on the usefullness of 3/4" tweeters

I would like to discuss the lowest practical crossover frequency for 3/4" tweeters.

[IMGDEAD]http://i35.tinypic.com/fliyqx.jpg[/IMGDEAD]

For seamless synchronization between the woofer "W" and the tweeter "T" the distance from both acoustic centers "D" should be no greater than half the wavelength of the crossover frequency.

To illustrate a practical example I will use a very small tweeter - the Dayton ND20FB-4 Rear-Mount
Dayton ND20FB-4 Rear-Mount 3/4" Neodymium Dome Tweeter | Parts-Express.com
The impedance graph shows a peak at around 2kHz prompting for a crossover somewhere above 3kHz. To be safe and to simplify the calculation let's assume a crossover at 3440Hz. That leaves us with an acoustic centers distance D of 50mm. The overall diameter of the Dayton ND20FB is given as 35.7mm so the distance from edge to the center "dt" is about 18mm. Assuming no space wasted between the tweeter and woofer "ds", the resulting overall diameter of an appropriate woofer should be no greater than 64mm. That leaves all woofers greater than 2" out of the picture. Even the tiny 2.5" Peerless 830985 can't make it by 5mm.

Moving the crossover back to 3000Hz lets us choose from drivers with outer diameter less than 78mm which barely makes the Hi-Vi B3S the only reasonable choice at the expense of tweeter distortion.
 
You've pretty coralled the external tweeter mount problems. You can help those by constructing a shorting plug for voice coils over about an inch and a quater and mounting such small tweeter onto it. If it gets long enough you can start calling it a phase plug. Of course the cone has to behave properly without a dustcap, and you have to better like the problems caused by having the tweeter mounted on a post and/or with a cone around it. Other than that, you're kinda stuck.

You can see what I did to one of those poor Daytons here:

http://www.diyaudio.com/forums/parts/170064-smallest-tweeter-available.html
 
Last edited:
I might point out that if used with a dome midrange driver (like, say, the 2" Dayton RS52), they can cross over much higher - especially if you trim down the mounting flange. I've also seen it used with various 4" midrange drivers with fairly good results - and there's no shortage of those.

There are two advantages to a tiny tweeter: One, the distortion from a $12 Aurasound NT1 above 4khz is lower than anything else you'll find under $50; and two, the dispersion from such a tweeter is going to be better above 10khz than a 1" tweeter.

The cost-effectiveness thing is probably the largest part of it, though.
 
Last edited:
I would like to discuss the lowest practical crossover frequency for 3/4" tweeters.
For seamless synchronization between the woofer "W" and the tweeter "T" the distance from both acoustic centers "D" should be no greater than half the wavelength of the crossover frequency.

Now you learn the art of compromise.
There are many speakers on the market (probably the majority), including lauded high end models, that do not follow that rule.
 
I should exclusively state that a 3-way design is preferred since 3/4" tweeters definitely make sense for 4-way. I like the flow of ideas this is generating.
It seems that a wide surround helps the dispersion in the top octaves while keeping the Fs further low so that's one alternative. A 1" tweeter with a wide surround operates similarly to a 3/4" off-axis.

I did a quick and dirty experiment with my current system. It's a 3-way computer-XO active. I moved the crossover frequency following the half wavelength rule and there's a very audible improvement in coherency. I didn't take measurements and I'm sure there would be dips in the FR with the current crossover but I like it much better.
 

Dr_EM

Member
2006-10-18 12:00 pm
Swindon
What might the merits be of the attached arrangement? 1" 'fullrange' units surrounding a 3/4" tweeter. The use of 5 (or 6) 1" drivers gets the sensitivity and max SPL up to reasonable amounts yet the small size allows a high crossover point. A possible pitfall is the spacing of the 1" drivers from each other, but how is this different to a tweeter mounted in the centre of a conventional driver as per coaxial units? All the 1" units are low-passed.
 

Attachments

  • driver circle.png
    driver circle.png
    6.6 KB · Views: 570

Dr_EM

Member
2006-10-18 12:00 pm
Swindon
Actually, looking into it, this half-wavelength rule restricts us to crossovers of 1kHz or lower when using a 5-6" midrange unit? Many systems use such sized midrange units and much higher crossovers, what is the implication of doing this? Of course crossover slope makes a difference, but assuming 12dB here.

Incidentally, what is the pretence behind the larger than necessary driver spacing employed in this design?

HTGuide Forum - Design Concept Study: M8td3
 
Actually, looking into it, this half-wavelength rule restricts us to crossovers of 1kHz or lower when using a 5-6" midrange unit? Many systems use such sized midrange units and much higher crossovers, what is the implication of doing this?

That's my point. I'd love to see some research on the effects of the half-wavelength rule. It's obvious that the closer the drivers are the better the transition but is the half-wavelength distance really crucial?
 
The reason I'm going to use 8 of the Dayton ND20FB 3/4 inch domes (in a Bozak arrangement) for my center channel speaker under my 42 inch LCD TV is because I want to cross over from a 2 inch HiVi DMN-A cloth dome mid at 6-7kHZ 2 pole. The laws of physics suggest that the distance between the mid dome and these tweets is too long for that xover frequency. Hence the array. Any nulls created by any single tweeter and the 2 inch dome will be largely filled in by the other tweeters/mid dome combinations, since the distances are all different (so their nulls will all be at different frequencies). Further, I chose 6-7kHZ because that's just above where our ear-brain mechanism is good at sensing image location. Above that frequency it gets real "ambient". Below 1kHZ I'll be crossing over to four 5 inch drivers on panels that are angled outward by maybe 15 degrees (2 on each panel - the tweets are all on the center panel). I don't see why this wouldn't work real well. Criticisms encouraged.
 
The reason I'm going to use 8 of the Dayton ND20FB 3/4 inch domes (in a Bozak arrangement) for my center channel speaker under my 42 inch LCD TV is because I want to cross over from a 2 inch HiVi DMN-A cloth dome mid at 6-7kHZ 2 pole.

I have yet to hear a speaker with multiple tweeters that sounds right (except dipoles).

Any nulls created by any single tweeter and the 2 inch dome will be largely filled in by the other tweeters/mid dome combinations, since the distances are all different (so their nulls will all be at different frequencies).

I can't verify this but my intuition suggests adding more drivers will not easily solve the problem with cancellations. I think for this approach an arrangement based on luck can be more successful than one based on science.

Further, I chose 6-7kHZ because that's just above where our ear-brain mechanism is good at sensing image location. Above that frequency it gets real "ambient".

http://www.diyaudio.com/forums/multi-way/147013-tweeters-dont-drill-your-ears-5.html#post2694059
I did a quick and dirty experiment to evaluate different crossover points. Basically they all sounded similar but the ones higher than 4kHz were not so great. I also found that wide dispersion at the top of the highs does not improve the sound dramatically. It may not be worth the extra trouble to get it right.

I'm sorry for bashing your ideas without providing hard data but I thought you might be interested in my opinion.
 
I have yet to hear a speaker with multiple tweeters that sounds right (except dipoles).
I can't verify this but my intuition suggests adding more drivers will not easily solve the problem with cancellations. I think for this approach an arrangement based on luck can be more successful than one based on science.
http://www.diyaudio.com/forums/multi-way/147013-tweeters-dont-drill-your-ears-5.html#post2694059
I did a quick and dirty experiment to evaluate different crossover points. Basically they all sounded similar but the ones higher than 4kHz were not so great. I also found that wide dispersion at the top of the highs does not improve the sound dramatically. It may not be worth the extra trouble to get it right.
I'm sorry for bashing your ideas without providing hard data but I thought you might be interested in my opinion.
Boris, I appreciate your willingness to politely tell me what comes to your mind. As with most of my projects, I will probably learn a few things the hard way on this one too. I'm actually more worried about the psycho-acoustic effects of having two midrange drivers (100HZ - 1kHZ, 5 inch cones) spaced apart horizontally by about a foot. I can't quite imagine what that will do, but I bet it does something. Maybe excessive sense of diffusion at the high end (?).
 
Actually, looking into it, this half-wavelength rule restricts us to crossovers of 1kHz or lower when using a 5-6" midrange unit? Many systems use such sized midrange units and much higher crossovers, what is the implication of doing this? Of course crossover slope makes a difference, but assuming 12dB here.

Incidentally, what is the pretence behind the larger than necessary driver spacing employed in this design?

HTGuide Forum - Design Concept Study: M8td3
Go to post #20. He's talking about midranges and "dual 8" configuration"... for the new (renamed) M8td3.
And cutting the baffle, and problems with sawdust in the cabinet.:xfingers:
Where was the original (2-Way) xover, anyway?
M8ta it's a project i did a few years ago not expecting anyone else to build.
JonMarsh
Hi,hi...
http://www.youtube.com/watch?hl=en&v=fy4kSpqNka8
 
Last edited: