Cabinet Impedance Mythology

[Boogie]

Speaker cab impedance is just a number you match to your amp's output, right (Or so the general consensus seems to be)? Oh, contraire! This was a lesson I learned pretty late in my musical life but pay special attention to nowadays. It made a huge difference, tonally. Here's my question...

In my experience, certain amps seem to "breathe" differently at different loads. The Super Dallas feels more relaxed at 16 ohms with my 1x12 than another 1x12 cab, different speaker, at 8 ohms (that's the best I can do). Another way to put it: it seems to be singing, comfortably, in its normal register at 16 ohms, and yelling at 8 ohms. Does this phenomenon change with the number of speakers? Or should I wire my 2x12 to 16ohms?

I have an outdoor gig tomorrow and am considering taking the 2x12 but don't want to jeopardize my "well oiled machine" of a rig without considering the impact. Ideas?

 

[Dusty Chalk]

I conducted an experiment once -- I had variable input impedance on a pair of powered monitors, so while some music was playing, I varied it through its entire range while manually keeping the volume the same through a mixer level. If you have this capability, you should do it, it's very enlightening. It goes from not sounding right, to sounding right, to not sounding right again -- I presume 'sounding right' is when my impedance matched. One side of 'not sounding right' sounded thin and reedy, but I honestly don't remember what the other side sounded like.-- it wasn't like a straight filter, it didn't just sound muffled, it was just wrong. Since then, I've always made a point when possible to match impedances, but this is more in mix'n'master-land. In guitar-land, there's certainly no problem with plugging your lower-impedance cab to a higher-impedance output jack of your amp, it'll just sound different, and it may even be a desired effect.

Is it possible to try it both/all three/whatever ways?

 

[László]

Impedance is a nominal specification at best. The impedance of a speaker varies with frequency. This is why all speakers are rated for impedance at a certain frequency.

And you have the issue damping factor in the amplifier to control the speaker's tendency to "ring", which just means that the amp controls the loudspeaker's tendency to stay in motion after the signal stops (Newton's Law of Motion, etc.).

Damping factor does this because it is a result of the amp reacting to backwards EMF from the speaker; a well-damped amp stops the ringing.

In addition to the impedance of a speaker varying with frequency, the output impedance of an amplifier varies with frequency (heck impedance in general varies with frequency). The amplifier's own output impedance is a factor affecting damping of the speaker.

Because the impedance will vary with frequency, and because different speaker loads present different impedances to the amp, I believe this will affect the damping factor to some degree.

Clearly the interaction between amp and speaker is one key to good tone, and this is a thing hi-fi enthusiasts want an amp to deal with, high damping factors being desirable for tight bass and fewer smeared sounds.

And of course, even the cables will affect the damping factor of the amp.

High damping factor - tighter bass, less "wooly" quality to the tone. Less disruption of the frequencies being reproduced. Clearer high end because less ringing.

In any case, some amps will sound best with certain speakers, at certain nominal impedances, and not with others. I didn't like my HXDA as well with the 16 ohm 1x12 I had it connected to for a while, and prefer the 2x12 at 8 ohms, but whether that is the amp's output impedance, the damping factor, the input impedance of the speaker, the fact that one is a 1x12 and the other a 2x12 in a much larger cabinet...I mean, who knows?

There are endless variables to all this, but the bottom line is that when damping factor varies, there is going to be some amount of ringing and smearing of tone, tighter or looser bass, etc., and that's going to affect how the player perceives the tone of the amp.

Then there is the whole issue of transformers, and their output taps, but how differences in these output taps can affect tone is far above my pay grade.

 

[Deleted member 5962]

In my experience, certain amps seem to "breathe" differently at different loads.

Let me preface this by saying again... I know enough about this stuff to be dangerous, and to possibly be wrong here and there. However, I know way more than a guy who doesn't know as much as I do.

This is almost always true from what I've read. In fact, I read an exact quote from a well respected amp maker who said that in essence whatever the OT was wound for, anything lower that was used was a compromise. Example, if an OT had 16,8 and 4 ohm outs, it would typically sound it's best at 16 ohms ALL OTHER THINGS BEING EQUAL, because that was the full transformer output and the lower two were early taps on it. He even said that "ideally" you would make a transformer that only had one output, that would match that amps matching cab, but that to allow for use of different cabs, multi taps were used "as a compromise." One maker (might have been the good doctor...) said that he typically only used 8 and 4 so that the 8 would be the full OT and 4 would be less of a compromise, than a full 16,8,4 would be to use in 4 ohm mode, and that 8 ohm was the most common cab impedance so he used what was best with the most cabs. He even said that in a few of his amps he did only the out output since it was simpler and better... MiniZ with 4 ohm speaker at the time had only a 4 ohm tap, or ones with an 8 ohm speaker had only an 8 ohm tap... as example.

