I'm trying to design a 3 way speaker with a passive crossover for the mid+tweeter, and use active crossover with the twin woofers using minidsp 2x4 HD. Will be used with 4 ohm stable amplifiers.
The parts I'm considering are
Tweeter: SEAS Prestige 27TBCD/GB-DXT
Midrange: SB Acoustics SB12MNRX2-25-04
Woofers: 2 of SB Acoustics SB15SFCR39-8
I'm getting a feel for how the passive crossover might look like, using the manufacturer supplied graphs and vituixcad.
Attached is my first attempt. Please ignore the response below 300 Hz. I don't have any response shaping components in the network because I'm planning to do that (and room correction) using the DSP.
All I'm looking for right now is
I'd like to get some feedback on how this network will behave. Am I going in the right direction? Any suggestions to improve it?
The parts I'm considering are
Tweeter: SEAS Prestige 27TBCD/GB-DXT
Midrange: SB Acoustics SB12MNRX2-25-04
Woofers: 2 of SB Acoustics SB15SFCR39-8
I'm getting a feel for how the passive crossover might look like, using the manufacturer supplied graphs and vituixcad.
Attached is my first attempt. Please ignore the response below 300 Hz. I don't have any response shaping components in the network because I'm planning to do that (and room correction) using the DSP.
All I'm looking for right now is
- A reasonable response
- Good off axis response
- Reasonable phase response
- Not a big loss of efficiency
I'd like to get some feedback on how this network will behave. Am I going in the right direction? Any suggestions to improve it?
Attachments
I think I understand what you are doing and seems smart (OK, it would be better to just buy another second-hand small power amp for the treble. Why not?).
With your mini-DSP, you gain all kinds of wonderful EQ and level control and these should influence your XO design. Deserves some thought on the part of this forum, not to just assume the standard XO makes sense here.
Like almost all the "precisely" engineered XOs here, you are lacking an L-pad which you inevitably will want. Or just get another amp.
B.
With your mini-DSP, you gain all kinds of wonderful EQ and level control and these should influence your XO design. Deserves some thought on the part of this forum, not to just assume the standard XO makes sense here.
Like almost all the "precisely" engineered XOs here, you are lacking an L-pad which you inevitably will want. Or just get another amp.
B.
My plan is to have the woofers placed behind the midrange, in dual opposing configuration and radiating sideways. One hope of mine is to play around and see if I can achieve a cardioid pattern.
I have thought about the process of combining an active crossover at low frequencies with a passive crossover at higher frequencies... exact situation you are looking at now.
You will have powerful EQ capability on the lower channel (woofer) and the upper channel (mid+tweeter). Therefore, your passive crossover will not need to provide EQ such as baffle step compensation or managing overall tonal balance... your DSP can do that. However, any EQ you apply to the upper channel will hit the mid and tweeter both.
I believe the most important job that your passive crossover needs to do is to create a good Directivity index between the mid and the tweeter. If this is done well, you can EQ the mid+tweeter channel to be flat, and that flatness will carry over from the on-axis to the off-axis response, both will be good, and the DI will be good.
j.
You will have powerful EQ capability on the lower channel (woofer) and the upper channel (mid+tweeter). Therefore, your passive crossover will not need to provide EQ such as baffle step compensation or managing overall tonal balance... your DSP can do that. However, any EQ you apply to the upper channel will hit the mid and tweeter both.
I believe the most important job that your passive crossover needs to do is to create a good Directivity index between the mid and the tweeter. If this is done well, you can EQ the mid+tweeter channel to be flat, and that flatness will carry over from the on-axis to the off-axis response, both will be good, and the DI will be good.
j.
I do something similar but with an Ashly XR-1000 crossover. I wouldn't spend too much time with the simulations for the woofer to mid crossover. I found once the speakers are in good starting listening position it was best to reposition the speakers and tune the active crossover to avoid any room issues that can be fixed with positioning and frequency adjustments reducing EQ needs. In my case the woofer had a dip at 300hz that the mids didn't have. Using REW the best I could I tuned the crossover to minimize the dip. So it was crossed just below 300hz.
Also, even though the cross over is fairly low the woofer to mid distance was still important. A few inches made a difference. Having crossover flexibility at midbass frequencies is a blessing and a curse. A lot of speaker and chair moving with crossover adjusting to drive you mad and moving the mic short distances shows different responses and can take you down some rabbit holes until you find what you like best. I think it was worth the effort for me anyway. I bit off a black art. YMMV
Also, even though the cross over is fairly low the woofer to mid distance was still important. A few inches made a difference. Having crossover flexibility at midbass frequencies is a blessing and a curse. A lot of speaker and chair moving with crossover adjusting to drive you mad and moving the mic short distances shows different responses and can take you down some rabbit holes until you find what you like best. I think it was worth the effort for me anyway. I bit off a black art. YMMV
Last edited:
Thanks everyone for all your suggestions.
I also went ahead and tweaked the crossover a little, with the goal of recovering some efficiency, and ended up with this. On axis, I'm rolling off after 7k, but I'm fine with it. I plan to bring it up in dsp if necessary, based on measurements in the listening window.
Thanks for pointing out my mistake. I corrected it. I'll model the woofer crossover after I get a mock enclosure built.
I also went ahead and tweaked the crossover a little, with the goal of recovering some efficiency, and ended up with this. On axis, I'm rolling off after 7k, but I'm fine with it. I plan to bring it up in dsp if necessary, based on measurements in the listening window.
Hi Jim, is the DI of the new crossover any good? Off axis responses are within 5 db. It looks ok to me but I can't say for sure.
I can't really tell what you are doing, how far along your simulation is... Usually I first use the diffraction tool to simulate the horizontal and vertical responses of each driver, based on the baffle shape I have chosen. Then when I go to the main crossover page, I load all those responses in for each driver, something like 36 responses for each driver. Have you modelled your baffle shape, and can we see it?
Tigeranand, I abandoned this particular design.
I’m convinced that a coaxial design would be have better vertical directivity and decided to explore them instead (with seas or kef drivers). I also decided to start with a full range based 2 way waw to keep things simple. I’ll shortly post my build thread.
I’m convinced that a coaxial design would be have better vertical directivity and decided to explore them instead (with seas or kef drivers). I also decided to start with a full range based 2 way waw to keep things simple. I’ll shortly post my build thread.
