Cabs with 4 speakers and wattage

[fossilshark]

When pushing 100 watts through a cab, and the 4 speakers are wired in series parralel, you dont need each speaker to be rated for 100 watts right? Only like 50 watts?

[HotBluePlates]

Right.  A good conservative rule is total speaker wattage which is double the amp's power output. So 4x 50w speakers give 200w total power handling, and a good match to a 100w amp. 

[pompeiisneaks]

This is one of the areas that has always confused me.  Does the load hit all speakers evenly?  I.e. does series or parallel matter here?  or if you have 4X50W you just get 200W of handling?

~Phil

[drgonzonm]

Check out the Dummy Load Question thread in amp tweeks, the questions answered in that thread are similar to those you are asking in this thread.

[pompeiisneaks]

Check out the Dummy Load Question thread in amp tweeks, the questions answered in that thread are similar to those you are asking in this thread.

I was actually talking in that thread and thought I understood it, but then I've seen some data there that seemed confusing to me as well, thus why I asked here...   

At any rate, I think I understand the concept, but I wanted clarification.  Basically, what I think I understand is that any speaker in the entire chain of the loop (series or parallel) takes part of the load. The problem that I recall is that in parallel the two sides dont' guaranteed have equal resistance/response/imedance and therefore one half can end up taking a bit more of the brunt of the load, thus why 2x is generally advised, because you can't guarantee equal loading. 

BUT, I may be mistaken :)

~Phil

[drgonzonm]

Check out the Dummy Load Question thread in amp tweeks, the questions answered in that thread are similar to those you are asking in this thread.

I was actually talking in that thread and thought I understood it, but then I've seen some data there that seemed confusing to me as well, thus why I asked here...   

At any rate, I think I understand the concept, but I wanted clarification.  Basically, what I think I understand is that any speaker in the entire chain of the loop (series or parallel) takes part of the load. The problem that I recall is that in parallel the two sides dont' guaranteed have equal resistance/response/imedance and therefore one half can end up taking a bit more of the brunt of the load, thus why 2x is generally advised, because you can't guarantee equal loading. 

BUT, I may be mistaken :)

~Phil

You are probably correct when looking at whats happening at the mili-sec, and sub milisec times.  Assuming that all speakers are the same model, the same week of manufacturer (let's excuse any made on Monday's ), the speakers will pretty much handle equal loads within the one or two significant figure range. 
However, if brands are mixed, or models are mixed and maybe the same model from different production cycles, I believe you would find some differences between speakers.  My head hurts just thinking about setting up the testing, much less modeling the interactions.  The data shown for response and other characteristics probably should be treated as "bogey" data in tubes. 

Unless you mix speakers with drastically different characteristics, for example resonant frequencies, shifted peaks, etc., That rule of thumb, stated by HBP, should minimize damage to an individual speaker.

NOTICE: my choice of words like assuming, should, probably, believe, etc., means my comments are qualitative at best  .  Good questions and comments 

[HotBluePlates]

... Does the load hit all speakers evenly?  I.e. does series or parallel matter here?  or if you have 4X50W you just get 200W of handling? ...

For practical purposes in a guitar amp:

Yes, No, Yes.

... any speaker in the entire chain of the loop (series or parallel) takes part of the load. ...

Yes.

... in parallel the two sides dont' guaranteed have equal resistance/response/imedance and therefore one half can end up taking a bit more of the brunt of the load ...

My recommendation assumed we're talking 4 of the same speaker.  If you have mixed speakers, the ones which are most efficient will be louder than the others and could mask them.

We're assuming the "cab for a 100w amp" isn't using 3 50w speakers and 1 10w speaker; if you did the 10w speaker would tend to blow first.  But you intuitively know that's a silly idea.

You may need to practice basic electronics figuring with some Ohm's Law puzzles.  Replace the "speakers" with "resistors". If we have 4x 8Ω resistors in series-parallel, there are two branches of the circuit with two 8Ω resistors in series.  These can be simplified to two parallel 16Ω resistors, which could again be simplified to a single 8Ω resistor as a load.

If you adjust the values of any of those "resistors" within a reasonable range, you'll find the power dissipated across the changes very little.  It's probably good practice to figure the current through and voltage across each of the resistors.  So production tolerances don't amount to much change in-use.

... in parallel the two sides dont' guaranteed have equal resistance/response/imedance and therefore one half can end up taking a bit more of the brunt of the load, thus why 2x is generally advised, because you can't guarantee equal loading. 

No.

Volts-RMS is converted to Peak volts as √2*Volts-RMS; Current-RMS is converted to Peak current as √2*Current-RMS.  Power equals Voltage * Current.  For a sine wave, average power is equal to RMS power, and the power implied by the peaks is a brief instant.

But if you distort a sine until it is a square wave equally-large (the square's peaks are as-tall as the sine's peaks), the RMS power of the square wave = √2*Volts-RMS * √2*Current-RMS = (√2*√2)*Volts*Current = 2*Power RMS of the sine wave.

You rate the speakers for 2x the rated clean power of an amp because if you distort it hard enough, it will kick out twice the RMS power to the speakers.  100w of speaker load will burn up fast in such a situation with a distorted 100w amp.

[pompeiisneaks]

Thanks that helps a ton!

~Phil

[HotBluePlates]

Oh yeah... And easy way to remember this "rule" is that old Marshall cabs typically used 25w Celestion speakers.  A single 4x12 cab (100w total power handling) was paired with a 50w amp head, while the 100w heads were used with two 4x12 cabs (200w total power handling).

... The problem that I recall is that in parallel the two sides dont' guaranteed have equal resistance/response/imedance and therefore one half can end up taking a bit more of the brunt of the load ...

This wasn't entirely wrong... If you were using a nominal 4Ω speaker and a nominal 8Ω speaker, the output voltage applied by the amp would tend to push more current through the 4Ω speaker than the 8Ω speaker.  The 4Ω speaker would wind up feeling 2/3 of the amp's output power, with only 1/3 delivered to the 8Ω speaker.  You can confirm this easily with Ohm's Law and Kirchoff...

But the minor inconsistencies between a pair of same-rated speakers isn't worth worrying about.

[pompeiisneaks]

ok that's what I remember now too that you've stated it that way.  For similar loads it is inconsequential, but for differing resistant loads it does adjust. 

Cool, thanks.

~Phil