There are a metric crap ton of variables here that make it almost impossible to simple "test and verify" this.

 

[dmatthews]

I have always been under the impression (possibly falsely) that it is OK to have a speaker with a higher impedance connected, but not lower than the amp impedance.
In other words
Amp 8ohm ----> speaker 8ohm good
Amp 8ohm ----> speaker 16ohm OK
Amp 8ohm ----> speaker 4ohm bad
Amp 8ohm ----> speaker 2ohm worse

As I understood it, the amp is expecting to have a matched or higher load. It needs the proper load in order for the tubes and circuitry not to blow. That's why you can't/should never run a tube amp with the speaker disconnected in order to have a DI on the amp output. Palmer, Radial, Rivera make loadboxes for this purpose.
I welcome a correction if I'm off base.

 

[Kred]

An engineering insight is there is no such thing as a specific impedance output transformer. The true specification is really the "impedance ratio" which means an output transformer with a 5,000 ohm primary/4 ohm secondary is the same transformer as a 10,000 ohm primary/8 ohm secondary - it is simply a transformer with a 1,250:1 impedance ratio which is the square of the turns ratio (35:1 or 36:1 in this example). The second insight is changing the speaker impedance (secondary load) reflects back a different primary impedance load to the power output tubes. So putting a 4 ohm speaker on the above transformer reflects back a 5,000 ohm load to the power output tubes and putting an 8 ohm speaker on the same output transformer reflects back a 10,000 ohm load to the power output tubes.

Generally speaking, good impedance matching maximizes power output transfer with minimum distortion. Reflecting back a higher impedance than the power output tubes are designed for (a lighter load) transfers less power (more voltage, but less current) which may sound better if it brings out the even-order harmonics. Car analogy is hitting the gas pedal in 1st gear. Reflecting back a lower impedance than the power output tubes are designed for (a heavier load) also transfers less power (less voltage, but more current) and may sound worse by overloading the power output tubes. Car analogy is hitting the gas pedal in 4th gear. Being mismatched by a 2:1 ratio may improve the tone, but may shorten power output tube life. Very little risk in trying a 2:1 mismatch, I don't recommend a 4:1 mismatch even though Stevie Ray Vaughan did it. Not connecting any speaker reflects back an infinite impedance to the power output tubes (no load) which creates such wide voltage swings (very high voltage, but no current) which may destroy the power output tubes and/or output transformer. Car analogy is hitting the gas pedal in neutral - kaboom!

 

[Boogie]

Just to clarify, I'm not talking about impedance mismatching. The amp output impedance is always the same as that of the cabinet rating...16 ohm to 16 ohm, 8 to 8, 4 to 4. Will a single speaker, 16 ohm cab "load" the amp the same as two speaker cabs with the same final impedance?

Nearly every Marshall amp I've ever seen was made for 16 ohm cabs. Sure, they may have taps for other loads on newer amps but the old 50w and 100w amps, like the Plexi, were made for 1 or 2 1960 cabs at 16 ohms each. Considering the 25th Anni, HXDA, and Super Dallas all share output transformers, wouldn't it be a logical conclusion that they, too, would be optimized for 16 ohms?

Great feedback from everyone. Thanks!

 

[László]

Nearly every Marshall amp I've ever seen was made for 16 ohm cabs. Sure, they may have taps for other loads on newer amps but the old 50w and 100w amps, like the Plexi, were made for 1 or 2 1960 cabs at 16 ohms each

This is from memory, but...

If you have two 16 ohm cabs, and a vintage style Marshall head, don't you connect each cab to the parallel outputs on the head and set the switch to 8 ohms?

I believe that's the way it's supposed to be done.

 

[Dusty Chalk]

This is from memory, but...

If you have two 16 ohm cabs, and a vintage style Marshall head, don't you connect each cab to the parallel outputs on the head and set the switch to 8 ohms?

I believe that's the way it's supposed to be done.

This is my understanding. Full stack of two 16 ohm cabs ==> 8 ohm tap; half stack (one 16 ohm cab) ==> 16 ohm tap.

 

[G-Man]

Help me here. I was always was lead to believe that you have a 100 watt amp matched to an 8 ohm cabinet you can produce 100 watts. If you dropped the resistance to 4 ohm, the amplifier would produce roughly double the wattage but become more unstable and introduce some type of distortion as the resistance drops unless the power supply can deliver additional current to stabilize the amp. Sound volume would not really change dramatically but the stability of the amplifier if it did not have a really good power supply would change - aka high current amps. I could be totally off base here.

