Is it better to.....

[oilcanracer]

First of all, DON'T use passive sub crossovers. Due to impedances of subs and satellites, they simply cannot work. Been there, simulated it, tested it...er sold it anyway...but it just doesn't work worth a damn.

Second, somewhere you need an electronic summing stage. Don't just connect L & R positive leads together, because then any out-of-phase information can basically act like a bit of a short circuit to the preamp or crossover.

Really, you need a crossover with a built-in sub output. Most receivers have such an output, driven by home theater considerations.

i knew there was a reason not to connect left and right together after the amp, but was missing the technical part. this is going in my car. i have 5 1/4" speakers in the doors and they can handle the mids/highs fairly well, but fail in the subharmonic area.

i am using a lowpass crossed at 100hz then the signal then splits between the sub speakers and going to a highpass at 150hz getting sent to the door speakers.

just using the front l/r coming out of the radio amp(60 watts i think) and its enough to fill the truck up with good sound level.

 

[head_unit]

Maybe we are going to flame into Golden Ear territory, but there sure isn't sound localization with a 140 Hz 18 dB/8ave crossover woofer. i'd guess that with a woofer sitting sort of near the mid-ranges and a sharp slope cross-over, 200 Hz would be OK.

Some very solid research was done (Tom Holman/THX, I think) and they settled on 80 Hz as a crossover frequency which allowed positioning the sub in a variety of locations but the listeners could not tell where. I believe the slope was 18 dB/octave.
- Note this research presumed the sub did not emit noise or distortion products out of the crossover range. That means not overdriven, or certain bandpass designs if not on-axis.
- Higher crossover frequencies resulted in the sub position definitely audible in some cases.

It's certainly possible that higher crossover frequencies might not be localizable. That would depend on the specifics of the room, sub placement, main speaker placement, relative phase between mains and sub. And quite possibly, on the ear/brain of the listener as well as the music being played.

--> I'm not disagreeing with you, but I think your statement is based on a "happy" situation, and you can't project it to all systems.

To generalize, the sub should cross over at 80 or below to be on the safe side.

 

[bentoronto]

After the rude people who post to my BMW motorcycle forum, I appreciate your measured, data-referenced, and respectful reply.

I bet it isn't hard to create some kind of legitimate lab-like experiment where listeners can detect a sub location at frequencies lower than 80. Or real low if the test question posed to the testee is just "same" or "different."

But what if you posed the question, how high can I go and "get away" reasonably without harming the stereo illusion?

In rooms of my acquaintance, with midrange speakers (from 140 Hz) playing commercial recordings and my Klipschorn (1) directly behind one of those speakers or (2) pretty far away in a distant corner to one side... I have zero sense that the localization of any sound is distorted at all by the woofer location.

Footnote: Not long ago, I found myself in a great flame war about the question: if you had no output below some frequency, would you notice? You couldn't believe what great organ music I hear when my woofer is off.

 

[head_unit]

After the rude people who post to my BMW motorcycle forum, I appreciate your measured, data-referenced, and respectful reply.

Oof, know what you mean. Posted some very detailed questions to a Mercedes forum. Snotty answers basically like "don't you know how to look anything up, it's all there!" NO, it's NOT you morons! And gee, on a different forum the nice people just answered my questions. Sheesh!

But what if you posed the question, how high can I go and "get away" reasonably without harming the stereo illusion?.

I'd say the answer to that REALLY depends on the sub, the satellites, the room, and the listener(s). Those Bose Acoustimass get criticized for having a hole in the response between the subs and the satellites, but listeners who are not doing an A/B comparison wouldn't necessarily notice. In the case of those bandpass subs, the sub rolls off quickly at high frequencies AND the port acoustically filters higher frequency distortion --> hardly to "locate" than a distorting sealed box facing right towards the listener.

Footnote: Not long ago, I found myself in a great flame war about the question: if you had no output below some frequency, would you notice? You couldn't believe what great organ music I hear when my woofer is off.

Even if you HAVE output below some frequency, there was a paper by Louis Fielder at Dolby showed that you wouldn't hear it if it wasn't loud enough. Sounds obvious if you think about Fletcher-Munson hearing detectability curves, but Fielder's point was that extending the low frequency response of subs is useless unless they can generate a certain minimum SPL at those low frequencies.
The other wrinkle is that if the fundamental note is missing, your brain "fills it in" from the harmonic overtones, and you still "hear" the note anyway. All those MAXXBASS and similar processings rely on that psychoacoustic wrinkle.