Crossover for floorspeaker based on Paul Carmody's ZX Spectrum

Hi to all!
I'm new to the forum and also i'm new to diy speakers :)

My project starts from Paul Carmody's ZX Spectrum/Amiga, a floorspeaker 102cm height, 21.7 wide and 29.3 depth with HiVi M6N woofer and Monacor DT-300+waveguide WG-300. I choose this tweeter with his waveguide because i found it at a really good price and it seemed to me good for crossing with M6N at 2kHz.
Vent is 5cm diameter (length will be probably between 8.9-3.5" as in Paul's project for a F3 of 42Hz).

This is the cabinet sketch:

[IMGDEAD]http://zeppelintest.altervista.org/Progetto.jpg[/IMGDEAD]

I added a close part at the bottom to have the speakers a little bit higher: here i will put the crossover.

I need some help for the crossover because i found the specs of the DT-300 tweeter but i can't find the specs of DT-300+ the waveguide WG-300. I found only some data about SPL increment:

Hz / dB Increment
600 / +4,5
1.000 / +5
1.500 / +8
2.000 / +8
3.000 / +7
4.000 / +5
8.000 / +2
10.000 / +0

The waveguide will improve response of the tweeter at lower frequencies, probably facilitating the crossover with the woofer but i don't want the tweeter to be too loud...

This is the graph of the DT-300 without WG-300:

[IMGDEAD]http://zeppelintest.altervista.org/dt300.jpg[/IMGDEAD]

Anyone could please help me designing the correct crossover?

Many thanks!
 
Mmmhhm it's not so easy. I see that in Carmody's design the crossover is at around 1.8 kHz, same as yours. I would keep the same lowpass ( 3 mH+(100 nF)+15 uF ) for the hi-vi and try different values for the 3rd order filter on the tweteer. The 3 mH coil is wrong, 0.3 is right. Or you can start with 1st order filter, as the waveguide might be more "friendly" with 6 dB slopes and so 1, 1.5 and 2.2 uF might be just right
 
Mmmhhm it's not so easy. I see that in Carmody's design the crossover is at around 1.8 kHz, same as yours. I would keep the same lowpass ( 3 mH+(100 nF)+15 uF ) for the hi-vi and try different values for the 3rd order filter on the tweteer. The 3 mH coil is wrong, 0.3 is right. Or you can start with 1st order filter, as the waveguide might be more "friendly" with 6 dB slopes and so 1, 1.5 and 2.2 uF might be just right

Thanks for the reply picowallspeaker! Consider that i'm a "real novice".... I didn't understood how to modify the scheme: where do i have to put (or change) the value 1, 1.5 and 2.2 uF ?
 
I'm no expert at x-over design, thou I enjoy very much reading about technical design of how crossovers are implemented.
When you see a metal woofer, as the hi-vi, you'll surely find some method to overcome the "nasty early break up effects" such ill resonances coming from the metal cone; so a filter has to put those nasty effects under udibility threshold.
The game here is: a very big coil so that the woofer response starts to decay very early; also the tweeter's response has to rise in the same manner to sum acoustically flat with the woofer. Moreover, the tweeter is fragile and to withstand some more power a 3rd order might be required.

What I suggest is to try with 1st order on the tweeter -power might be an issue !
Check also the polarity ( phase)
Put some 3.3 Ω before the cap

When designing with pc, I see that most of the utilities are used: not only Frequency Response but same importance is shared by the Phase graphs and Impedance graph
 

jReave

Member
2012-10-30 4:34 pm
Because your simulation doesn't have real SPL values, I can't tell if have included baffle diffraction effects, which you must.

As it stands, you need to increase the inductor on the woofer and include a tanking capacitor for the cone resonance. See the ZX Spectrum xo.

Add an L-pad or some series resistance either before or after the xo (or both) to the tweeter to pull it's level down to match the woofer. Then see how it looks.

And picowallspeaker is absolutely correct, you also have to align the tweeter and woofer phase in the xo region and keep your eye on the summed impedance.

2,000Hz looks like a good xo point. I might try LR4 acoustic slopes using 2nd order electrical on the woofer and 3rd order electrical on the tweeter. Again see Carmody's spectrum.
 
Last edited: