How high can drivers with 105mm CTC be crossed?

Hi All,

I plan to use this driver for an array.
https://www.madisound.com/store/product_info.php?products_id=8990

Its CTC (105mm) suggests a crossover frequency no higher than 3130Hz. The crossover slope to the tweeter will be 24db/Octave.

I would like to understand:-
1) The sonic impact of crossing higher at 3500Hz?
2) What do you abt the highest xover frequency in an array without negative audible impact?
3) Does a deviation like this benefit (lets me get away) from high order crossover or low order one?

Thanks,
Goldy
 
hi ideally you would actually cross this over at around 1/2 wavelength so that would mean crossing at around 1.5 khz.

in line array use crossing higher will introduce increased comb filtering the futher up the frequency range you go and this is where the tweeter should be used.

by curving the array like so ( the distance from the listener to each individual speaker can be kept the same however, this will only work perfectly for a specific listening position.

by tapering the amount of power going to each speaker i.e. less power going to each speaker as it moves out from the centre of the array you can also lessen the combing effect.

the higher order the crossover the better as this will mean there is less 'stray' sound above the cutoff frequency and less likelyhood of combing.


All this being said I ran a 12 4" mids per side for quite some time - this was a vertical line with no curve and only mild power tapering and I could not hear any combing.


you should also ideally run a tweeter array if possible in order for the wavefront from the tweeters to mimic the wavefront from the mids of course this is not always possible.
 

CLS

Member
2005-06-17 6:58 am
Taiwan
I'm wondering how audible the combing effect really is.

I guess few is worse than what I have -- 50cm c-t-c distance between midrange and tweeter xover'ed at around 2.5~3kHz range (the most sensitive region of hearing BTW).

According to the diagrams in the books or various articles, the combing lobes look like many fingers pointing everywhere - it depends on the system design, but mostly spreaded in the vertical plane. So I imagine it'd sound like some kind of fluctuation in SPL at specific frequency range when I'm moving up and down. By the proportion of peaks and dips on those diagrams, they should be very obvious.

But, in reality, playing normal materials (music), I can't hear it :eek: Maybe it's because I'm so used to it, or my space is all too reflective.

How is your experience?
 
I Can measure it but i can't hear it, for technical perfection you would of course cross well below the beginning of combing frequenies or use a very small high midrange, a very small tweeter or planar style speaker for the tweeters and a larger low midrange. Even then there will be combing at very high frequencies.

Any speaker system will comb to a greater or lesser extent but personally i dont hear it.

It would be more of an issue on the horizontal plane than on the vertical particularly since most folks will be sitting at a specific height when listening.

Someone with better trained ears than me would probably hear it but i'm luck in that respect.

I ran 12 mids per side with no issue as a preliminary test for my surround system and as you may have seen i also used 6 per side for my open baffles.
The reason for that was to try and extend the midrange down from a 500hz crossover to the woofer to around 200hz.

This did work and also evened out the off axis performance but not enough to warrant the use of different speaker configurations i.e. running 2 woofers 6 mids and one tweeter.
 

CLS

Member
2005-06-17 6:58 am
Taiwan
Ah! I remember some now. I just forgot.

Many years ago, I had tried TTM in a vertical arrangement. 2 Dyaudio D21 tweeters were used for the sensitivity compensation. It has 11cm flange - an old school 3/4" dome tweeter, thus huge ctc between 2 of them.

The even larger and different ctc between the 7" mid and 2 tweeters individually made it worse.

It did behave very badly in the vertical dispersion. Lobbing effect was clearly audible. I could hear the treble changing obviously when moving up and down in just a few cm. Horrible, the worst I've had.

So evetually I changed the tweeter to a single one with higher sensitivity.

-------

Later I tried 2 tweeters once more on a center channel - an MTTM with 2 midbass drivers lining up horizontally and 2 tweeters in vertical between them.

In this case, the treble lobbing was audible when listened at different vertical positions but not as severe as the previous one. I lived with it for some time.

So I did hear them!

----------

However those were 2 tweeters and xover'ed to mid. If in a line array (with more than 4 drivers in a line), the situation is probably very different. I'd imagine there might be very very many lobes overlapping one another and flatten out. No?
 
Ah! I remember some now. I just forgot.

Many years ago, I had tried TTM in a vertical arrangement. 2 Dyaudio D21 tweeters were used for the sensitivity compensation. It has 11cm flange - an old school 3/4" dome tweeter, thus huge ctc between 2 of them.

The even larger and different ctc between the 7" mid and 2 tweeters individually made it worse.

It did behave very badly in the vertical dispersion. Lobbing effect was clearly audible. I could hear the treble changing obviously when moving up and down in just a few cm. Horrible, the worst I've had.

So evetually I changed the tweeter to a single one with higher sensitivity.

-------

Later I tried 2 tweeters once more on a center channel - an MTTM with 2 midbass drivers lining up horizontally and 2 tweeters in vertical between them.

In this case, the treble lobbing was audible when listened at different vertical positions but not as severe as the previous one. I lived with it for some time.

So I did hear them!

----------

However those were 2 tweeters and xover'ed to mid. If in a line array (with more than 4 drivers in a line), the situation is probably very different. I'd imagine there might be very very many lobes overlapping one another and flatten out. No?

sort of. A vertical tweeter array will project a sound field which mimics that of a mid range arrays it's use in the pro sound field is due to having to cater to multiple listening positions simultaneously so the entire speaker has to match. That being said you could get away with a single high power tweeter in order to match the efficiency of the line array if you were in a specific listening position.

More drivers will increase the amount of interference above a wavelength approximately half the ctc spacing of each driver so in an ideal world the tweeter would have a ctc spacing of 1cm in order to play by the 'rules' up to 20khz. The only way this would really be possible is by using a ribbon or electrostatic panel.

My line array was open baffle as i wanted to test the effect on high frequencies to see if with a narrow baffle i would even out the high frequency dispersion due to the rear frequencies wrapping around the baffle and augmenting the highs.

What i found was that up to around 5khz there was no loss or combing that i could measure but of course above this it was different as the rear of the speakers were not really generating any hf signal.
 

CLS

Member
2005-06-17 6:58 am
Taiwan
...

My line array was open baffle as i wanted to test the effect on high frequencies to see if with a narrow baffle i would even out the high frequency dispersion due to the rear frequencies wrapping around the baffle and augmenting the highs.

What i found was that up to around 5khz there was no loss or combing that i could measure but of course above this it was different as the rear of the speakers were not really generating any hf signal.

Yes! this more or less reflects my own experience, too.

For making it simpler, I didn't mention earilier that my current speakers are actually sort of dipole - probably a very messy type of asymetric quasi-dipole. Ah, maybe there's no such thing. It's just emitting sounds in both front and rear directions anyway.

See my avatar, there're 2 tweeter horns, one for front and the other is firing back (and they are out of phase). And the mid horn driver is open back (no chamber). The rear firing tweeter horn is attenuated somewhat to match the lower level of rear-firing midrange sound. (the bass section is not shown in the avatar, it's simple and pure dipole with small open baffle)

I've previously tried without back-firing tweeter and "sealing" the back of midhorn (by several layers of wool felt covering the driver). It sounded OK, and I couldn't say I heard the lobings clearly. But the later configuration with back-firing tweeter and open back mid is obviously much better in tonal balance across the room and overal coherence, also more stable in the vertical plane. Now it's not much different by hearing from lying in the couch or standing.

So I'd guess the dipole line array should perform similarly in this regard. :D
 
Last edited: