Midrange driver flat between 600Hz - 7Khz

Status
Not open for further replies.
What dispersion do you want, only flat on axis or flat off axis too? Will it be front firing only or mounted in a dipole?

I think a small 3-inch cone driver with a phase plug should do about what you want. Those small drivers rareley have super flat responses though but they are still pretty good. If you could raise the lower frequency to about 1000-1200 you could use an OSWG-12 waveguide and a compression driver. That, would give you very uniform dispersion which is really nice.

That combo wouldn't shine unless you control directivity lower down though, so unless you have a big say 15-inch woofer or cardioid woofer it probably wont mate that well.
 
Front firing only, flat off axis would be a bonus as I seldom sit bang on in front.

I was thinking minimum 4" - 5" ?

The most uniform dispersion on a cone driver I've seen is those with a phaseplug, and a 6.5 inch driver falls off at 2-3 khz. If you use a dipole or cardioid then you can probably push the XO closer to 3 khz but if you want 90 degree dispersion then closer to 2 khz.

So if you want 6-7 khz dispersion then you'd have to halve the driver diameter, hence a ~ 3 inch driver. Phase plugged small midranges are very rare though. There are non phaseplugged drivers that are almost as good though, the FF85WK is about as good as it gets in that form factor. http://www.fostexinternational.com/docs/speaker_components/pdf/ffwk.pdf

It's cheap too which is nice, not very efficient and has some hiccups but the overall response is very good and consistent.

Also, a 4-5 inch driver that only covers 600 and upwards is a bit of a waste unless you want really high SPL =)
 
Last edited:
Some thoughts
- flat on-axis is the easy part, if other considerations are nulled!
- to make a good speaker to be used in a domestic room, power response must be smooth too ( smooth roll-off without steps and undulations)
- how much max spl - it translates to distortion at moderate/high levels

These rquirements are reasonably well filled By some of the best 4-5" drivers. When crossover is applied usually around 400 and 3-4000Hz, distortion and dispersion is ok. This is near ultimate of any real world driver and at least LR4 crossovers must be used.

I have measured Audax HM100Z0 and it is a good candidate. Measured at 1m in an oversize closed box.
 

Attachments

  • 3wprotoc m 0-90¤ 5ms 124.png
    3wprotoc m 0-90¤ 5ms 124.png
    132.5 KB · Views: 534
  • audax hm100 disto 90db 1m.png
    audax hm100 disto 90db 1m.png
    116.5 KB · Views: 536
*cough* spherical speaker made from IKEA bowls *cough* 😉

-you'll still have pressure loss on the low end depending on the size of the sphere, so it's not going to be "flat" anymore.

Still, a spherical shape (a larger one anyway relative to the driver) substantially reduces diffraction effects. :up:

The problem with a larger sphere is that it usually moves the driver away from the next driver - with a resulting irregular response due to combing/interference. :down: On the other hand if it's a small fullrane driver (with a lower freq. crossover to a bass driver) or a coaxial driver (perhaps also with a lower freq. crossover to a bass driver) then it's likely to not present any problems.

Speaking of a coaxial driver:

COAX-18
 
-you'll still have pressure loss on the low end depending on the size of the sphere, so it's not going to be "flat" anymore.

Still, a spherical shape (a larger one anyway relative to the driver) substantially reduces diffraction effects. :up:

The problem with a larger sphere is that it usually moves the driver away from the next driver - with a resulting irregular response due to combing/interference. :down: On the other hand if it's a small fullrane driver (with a lower freq. crossover to a bass driver) or a coaxial driver (perhaps also with a lower freq. crossover to a bass driver) then it's likely to not present any problems.

Speaking of a coaxial driver:

COAX-18

That loss will be global though, so easily fixed with some EQ =)
 
Status
Not open for further replies.