The 312s wiring conundrum...

[MikeL]

I recently obtained a 312s cabinet in decent shape; inside were three different Eminences--a square magnet 16 ohm and two round magnet 8 ohm---however, the round magnets were of different thickness.  All sprouted "Magnalab" stickers; I found out that was a 'house brand' of Carvin's manufactured by Eminence, so no relation to the Sunn Magna speaker line. For the sake of consistency, and to upgrade, I'm thinking of picking up three of these on clearance from Avatar:
 

 CELESTION G12T100    12 inch guitar speaker. 100 watts. AKA  'Hot 100' Big clear tone with extra bass in the bottom end. Great for all kinds of styles of music especially hard rock and drop tune.  4 ohm

One $59 +  $15 shipping Two $99+$25 shipping. Four $179 + $35 shipping. Ten dollars more per speaker for 8 ohm or 16 ohm versions.

Now, would there be any problem in wiring these three(4 ohm)drivers in series, resulting in a 12 ohm load? If I powered it with a Sunn tube head with the 8/16 ohm set up, which speaker out would I use? Would it be better to kick out the extra bux and get the 16'ers for a 5.2 ohm(or whatever it is)load in parallel?  This would be used not only for guitar, but for the top end of a Sunn "Rig Of Doom" in conjunction with a couple of 2x15's. Even though the impedance would not exactly match, doesn't it work the amp a little less with a higher load anyway? Also, I do realize that putting 60 tube watts into a 300 watt rated cab(with modern, less efficient drivers)is a bit of a mis-match in itself, but I will also occasionally power it with a modern solid-state head.

[EdBass]

I think you would be OK with the series for 12 ohms, I would use the 16 ohm tap. IMO a mismatch is better down with a tube amp, I doubt you would have any problems at all blasting away with a 12 ohm load on the 16 ohm tap.
That 12 ohm load might throttle down a transistor amp a bit though.

[MammothVolume]

Get two 8s, and one 16.

Series the 8's to 16, parallel both 16 ohm loads for a total of 8. No mismatch, no issues. Just shell out the extra cash. I firmly believe in NEVER mismatching, regardless of what the internet says.

[EdBass]

Get two 8s, and one 16.

Series the 8's to 16, parallel both 16 ohm loads for a total of 8. No mismatch, no issues. Just shell out the extra cash. I firmly believe in NEVER mismatching, regardless of what the internet says.

I can think of an issue with your wiring suggestion, one speaker will get half the amps power while the other two drivers will only get 25% of the amps output. Not exactly an ideal situation; in fact it's a bad idea "regardless of what the internet says". 

I think a matched load is preferred also, but speaker impedance is rated as "nominal".  Speaker impedance rating is a somewhat nebulous number, study some factory impedance curves for drivers and you will see that the actual impedances of a given driver varies tremendously at different frequencies, and those manufacturers rating will skew even more when you load the driver by putting it in a cabinet since it’s a reactive load; it's not like the DC resistance of a resistor in an amp circuit.

Here’s another interesting tidbit; most manufacturers that offer odd numbered driver configurations have cabs nominally rated at odd impedances, and they are used with tube amps all the time. In fact, the cab being discussed here was initially a 6 ohm cabinet that was sold with the Sunn Sceptre and included a lifetime warranty.
How worried could the designers been about the impedance mismatch? 

Tube amps are generally pretty resilient, and tolerant of impedance abuse. Many tube amps even default to a dead short (ZERO impedance) as safety device if the amp is accidently run without a speaker cable plugged in.
An open jack (+/- infinity ohms) is NOT a good thing however, which is why the dead short default is used.

I have firsthand experience with amps that have run mismatched loads hard in touring situations, and are not adversely affected in the slightest, but I’ve posted those examples before and don’t really feel like belaboring the issue further; a search will find them if anyone is interested.
I wouldn’t hesitate to run any of my amps as I suggested to Mike, although I would probably go with 16 ohm drivers paralleled instead of 4 ohms in series.

[mckinnon audio]

   Hi there,I agree with Ed,I'd use 3 x 16 ohm drivers in parallel,if one dies,the other 2 keep working and the impeadance rises to 8 ohms,still usable.Good luck,Mel.

[MammothVolume]

Get two 8s, and one 16.

Series the 8's to 16, parallel both 16 ohm loads for a total of 8. No mismatch, no issues. Just shell out the extra cash. I firmly believe in NEVER mismatching, regardless of what the internet says.

