Running 4 speakers from a 2-channel amp

AL - I would never connect speakers in series but I believe DF doesn't suffer - at least with woofers in series within the same box. It appears that each woofer sees impedance of the other woofer in series (JBL wrote paper on that - that is incorrect) but in reality EMFs are in opposite phase and cancel. If amp's output impedance is zero - it looks like two woofers are connected in parallel in opposite polarities (EMFs cancel). Connecting woofers in series makes sense only when they are identical (for cancellation). Connecting speakers boxes in series will be a mess.

More info here: http://www.cartchunk.org/audiotopics/SeriesSpeaker.pdf

or here: http://www.monstercable.com/mpc/stable/tech/A2538_Wiring_Woofers.pdf

Jim -- Thanks for the update. Enjoy!

Kijanki -- That's an interesting thought, which I hadn't previously seen stated.

A series connection, whether of two woofers within an enclosure, or of two series connected speakers, would have to be done such that the plus (or red) terminal of one woofer or speaker is connected to the minus (or black) terminal of the other woofer (or speaker). Otherwise the two drivers would operate out of phase and cancel acoustically.

Given that arrangement, the voltages corresponding to the back-emf of each driver would add together, resulting in double the voltage assuming the drivers are identical. And the papers you referenced appear to say that the doubled voltage compensates for the doubled total impedance, resulting in the same current and hence the same dissipation of energy from each woofer (or speaker) that would occur with a single woofer (or speaker) connected to the same amplifier.

That seems to make sense. Thanks!

In any event, we (you, me, Rodman, HifiTime) are all in agreement that for other reasons series connection of speakers, especially if they are not identical, is a no-no.

Best regards,
-- Al

Thanks Kijanki . . . A good explanitation of why it can be appropriate to wire speakers in series under certain circumstances. It should be noted the main assumptions/limitations . .

First, it makes the loudspeaker performance more variable with manufacturing tolerances, so in i.e. the D'Appolito array mentioned the effective tolerance for the drivers themselves must be better than for the parallel connection to get the same average results. And for a professional sound reinforcement application, there's one-quarter the statistical redundancy in the event of driver damage.

Second, it assumes that the driving source (amplifier) have essentially a zero output impedance. Although one wouldn't expect to see an SET or OTL amp for this kind of application, series volume controls should definitely be avoided, especially because they introduce enough frequency-response variations of their own.

So for something like putting four pairs of identical outdoor speakers on a single receiver ... The series/parallel scheme works great, provided they're hooked directly to the (typical solid-state) amplifier.

Kirkus - I remember seeing large bass guitar stacks that had a lot of small (about 10") speakers (10-12). They must have been connected serial/parallel to obtain any drivable impedance.

I've read on this forum that before SS popularity speakers/drivers had very high impedance (16 or even 32ohm). Maybe it was to match better with tube gear but it could be for other reasons. Lower impedance allows to squeeze more power from SS amps.

Kirkus - I remember seeing large bass guitar stacks that had a lot of small (about 10") speakers (10-12). They must have been connected serial/parallel to obtain any drivable impedance.

Actually, just about every instrument cabinet I can think of is parallel connected . . .the classic Ampeg SVT bass cabinet used eight 10" 32 ohm speakers, and I think the modern versions do as well. In Fenders, many of the differences between the (ostensibily very similar) amplifiers used in heads and combos were different output transformers . . . i.e. a Super Reverb combo had 4 8-ohm speakers, and the amp had an OPT designed for a 2-ohm load.