OK I ran the Sim and the results are not nearly as pretty or as flat as Troels' results. I'm going with his as he know a hell of a lot more than I do lol. I'll build up his version and try it out.
Member Troels used a MW19 and MW16 in 2.5 configuration, not two MW16’s. This was just to give you a starting point to work from, unless you are going to build the SBA761 as per the plans then any change to the driver, not using the step baffle, changing the baffle width or changing the drive positions will need a crossover modification.
Units of such construction reaching response far in the hf zone makes oneself more trouble in designing a proper filter. That's it.
Last edited:
I believe the 1.3mh coil is not critical. However the R4 and C2 appear to be a zobel or notch? He states that he is trying to suppress energy in the 5-10khz region. In my system I will have to modify.
Last edited:
So I redid the sim. Values are recalculated. Does this seem reasonable? I replaced the 5mh with a 3.6mh
Attachments
Last edited:
R4 c2 can be most likely omitted for mw16.
Is your baffle tilted or stepped? If not I am afraid lr2 for tweeter will not phase match upper midwoofer.
Is your baffle tilted or stepped? If not I am afraid lr2 for tweeter will not phase match upper midwoofer.
Agree R4,C3 is a tank to reduce the MW19 breakup at around 8khz, the MW16 has a smoother roll and so not really needed.
How’s the phase alignment, when I modelled the SBA761 it looked pretty good. Not sure what the 2xMW16 will do to that plus driver spacing may also affect it a little.
How’s the phase alignment, when I modelled the SBA761 it looked pretty good. Not sure what the 2xMW16 will do to that plus driver spacing may also affect it a little.
Last edited:
Pida, thank you for that observation. It is not tilted or stepped back. This is where I was reading about addressing it in the crossover. The example I read was where you make the tweeter section one pole greater than the Mid to create delay.
Or one pole less. In case of tw29, just a capacitor plus rlc to adress Fs peak work well. But this is based on my projects. I guess you will need to make real measurements, I find it actually simpler way than trying to make crossovers without measurements.
Are you doing the crossover based on measurements or a traced/downloaded frd file?
Use xSim 3d and place the drivers exactly as you would place them in a box. to get a better feel for actual final response. Driver offset is to be entered for correct time alignment.
Use xSim 3d and place the drivers exactly as you would place them in a box. to get a better feel for actual final response. Driver offset is to be entered for correct time alignment.
I was using a traced file from Satori. XSIM 3D?
I do have the ability to measure I have Omnimic V2
I do have the ability to measure I have Omnimic V2
If you have Omnimic, Use measured values, you dont have to "estimate" the baffle step.
Xsim 3d.
Index of /Xsim/BetaTest
Click baffle on the top and place the drivers and offsets correctly.
you can still use the traced zma files,
Xsim 3d.
Index of /Xsim/BetaTest
Click baffle on the top and place the drivers and offsets correctly.
you can still use the traced zma files,
OK got it. Yea I can see the effect o Fr when I am able to set back the tweeter 0.6 inches. Also allows boundary effects. The thing that really affects the response is floor bounce.
It looks like there are many ways to skin a cat. But me not being as experienced as members here, I dont know what the "correct" or preferred solution is. I have 3-4 models that are pretty decent Fr.
It looks like there are many ways to skin a cat. But me not being as experienced as members here, I dont know what the "correct" or preferred solution is. I have 3-4 models that are pretty decent Fr.
Just found this Troels project that more cloely matches my project. SBA-16-MTM
I dont know the values but maybe I can reverse engineer it?
I dont know the values but maybe I can reverse engineer it?