L100 out-of-phase midrange

[Zeb Norris]

I have a pair of L100s and a pair of 4312Bs that I love.

I've been reading a lot about these designs. Many people think that they benefit from an upgraded crossover, but I've been hesitant. I love the way they sound already, 6k spike and all.

Proponents of a crossover upgrade note that there's very little to restrict all 3 speakers from playing everything. With modern crossovers one needn't do this, leading to designs that deliver flatter performance.

In discussing upgrading crossovers it is noted that JBL wired these speakers so the midrange driver is out of phase. That struck me as odd.

But suddenly it makes sense.

Is it possible that JBL engineers were using phase cancellation as a form of crossover? That is, where the woofer and midrange overlap, the phase cancellation leads to a form of attenuation of the shared frequencies, thus staying "flatter" (seeming) than they would reinforcing each other? And the same for the frequencies shared between the mid and the tweeter?

Further, could the odd "spatial" effect of speakers running out-of-phase... in this case, not from side to side but equally inside each cabinet, create the "magic" of the L100 sound? It's obviously not physically hi-fidelity... yet it's just the sort or ear candy that could be perceived as pleasing with greater perception of depth. Could the 6k hump caused by the asymmetrical placement of the midrange driver in the baffle also be intentional; to create the perception of definition and "pull the ear away" from the phasey mids below 6k?

It's just a hypothesis; thoughts? Could this be so? Or have I misunderstood something?

[mixsit]

Interesting.. so subscribing ;)

[more10]

Its a standard trick to avoid cancellation at dividing frequency. A 6 dB filter shifts 45 degrees at -3 dB. A 12 dB shifts 90 degrees. A 18 dB shifts 135 degrees. A 24 dB shifts 180 degrees.

A 24/24 dB filter shifts total 360 degrees. The lowpass and highpass shift in opposite directions. So in total the signal will shift 360 over the crossover point. But there will be no cancellation since both channels are in phase at crossover frequency.

A 12/12 dB filter shifts total 180 degrees. In order to avoid cancellation at crossover point, one driver is phase shifted.

A 6/18 dB filter also shifts 180 degrees in total. Here you also have to shift one driver.

[Zeb Norris]

Does that apply to the mere 2 caps that constitute the "crossover" in these models?

[more10]

Find a drawing of the crossover and post it here.

[audiomagnate]

Find a drawing of the crossover and post it here.

It's literally two caps (and two l pads) , one in series with the mid, one with the tweeter. All three drivers are wired in phase.

http://forums.audioholics.com/forums/attachments/diy-corner-tips-techniques/3467d1156447567-vintage-jbl-west-coast-sound-becomes-%85-jbl-l-100-original-crossover-jpeg.jpg

[more10]

Could there be terminal color confusion?

[audiomagnate]

Could there be terminal color confusion?


Always a possibility with JBL.

[DavidF]

I have a pair of L100s and a pair of 4312Bs that I love.

I've been reading a lot about these designs. Many people think that they benefit from an upgraded crossover, but I've been hesitant. I love the way they sound already, 6k spike and all.

Proponents of a crossover upgrade note that there's very little to restrict all 3 speakers from playing everything. With modern crossovers one needn't do this, leading to designs that deliver flatter performance.

In discussing upgrading crossovers it is noted that JBL wired these speakers so the midrange driver is out of phase. That struck me as odd.

But suddenly it makes sense.

Is it possible that JBL engineers were using phase cancellation as a form of crossover? That is, where the woofer and midrange overlap, the phase cancellation leads to a form of attenuation of the shared frequencies, thus staying "flatter" (seeming) than they would reinforcing each other? And the same for the frequencies shared between the mid and the tweeter?

Further, could the odd "spatial" effect of speakers running out-of-phase... in this case, not from side to side but equally inside each cabinet, create the "magic" of the L100 sound? It's obviously not physically hi-fidelity... yet it's just the sort or ear candy that could be perceived as pleasing with greater perception of depth. Could the 6k hump caused by the asymmetrical placement of the midrange driver in the baffle also be intentional; to create the perception of definition and "pull the ear away" from the phasey mids below 6k?

It's just a hypothesis; thoughts? Could this be so? Or have I misunderstood something?

The 4310 variants did use a minimal xover design as far as component count and response manipulation. The caps in the mid and tweeter legs were to protect the driver and roll up the response, both. But there was no attempt to alter the high end response of any of the drivers.

I think you may be reading too much into the phase aspect of the xover as it relates to the overall system "magic". Remember that the woofer is used full range in the design which means it is responding up into the low treble. Having a very large segment of the critical mid range handled by the one driver is, I suspect, what you find pleasing in the overall system sound.

Any humps in the response around 6k would be the response peak of the mid driver that is not attenuated at the high end of its response.

[audiomagnate]

The 4310 variants did use a minimal xover design as far as component count and response manipulation. The caps in the mid and tweeter legs were to protect the driver and roll up the response, both. But there was no attempt to alter the high end response of any of the drivers.

I think you may be reading too much into the phase aspect of the xover as it relates to the overall system "magic". Remember that the woofer is used full range in the design which means it is responding up into the low treble. Having a very large segment of the critical mid range handled by the one driver is, I suspect, what you find pleasing in the overall system sound.

Any humps in the response around 6k would be the response peak of the mid driver that is not attenuated at the high end of its response.

It's probably cone breakup. That's why you don't want to drive a speaker beyond its operating range. There's only one reason the crossover in the L100 is so simple and that's cost, so there is zero downside to upgrading them exept the expense and work involved. At the very least coils in series with the woofer and midrange would be a huge improvement.

[ratitifb]

At the very least coils in series with the woofer ...would be a huge improvement.:hmm: not so sure regarding the punchy response of the 12" without any X-over resistor value in serie with the voice coil:rockon1: