one possible benefit of the large distance, which is the lesser diffraction effect caused by other driver(s) near the tweeter (this is less of a problem with controlled directivity tweeters). But usually that diffraction effect is not a problem anyway imo.
Many older designs had large spacing, it may not be perfect but it can be ok. Slopes are not the answer and there is not a formula to give a perfect wide spaced option.
Hmm, don't know the math without drawing it to scale to figure a ratio, but for typical HIFI/HT room sizes and if speakers are spaced to be in front of the outboard seats and toe'd in to cross at some point in front of the 'sweet spot', the spacing need only be perceptually? close at one's ears. Even better if the HF is 'focussed' a bit for closer seating.
This works well with horns, but tweeters sometimes need to be a small [usually vertical] array to mimic a horn's pattern.
GM
This works well with horns, but tweeters sometimes need to be a small [usually vertical] array to mimic a horn's pattern.
GM
Yes, I would love to Hear the sound of TW with the woofer removed and replaced with Felt but the woofer is not removed🙂
Althou rarely possible getting the mid & tweeter centre to centre (in 3-space) should be less than a quarter wavelength. That makes the drivers essentially coincident.
dave
dave
Has anyone ever tried what happens if CTC is exactly one wavelength at the crossover frequency ?
Regards
Charles
Regards
Charles
There are a lot of good measurements of vertical directivity on audiosciencereview dot com, which are pretty telling of results of different spacing. Results are not hopeful...the absolute majority of speakers will have about 20 (maybe 30 at most) degree listening angle if they don't have a coaxial mid-tweeter. I wonder if it's that problematic in practice if one is sitting in a sweet spot, as the site owner Amir seems to prefer non-coaxial wide horizontal directivity designs himself.
If you're using conventional drivers (i.e. not horns) you have a chance without digital delay; the best you can do is to physically align the fronts of their voice coils on the same plane and mount them vertically and closely (since most of your travels will be horizontal). Certainly do this for the tweeter and whatever it crosses over from at least. That should minimize parallax.
I just posted a summary of my system which is multi-way and was predicated on being phase linear (including of course correcting for propagation delay). It does use horns and a great deal of digital processing, but there's a section on drivers under 'setup' that gives a speil on how and why I set them up as I did; also if you're interested on how they were digitally corrected and tested. Might be of use.
An Unorthodox and Affordable 1st Class System
I just posted a summary of my system which is multi-way and was predicated on being phase linear (including of course correcting for propagation delay). It does use horns and a great deal of digital processing, but there's a section on drivers under 'setup' that gives a speil on how and why I set them up as I did; also if you're interested on how they were digitally corrected and tested. Might be of use.
An Unorthodox and Affordable 1st Class System
Depends on your point of view...
No, really, this is a rather simple geometric equation and all simulation software calculates the phase and amplitude effects on- or off-axis by calculating the size of the triangle formed by the two speakers and the ear (or microphone). And reality follows this pretty close (good software also takes secondary sources from baffle edge diffraction into account).
Does using a Horn change the CTC relationship between tweeter and woofer? I'm using a 12" tall horn with a 12" dia woofer... so technically the CTC is 13 inches... (including a 1" gap between driver and horn edges) - so does that mean that the crossover has to be lower than 1khz? I'm planning to cross at around 650 hz, so that should be OK, but could I go all the way up to 1khz safely?
Thanks!
Six - Mpls.
Thanks!
Six - Mpls.
This is the important bit, as well as considering whether acoustically coincident drivers is actually a design goal.
I'll take a pair of speakers I designed and built as an example. They're a pair of Faital 10FH520 and a 18Sound ND1460 on an RCF HF94 (90x40) horn. The crossover is around 1kHz, which would be considered much too high, by HiFi standards.
However, the aim of the TMM layout was to maintain vertical directivity down to a frequency below what the HF horn itself would support. So, the pair of 10" drivers, being an entire wavelength apart at 1kHz, exhibit a narrow vertical directivity that matches the HF horn pretty well. The directivity gradually widens as we go down the frequency range, until the drivers are within 1/4wavelength at around 250Hz.
Why's that useful?
Because this is a PA speaker. By raising the speakers reasonably high up and using something with a fairly tight vertical pattern, I can ensure that the people sat near the front aren't being deafened.
I often got comments about how the sound "throws" really well, and there's no magic. It's just that the speakers aren't throwing their full output at the front couple of rows - as you move back, you're getting closer to on-axis, so you've got more sound heading towards you.
Whether all of that work is useful in HiFi is up for debate. I'm in the controlled directivity camp, but others like dipoles and omnis. I suspect that partly comes back to individual listeners - some people might like the extra acoustic info, while others find it turns the stereo image to mush.
Chris