You can not insult an engineer if you can not prove that you better than him/her.
Last edited:
It's easy to build a system that is very accurate when you use an active xover. A 4th order LR active with its 24dB/octave slopes also minimizes comb filter effects in the crossover region, and distortion in general. Many drivers have a more limited range of good operation than you would think from looking at frequency response graphs, so the tight rolloffs are often important, especially when there's only two frequency bands.
Calibrating a passive crossover is a pain in the rumpster. Absolute SPL's have to be measured first, and level equalizing resistors determined, then you have to either measure impedance of each driver at the frequency you want to do the crossover at (it won't be the nominal impedance) or believe a published graph on that if there is one. Then you calculate the reactive devices, and then you might want to consider baffle step response EQ, at which point the math is over the heads of most people since everything is interactive. With active crossovers you can separate each function out (no interactivity) and do everything very accurately.
I built a 4 channel 3886 poweramp, about 50watts RMS into 8 ohms per channel, and Peerless makes some very great 6.5 inch woofers (I'm using the nomex cone one and love it - about $60 each). There's lots of good tweeters depending on personal preferences. Peerless and Dayton come to mind and might be great references to compare to. Ideally, the amplitude response of the tweeter should roll off an entire octave below the frequency you will cross the tweeter over at. The transient response and distortion level can get ugly right where the amplitude response starts seriously rolling off.
Adding active EQ to the woofer to make the speaker somewhat acoustically flat at the listening position down to 30 or 40HZ is a BIG plus, if you can come up with a circuit to do that (a Baxandall circuit can be designed to do that). I lifted parts of the Linkwitz circuits (and others), scaled the values of the reactive components to get the frequencies I want, verified the design in pspice software (the free version) on my computer, built it, tested it and calibrated the absolute levels with a calibrated mic and pink noise, and it's great. My own brew of tone controls (4 section Baxandall) then compensates for room acoustics, Fletcher-munson effects, and whatever else to give me the best possible enjoyment. I've been doing this since the 1960's, and had the good fortune of being able to work with and learn from some top notch EE's at Tektronix and Dolby labs.
If you want to play the speakers at very high levels, you'll want to go 3 way, adding 12 inch woofers for below 80-150HZ, and actively crossover the 6 inch and woofer in that range. Active EQ works great on closed box speakers. If they are bass reflex (vented), active EQ is no longer recommended. I don't recommend bass reflex. Mechanical damping is weak and varies a bunch, and can drift out of calibration, assuming it was ever calibrated in the first place. Speaker specs used for calculating port sizes aren't always real accurate or stable over time.
Calibrating a passive crossover is a pain in the rumpster. Absolute SPL's have to be measured first, and level equalizing resistors determined, then you have to either measure impedance of each driver at the frequency you want to do the crossover at (it won't be the nominal impedance) or believe a published graph on that if there is one. Then you calculate the reactive devices, and then you might want to consider baffle step response EQ, at which point the math is over the heads of most people since everything is interactive. With active crossovers you can separate each function out (no interactivity) and do everything very accurately.
I built a 4 channel 3886 poweramp, about 50watts RMS into 8 ohms per channel, and Peerless makes some very great 6.5 inch woofers (I'm using the nomex cone one and love it - about $60 each). There's lots of good tweeters depending on personal preferences. Peerless and Dayton come to mind and might be great references to compare to. Ideally, the amplitude response of the tweeter should roll off an entire octave below the frequency you will cross the tweeter over at. The transient response and distortion level can get ugly right where the amplitude response starts seriously rolling off.
Adding active EQ to the woofer to make the speaker somewhat acoustically flat at the listening position down to 30 or 40HZ is a BIG plus, if you can come up with a circuit to do that (a Baxandall circuit can be designed to do that). I lifted parts of the Linkwitz circuits (and others), scaled the values of the reactive components to get the frequencies I want, verified the design in pspice software (the free version) on my computer, built it, tested it and calibrated the absolute levels with a calibrated mic and pink noise, and it's great. My own brew of tone controls (4 section Baxandall) then compensates for room acoustics, Fletcher-munson effects, and whatever else to give me the best possible enjoyment. I've been doing this since the 1960's, and had the good fortune of being able to work with and learn from some top notch EE's at Tektronix and Dolby labs.
If you want to play the speakers at very high levels, you'll want to go 3 way, adding 12 inch woofers for below 80-150HZ, and actively crossover the 6 inch and woofer in that range. Active EQ works great on closed box speakers. If they are bass reflex (vented), active EQ is no longer recommended. I don't recommend bass reflex. Mechanical damping is weak and varies a bunch, and can drift out of calibration, assuming it was ever calibrated in the first place. Speaker specs used for calculating port sizes aren't always real accurate or stable over time.
Last edited:
- Status
- Not open for further replies.