My Satori based 2.5 way is up and running, and I have been tweaking the crossover. Sealed, single internal compartment (although I might add a divider in the future). I built the crossover in an external box. And use an 8pole Speakon connector to connect the external crossover to the speaker itself.
Tweeter: TW29DN-8
Midbass: MW-16P
I am using stock / DIN baffle FRD ZMA files, and modeling in Boxsim. (I bought a measurement mic, but haven't figured out how to use REW yet).
View of response (corrected for baffle step simulation) from the stock files.
Crossover:
So far. It sounds pretty great! Very happy with the sound so far. The top Midbass is running first order, with just a 1.25mH coil. It seems to blend pretty nicely with the tweeter, with the long freq overlap. The bottom midbass mostly provides the baffle step boost with the 7mH inductor. I tried a 2.5mH on the bottom woofer. Which makes it closer to a 2way MTM, but it was too bass shy. I also tried the 2.5mH coil and 10u woofer crossover in parallel with the main midbass, AFTER the 1.25mH coil. That gave a nice flat curve, but I did not like how congested that made the midbass sound in vocals. Not sure why.
Tweeter: TW29DN-8
Midbass: MW-16P
I am using stock / DIN baffle FRD ZMA files, and modeling in Boxsim. (I bought a measurement mic, but haven't figured out how to use REW yet).
View of response (corrected for baffle step simulation) from the stock files.
Crossover:
So far. It sounds pretty great! Very happy with the sound so far. The top Midbass is running first order, with just a 1.25mH coil. It seems to blend pretty nicely with the tweeter, with the long freq overlap. The bottom midbass mostly provides the baffle step boost with the 7mH inductor. I tried a 2.5mH on the bottom woofer. Which makes it closer to a 2way MTM, but it was too bass shy. I also tried the 2.5mH coil and 10u woofer crossover in parallel with the main midbass, AFTER the 1.25mH coil. That gave a nice flat curve, but I did not like how congested that made the midbass sound in vocals. Not sure why.
Attachments
I am glad it sounds good!
One thing that concerns me is that the tweeter filter has a low cutoff, very low in fact. The tweeter is down just 7 dB at 1000 Hz, and generally I like to see at least 12 dB attenuation at 1000 Hz with most dome tweeters such as this one.
j.
One thing that concerns me is that the tweeter filter has a low cutoff, very low in fact. The tweeter is down just 7 dB at 1000 Hz, and generally I like to see at least 12 dB attenuation at 1000 Hz with most dome tweeters such as this one.
j.
Have a look at Troels Gravesen’s SBA761, this uses the MW19, MW16 and TW29 in a 2.5way nominally first order setup. If you search the way back machine you can get the crossover values as it was originally an open source design. May provide some inspiration and at least a crossover for the MW19 mid and tweeter.
http://www.troelsgravesen.dk/SBA-761.htm
http://www.troelsgravesen.dk/SBA-761.htm
Thanks for that. I will try to steepen that cutoff. I was trying to get down close to 2000Hz for the crossover point, but still keep it 2nd order on that tweeter. I ave some 6uF caps, will model that drop. Maybe even a 3.3uf?
Yes! That was inspiration to trying the MW16 with a first order filter. I like that a first order is relatively flat in the midrange. If I model 2nd order on that driver with a 6uF or 10uF cap, it really boosts that midrange freq response.
Interesting. I tried swapping a 6uF cap in for the 10uf cap on the tweeter. It had minimal effect when looking at the overall freq response in the simulation, really just flattered the downward curve slope in the high freq
BUT. It sounded terrible. I would describe it as a "crispy" sound on the high end with the 6uF.
The 10uf cap makes it sound so much better balanced. Hopefully this tweeter can deal with the lower freq extension a bit better than most. It has a pretty low Fs.
(Maybe the modification on the tweeter allowed more of the midbass upper freq to dominate, and I need to roll that off too with a second order filter?)
BUT. It sounded terrible. I would describe it as a "crispy" sound on the high end with the 6uF.
The 10uf cap makes it sound so much better balanced. Hopefully this tweeter can deal with the lower freq extension a bit better than most. It has a pretty low Fs.
(Maybe the modification on the tweeter allowed more of the midbass upper freq to dominate, and I need to roll that off too with a second order filter?)
I tried altering the crossover to add a baffle step correction, placed before both midwoofers, to flatten the overall response. (And swap out that 7mH inductor on the bottom midwoofer) Also converted the upper mid to a second-order filter.
This gave interesting results - the two midbass are more coupled. Tweaks to one midbass filter now affect the other filter. I think I created a type of voltage divider between the filters?
Crossover:
Freq response:
Also attached phase response, in case there is anything interesting there. i dont know how to read that chart properly.
This weekend I finally plan to bust out the microphone and learn how to use REW, hopefully capture more accurate view of the drivers in the box.
This gave interesting results - the two midbass are more coupled. Tweaks to one midbass filter now affect the other filter. I think I created a type of voltage divider between the filters?
Crossover:
Freq response:
Also attached phase response, in case there is anything interesting there. i dont know how to read that chart properly.
This weekend I finally plan to bust out the microphone and learn how to use REW, hopefully capture more accurate view of the drivers in the box.
Attachments
This kind of crossover often tends to make the upper woofer have a higher order rolloff. Sometimes it will vary in different regions as it seems to here. If it works for you that's fine, otherwise you might need to find a way to influence them with more individuality.