I can think of an issue with your wiring suggestion, one speaker will get half the amps power while the other two drivers will only get 25% of the amps output. Not exactly an ideal situation; in fact it's a bad idea "regardless of what the internet says". 

"Bad idea" in terms of blowing them, or volume? They're hundred watt speakers (I actually have four in bass cab,) so they're pretty resilient when it comes to abuse. I'm just trying to think of ways to match the impedance to what the back of the amp says. Did I mention I don't like mismatching? 

[EdBass]

Get two 8s, and one 16.

Series the 8's to 16, parallel both 16 ohm loads for a total of 8. No mismatch, no issues. Just shell out the extra cash. I firmly believe in NEVER mismatching, regardless of what the internet says.

I can think of an issue with your wiring suggestion, one speaker will get half the amps power while the other two drivers will only get 25% of the amps output. Not exactly an ideal situation; in fact it's a bad idea "regardless of what the internet says". 

"Bad idea" in terms of blowing them, or volume? They're hundred watt speakers (I actually have four in bass cab,) so they're pretty resilient when it comes to abuse. I'm just trying to think of ways to match the impedance to what the back of the amp says. Did I mention I don't like mismatching? 

I think it’s just a “bad idea” all around. There are too many potential downfalls to list; a flawed solution, just not the way to approach MikeL’s conundrum IMO. 

The following is from the “Ask Dr. Decibel” section on the Celestion website, I found it when I was looking for the impedance curves for the drivers in question. I couldn’t find the impedance specs, or for that matter any reference to the G12T100 period.
It touches on a few of the problems with your “solution”:

“CAN I USE 8 AND 16 OHM SPEAKERS IN THE SAME CABINET?
It isn't advisable to mix speakers of different impedances in one cabinet. This is because power will not be shared evenly between the speakers and it can cause frequency dependent shifts in the power balance. This will sound terrible, can affect the amplifier and may damage the speakers. It is always best to use speakers of the same nominal impedance in one cabinet. (Similarly, it is also advisable to have extension cabs with speakers that have matching impedances).”

“Blowing them” wouldn’t be high on the list though, as long as you are mindful that one driver is taking the majority of the power.
That said, I’ll betcha I could  take out one of those 100 watt Celestion G12T100 drivers with a 50 watt amp.


OK, so I’ve made my case for why I think that mismatched load impedance isn’t a “life or death” situation for the amp…

Why do you seem to think it is?
What effect to you feel a nominal 12 ohm load on a 16 ohm tap would have vs. a matched nominal load on the tone, power, longevity, etc. of the amp?

[MikeL]

I'm guessing that there wasn't any info on those speakers on the Celestion website because they're being closed out? Anyway, I wasn't concerned with wiring for an exact match, as long as it was in the ballpark, and not resulting in too low of a load. After all, three speakers is a wacky configuration to begin with, otherwise only seen in an early Fender tweed amp, and more recently, a Fender Tone King cab(I think)and a Crate 'Vintage' or Palomino combo. There's no way to come up with an even number load unless speaker impedances are mixed, which was something I wanted to avoid due to the inequity of the power distribution between the three drivers. How did Sunn do it in the first place? In the '72 catalog it was listed at 6 ohms; did they have three 16 ohm speakers in parallel, and fudge a bit on the sum rating? There's also that thing about variances in ohm readings of different drivers; of the two 'Magnalab' replacement speakers with the round magnets that I found inside, the driver with the thicker magnet read 5.8, while the other one was 7.2; in addition, the square backed one read 14.7, and no wattage rating was found on any of them. Even though I want to keep my expenses down(plus this particular Celestion seems to fit the bill)and get the 4 ohm model, I did forget the bit about when in series wiring, one speaker goes down, they ALL go down. So, I guess I'll kick out the extra scratch and get three of the 16 ohm; the only other option in this price range is the heavy duty 16 ohm 12's that Carvin offers(irony all around, as they're manufactured once again by Eminence)but those are a little dull sounding. Unless you guys have some suggestions(the odds of finding three identical 12"s on Ebay is pretty low). Keep in mind that not a lot of rumbling low end is gonna go thru this cab; it will be used for either guitar or the high-mid/high part of a multi-cab set up in conjunction with some 15's, similar to the early/mid 70's Entwistle rig. Thanks to everyone for their info and viewpoints.

[Isaac]

"did they have three 16 ohm speakers in parallel, and fudge a bit on the sum rating?"

Yup.

[MammothVolume]

"OK, so I’ve made my case for why I think that mismatched load impedance isn’t a “life or death” situation for the amp…

Why do you seem to think it is?
What effect to you feel a nominal 12 ohm load on a 16 ohm tap would have vs. a matched nominal load on the tone, power, longevity, etc. of the amp?"

Two dead V4's in my past.  They're picky, but it was emergency situations, and both times they went south via a mismatch. An acceptable mismatch, 2:1.

I don't feel that there is a discernible difference in tone, or power. I'd just rather not play Russian Roulette with my output trannys. Can't fault me for being extra cautious. There's a reason it's printed on the amp... just sayin'.

I know this may be a small thing to think about, but there are also losses in cabling, and contacts. So your mismatch may be a bigger stretch than what you think it is. Everything has resistance of some sort, whether it's substantial or not doesn't matter, it's there.

Do whatcha want, though. 

[EdBass]

"OK, so I’ve made my case for why I think that mismatched load impedance isn’t a “life or death” situation for the amp…

Why do you seem to think it is?
What effect to you feel a nominal 12 ohm load on a 16 ohm tap would have vs. a matched nominal load on the tone, power, longevity, etc. of the amp?"

Two dead V4's in my past.  They're picky, but it was emergency situations, and both times they went south via a mismatch. An acceptable mismatch, 2:1.

I don't feel that there is a discernible difference in tone, or power. I'd just rather not play Russian Roulette with my output trannys. Can't fault me for being extra cautious. There's a reason it's printed on the amp... just sayin'.

I know this may be a small thing to think about, but there are also losses in cabling, and contacts. So your mismatch may be a bigger stretch than what you think it is. Everything has resistance of some sort, whether it's substantial or not doesn't matter, it's there.

Do whatcha want, though. 

Sorry about your loss, but I would suspect an “open” in the speaker circuit before an impedance mismatch if I were investigating a catastrophic amp failure. Usually those types of sudden death scenarios are caused by an amp running hard and seeing a “surprise” open circuit. I generally check my loads at the jack before I plug them in, just in case.
I don’t think of a reasonable impedance mismatch as Russian roulette; an open jack is Russian roulette.

Anyway, the key to this whole thing is the “reactive load” thing. Hopefully your cables, contacts, etc. (by comparison a resistance load) is insignificant enough to be pretty much a non factor in this equation; the final “jack reading” I mentioned earlier takes everything in the circuit into consideration in a DC resistance reading.
Even if you use a speaker rated for say, 8 ohms nominal load, the amp only actually sees “8 ohms” sporadically at best; it’s a reactive load. Sometimes current even actually flows from the cab into the amp, hook up a speaker to a voltmeter and pump the cone (very carefully) and see what the amp is seeing if you don’t believe me.

Yes, the output taps and jacks are marked for the nominal impedance and yes I make every effort to match my loads personally, but to effectively use the tools of the trade it’s helpful to have an understanding of what’s actually going on “between the lines”.

[tacklebox455]

you could see what the wiring specs are for 3rdpoweramps   3rdpower.com they have only 3 12 cabs

[jaywalker]

This is a different case than 3 x 12 but it demonstrates what can be done by a company that has a history of making amps that they know will be abused as rentals.
Different in the sense that we have here speakers with different "watts" and different "Ωs" combined. Just some food for thought.
Traynor has an interesting 3 speaker bass cab in current production. Model TC1510 600w.
1 x 15 (225w)
2 x 10 (150w ea.)
1" horn (40w w/L pad)
4Ω nominal impedance.
Birch ply construction, no MDF. Sounds like a plug doesn't it? I have a YBA-1 and a YGM-3 from the '70s that I'm quite fond of.
The wiring wasn't shown on the support page.
I would guess that this cab has the 2 10" combined (prob 2 x 16Ω paralleled) to make an 8Ω 300w leg which is then paralleled with an 8Ω 15" 225w leg resulting in a 4Ω nominal impedance cab. 300w + 225w + 40w horn? = let's call it 600w with the power fairly well distributed between the two legs.

[mckinnon audio]

  Hi there,I've used those Traynor cabs,and I really like them.What you don't see is a crossover inside the cab on the input board.The sinlge 15" and the dual 10"s are not receiving the same audio information,don't know the crxvr freq,but typically somewhere between 150Hz - 250Hz,and the tweeter could be 3500Hz - 5000Hz,although Peavey made a cab similar to this one,(no tweet) and the crxvr freq was 500Hz. So the 4 ohm impeadance is really the input impeadance of the crxvr,not the combined impeadance of all the drivers.Good luck,Mel.