Proper speaker quad wiring?

[Sowhat]

A customer has a Fender/Sunn 4 x 10 and it's wired in series parallel... no surprise there.

The thing that caught my interest is that the cabinet is wired with two series pair of speakers connected in parallel. I have always used 2 parallel pairs wired in series, you get the same resultant total impedance with both but the 2 series pair version of this wiring presents much different and more extreme results when there's a problem and one speaker goes out.

For example: if you have 4 ohm speakers each at 50 watts then the normal working series parallel wiring would result in a cabinet with 4 ohms impedance and 200 watts of power handling capacity ideally. If the cabinet is wired with the 2 series pair in parallel, the loss of 1 speaker will take out it's pair speaker as well resulting in two speakers going silent. The wattage will go down to 100 watts and the impedance will raise to 8 ohms. If you have the cabinet wired in 2 parallel pair in series, if 1 speaker goes out only that speaker will go silent, the wattage will drop to 150 watts and the impedance will rise to 6 ohms.

This makes me wonder why anyone would want to wire a quad in series parallel with 2 series pairs in parallel? Does anyone have any thoughts on this. I know when I scan the web I see both circuits recommended but I also know that stuff on the web is often "monkey see/monkey do" and a deeper understanding of the circuit analysis is almost always missing.

[J M Fahey]

A customer has a Fender/Sunn 4 x 10 and it's wired in series parallel... no surprise there.

The thing that caught my interest is that the cabinet is wired with two series pair of speakers connected in parallel. I have always used 2 parallel pairs wired in series, you get the same resultant total impedance with both but the 2 series pair version of this wiring presents much different and more extreme results when there's a problem and one speaker goes out.

For example: if you have 4 ohm speakers each at 50 watts then the normal working series parallel wiring would result in a cabinet with 4 ohms impedance and 200 watts of power handling capacity ideally. If the cabinet is wired with the 2 series pair in parallel, the loss of 1 speaker will take out it's pair speaker as well resulting in two speakers going silent. The wattage will go down to 100 watts and the impedance will raise to 8 ohms. If you have the cabinet wired in 2 parallel pair in series, if 1 speaker goes out only that speaker will go silent, the wattage will drop to 150 watts and the impedance will rise to 6 ohms.

This makes me wonder why anyone would want to wire a quad in series parallel with 2 series pairs in parallel? Does anyone have any thoughts on this. I know when I scan the web I see both circuits recommended but I also know that stuff on the web is often "monkey see/monkey do" and a deeper understanding of the circuit analysis is almost always missing.

Sorry but you (they) have it wrong.

Consider first a 200W/4 ohms SS amplifier, because it's more predictable.
Then weŽll extend it to a tube one.

In any alternative, the full cabinet impedance will be 4 ohms, each speaker will get 50W, total power handling matches the amplifier output into that load.
So far so good.

Now, suppose 1 speaker opens.

a) if you have 2 independent series , paralleled, one open speaker pulls its partner out of the circuit, cabinet impedance rises to 8 ohms, **available power** drops to 100W/8 ohms, each remaining speaker *still* receives 50W, you lose volume but you get finish your show.
I think a similar logic lies behind those Ampeg 10" 32 ohms speakers, all in parallel: you might have 1 or more die mid show, the rest would continue pumping until the end.

b) if you have 2 paralleled pairs, in series, and 1 opens, impedance rises to 6 ohms (2+4), power drops to 200*4/6=133W, and is split unevenly: 2/3 of that to the single 4 ohms speaker (89W) and 1/3 is split between the remaining two, or some 22W each, a most uneven distribution.

If the cabinet fails because it was so close to the limit, now you will *certainly* kill that poor speaker receiving almost 90W, and in a couple songs your cabinet will mute.

*If* you had a tube amp, bye bye output transformer, just to make a "happier" night.

I think it was more acceptable in the old days, where (tube) amp power was limited , so, say, those Silvertone 6x10" combos, 3 parallels in series with 3 parallels , driven by 2x6L6 were not in danger, but as soon as cheap SS power was available, now you could easily burn speakers and then some.

Edit: I forgot, in your example, cabinet wattage does not drop to 150W since there's no even distribution, but to 50x1.5=75W .... while in the other case it drops to 100W .... and power drops to match, so no big deal.
In a tube amp. it will not drop *linearly* but the general distribution analysis still applies.

[Sowhat]

Thanks for the analysis, there is certainly more to consider with something taken as so slam dunk simple. I never though about the uneven power distribution aspect... makes a lot of sense, still so many cabinets are wired as parallel pairs in series, Marshall quads for example come to mind.

[cbarrow7625]

Wait a second. If one of a parallel pair of drivers dies, the impedance of that pair changes to that ofbhe single remaining driver; 8 ohms. You are left with 8 ohms in series with 4 ohms. That is 12 ohms, not 6.

The voltage across the 8 ohm speaker will increase, but since the impedance of that driver, and the system, also increase there should be no significant change in power dissapation in that driver. The other pair of drivers would receive less voltage and have lower dissapation as well.

3 cones moving air at a lower power draw vs. 2 cones moving air becomes somewhat of a statistical dead heat. One will be slightly more than the other, but shouldn't be enough to easily detect with thebhunan ear.

The fundamental mistake here is assuming one configuration changes to 8 ohms and the other to 6 ohms. The series pairs paralleled does change to 8 ohms but if one driver goes out in the parallels put in series, the I mprdance goes up to 12 ohns, not down to 6 ohms.

Long story made short, they are both equally acceptable and safe under the condition described.

[Sowhat]

Wait a second. If one of a parallel pair of drivers dies, the impedance of that pair changes to that ofbhe single remaining driver; 8 ohms. You are left with 8 ohms in series with 4 ohms. That is 12 ohms, not 6.

