I have searched a bit for this but been unable to get a definitive answer;
What are the correct or textbook LP crossover slopes for the bass or .5 woofer when using DSP? For arguments sake, lets say all drivers are flat and match dispersion pattern well. LP for low woofer around BSC say 400hz, and midwoofer to tweeter XO at 1600hz. Full active amp direct to driver system.
Reason I ask is because all 2.5 way passive crossovers that I have seen "cascade" the .5 woofer, that is, the signal they receive is in series, after the network that rolls off the midwoofer. Which looks, to me, to create a unique roll off. Unless there's something I'm missing, this is not achievable in DSP with IIR filtering.
What I can simulate with a BW6 LP on the low .5 woofer seems still have some phase disturbance to the crossover between the midwoofer to tweeter. This seems to be fixable using FIR linear phase low passes, but I'm sticking with IIR at the moment due to using my systems with TV/HT as well as music and don't want the FIR delay.
Appreciate any comment
What are the correct or textbook LP crossover slopes for the bass or .5 woofer when using DSP? For arguments sake, lets say all drivers are flat and match dispersion pattern well. LP for low woofer around BSC say 400hz, and midwoofer to tweeter XO at 1600hz. Full active amp direct to driver system.
Reason I ask is because all 2.5 way passive crossovers that I have seen "cascade" the .5 woofer, that is, the signal they receive is in series, after the network that rolls off the midwoofer. Which looks, to me, to create a unique roll off. Unless there's something I'm missing, this is not achievable in DSP with IIR filtering.
What I can simulate with a BW6 LP on the low .5 woofer seems still have some phase disturbance to the crossover between the midwoofer to tweeter. This seems to be fixable using FIR linear phase low passes, but I'm sticking with IIR at the moment due to using my systems with TV/HT as well as music and don't want the FIR delay.
Appreciate any comment
You want a relative first order rolloff. If you go faster than that you risk setting up cancellations with the other woofer as the phase separates too much before the woofer is down far enough.
Of course this should take into account the phase of the midwoofer, but how much it matters depends on where and what the upper rolloff is. Also, as to why consider the upper crossover there's the issue of avoiding woofer output near the crossover, which may not be so bad a problem but is different from case to case.
DSP is no free ticket not to do all the same simulation work we always did with passive crossovers, just using different filter types. Simulation is also about so many other things besides electrical filters.
Of course this should take into account the phase of the midwoofer, but how much it matters depends on where and what the upper rolloff is. Also, as to why consider the upper crossover there's the issue of avoiding woofer output near the crossover, which may not be so bad a problem but is different from case to case.
DSP is no free ticket not to do all the same simulation work we always did with passive crossovers, just using different filter types. Simulation is also about so many other things besides electrical filters.
Any amplitude response generated by passive filtering is also achievable with IIR DSP. But if not, buy a more sophisticated DSP that can do it.
Of course you need to think slightly differently placing coils & caps and adjusting a DSP with mouseclicks.
Last edited:
FWIW, plenty published 2.5 way designs use completely separate filters for the bass/midbass drivers - a few examples:
Ekta-25
NOMEX-mkII
CNO-25-mkII
Test Selbstbauprojekt - Monacor Klang + Ton „Hobo II“
Test Selbstbauprojekt - Mivoc K+T Sixteen
As you can see, most of them are using 1st order for the bass LPF.
It's quite common to include extra components to notch out any higher frequency peaks that may be insufficiently suppressed by that shallow slope however.
HTH,
David.