Ah, the example given uses 4 ohm drivers, not 8 ohm. when 1 speaker in a parallel pair goes out you get 4 ohms, the other parallel pair is 2 4 ohm speakers in parallel so that's 2 ohms, 2 ohms plus 4 ohms in series is 6 ohms... providing I got my math right.

The consideration you make about the power dissipation is something I need to mull over, sounds plausible but JM's math looks correct to me as well, I have not double checked it to make sure.

This simple speaker hookup exercise has a lot more dimensions to it than one would initially think.

[Leo_Gnardo]

Besides the arithmetic, another factor in having a dud speaker in any multispeaker cab is, unless the voice coil is locked into place, the cone will continue to move but in opposition to the working speakers, driven by air pressure. So it more or less cancels one of the working speakers and makes the cab sound duff. Besides making rattling noises depending on what kind of failure it suffered. I've even encountered this with an Ampeg SVT 8x10 cab which is sectioned off into four 2x10 cabs. One bad speaker, and it ain't what it oughta be... Here's hoping to no blown speakers.

Good to see you posting Sowhat, you're too rarely seen around here.

[bob p]

when one driver opens it's voice coil, doesn't it just become a passive radiator?

[big_teee]

So if any one of these speakers in the diagram dies what impedance would you have?
https://www.google.com/search?q=how+...tm%3B326%3B289

[Tom Phillips]

If a single speaker fails open circuit then you would have 24 Ohms.

[flyingdutchman]

what impedance would you have?

Assuming the speaker becomes an open circuit when it dies, the total impedance would become 24 Ohms rather than 16.

That said, I've seen quite a few dead speakers that weren't open. I've recently had a 15" cheap chinese subwoofer that failed dead short.
Often, the voice coil just overheats and gets stuck. What this does to the speaker's impedance is anyone's guess.

[Sowhat]

Good to see you posting Sowhat, you're too rarely seen around here.

I only chime in when I got something to say, be it something stupid or just mindlessly ignorant. I get real busy most of the time with all the busted gear out here in SoCal. Only once in a great while does something dawn on me to be worthy of sharing in such a respected, trusted and esteemed forum... and I really mean that. MEF is the best no nonsense repository of music gear wisdom anywhere on the planet. It's bailed me out countless times.

[J M Fahey]

Wait a second. If one of a parallel pair of drivers dies, the impedance of that pair changes to that ofbhe single remaining driver; 8 ohms. You are left with 8 ohms in series with 4 ohms. That is 12 ohms, not 6.

The example used 4 ohms speakers.

The voltage across the 8 ohm speaker will increase, but since the impedance of that driver, and the system, also increase

Why would the impedance of each driver increase?

there should be no significant change in power dissapation in that driver.

But its dissipation *will* increase, please check the Math above.

The other pair of drivers would receive less voltage and have lower dissapation as well.

True, and Math confirms that.

3 cones moving air at a lower power draw vs. 2 cones moving air becomes somewhat of a statistical dead heat. One will be slightly more than the other, but shouldn't be enough to easily detect with thebhunan ear.

In the ecample, 2 working cones will receive 100W and movehalf the air as before.
To understand this better, consider the original cabinet as made out of 2 "half cabinets", each dooing 50% of the work.
Disabling one of them will leave the other still doing its 50%.

While connected the other way, the power loss would be somewhat less but power gets concentrated in a single speaker which will soon collapse.

The fundamental mistake here is assuming one configuration changes to 8 ohms and the other to 6 ohms. The series pairs paralleled does change to 8 ohms but if one driver goes out in the parallels put in series, the I mprdance goes up to 12 ohns, not down to 6 ohms.

Sorry but your fundamental mistake is erring the impedance value .

Long story made short, they are both equally acceptable and safe under the condition described.

Sorry but they are definitely not equally safe.
In fact the connection you suggest is inherently unsafe.

Please recheck the Math above.
It's not a matter of opinion.

[Enzo]

I have never considered one of the drivers in a 4x12 going open to be a major concern. I can't say I have seen it happen very often in my career.

[J M Fahey]

I have never considered one of the drivers in a 4x12 going open to be a major concern. I can't say I have seen it happen very often in my career.

Maybe I work with wilder musicians, because I have, many times.
*Specially* when they couple a "vintage" cabinets (80W RMS) with a 100W head and drive it through the roof with distortion, dropped tunings, etc.
These are my customers "La Renga" playing in Buenos aires "River Plate" Football (soccer) Stadium for 48000 people.

You see 4 x 100/120W heads (a Marshall, a Mesa Boogie and 2 Matchless) and 4 stacks.
Second and third, "Modern" , black front cloth, have 4xG12T75 each, total 300W RMS, you'll *hardly* see one of them blown, but cabinets 1 and 4, "vintage", "salt and pepper" checkered cloth (appears grey from a distance), have 4 "Grenbacks", 20W RMS if original; 25W RMS each if reedition.
As they print in the label "25W .... period" because they use old technology untreated brown Kraft paper coils , glued with nitro lacquer.
If you use one of these cabinets with a 100W head, full tilt (specially the non master volume ones), *definitely* you won't finish your show , not even a rehearsal.

[The Dude]

One thing that hasn't been discussed yet is phasing. When a speaker blows does not the cone movement become opposite of the working speakers in a sealed enclosure- i.e. passive radiator? I would think this effect would be less noticeable with only one nonworking driver as opposed to two. An argument for not using two series loads wired in parallel.

Edit: After further reading, I see this HAS been discussed/pointed out already......My bad. (Sorry Leo!)