I enjoy these posts and learn a lot so thank you guys.

 

[Kred]

A tube amp properly designed to deliver full power output into an 8 ohm load can't produce more power output into a 4 or 16 ohm load. The power output tubes idle at 60-80% of maximum plate dissipation (set by the bias) and reach maximum plate current on music peaks. Any more and they'll "red plate" and melt. Or the output transformer will melt.

Modern solid-state power amps have more current headroom and don't use output transformers, so they can push more power output into a lower impedance speaker load.

 

[dmatthews]

Thanks for the clarification Kerry, now I understand your query better.
I don't have a solid answer except to say that both my H and Custom 50 combos have 16 ohm speakers in them. PRS did not have to do that so the question in my mind now is why?
When I got my Avatar 2x12 it had two 8 ohm speakers in it and was wired for 4 ohms. I opened it up and changed it to 16 ohms.
That allowed me to set the amp to 8 ohms and have the internal and 2x12 hooked up for a 3x12.

Now you have me wondering if just my 2x12 at 16 ohms without the internal speaker might be a worthy experiment.
Hmmm...

 

[veinbuster]

Just to clarify, I'm not talking about impedance mismatching. The amp output impedance is always the same as that of the cabinet rating...16 ohm to 16 ohm, 8 to 8, 4 to 4. Will a single speaker, 16 ohm cab "load" the amp the same as two speaker cabs with the same final impedance?

Thanks for the clarification. I don't have an answer though.

Help me here. I was always was lead to believe that you have a 100 watt amp matched to an 8 ohm cabinet you can produce 100 watts. If you dropped the resistance to 4 ohm, the amplifier would produce roughly double the wattage...

No, the wattage does go up by changing the resistance. The power is what it is (as noted by Kred). More current flows through the lower resistance so the speakers take more of a beating.

 

[andy474x]

I don't understand the finer points of amplifier engineering. All I know is, as dmatthews stated above, it's generally regarded as "safe" to use a higher load cabinet than what your amp is set for (example, 16 ohm cab on an amp set for 8). It's an interesting sound, if I remember correctly, a little less volume and slightly warmer. I feel like I heard Paul say in a video that he likes to mess with the ohm matching in this way to get different tones.

 

[László]

Speaking only for myself, I think that every type of amp, speaker and cabinet will interact in different ways. Some will sound great, some will sound good, some will sound not so good. And the only way to know for sure is to experiment with them in an environment you're familiar with (this way the room issues aren't part of the equation).

 

[Deleted member 5962]

Just to clarify, I'm not talking about impedance mismatching. The amp output impedance is always the same as that of the cabinet rating...16 ohm to 16 ohm, 8 to 8, 4 to 4. Will a single speaker, 16 ohm cab "load" the amp the same as two speaker cabs with the same final impedance?

Nearly every Marshall amp I've ever seen was made for 16 ohm cabs. Sure, they may have taps for other loads on newer amps but the old 50w and 100w amps, like the Plexi, were made for 1 or 2 1960 cabs at 16 ohms each. Considering the 25th Anni, HXDA, and Super Dallas all share output transformers, wouldn't it be a logical conclusion that they, too, would be optimized for 16 ohms?

Great feedback from everyone. Thanks!

That is what I was addressing in my post. The thread has taken other, related paths, but my most of my post was directly addressing what you stated.

 

[Deleted member 5962]

A tube amp properly designed to deliver full power output into an 8 ohm load can't produce more power output into a 4 or 16 ohm load. The power output tubes idle at 60-80% of maximum plate dissipation (set by the bias) and reach maximum plate current on music peaks. Any more and they'll "red plate" and melt. Or the output transformer will melt.

Modern solid-state power amps have more current headroom and don't use output transformers, so they can push more power output into a lower impedance speaker load.

This is your answer to that question, Gman. You'll often see SS amps (especially PA Power Amps) rated something like... "100wpc at 8 ohms, 200wpc at 4 ohms, 400wpc at 2 ohms." Hence the confusion. Tube amps don't operate the same way and SS amps don't have OT's as Kred explained, so this doesn't apply to tube amps. And, not all solid state ones either, as many aren't designed to go lower and lower ohms without instability.

 

[vchizzle]

Speaking only for myself, I think that every type of amp, speaker and cabinet will interact in different ways. Some will sound great, some will sound good, some will sound not so good. And the only way to know for sure is to experiment with them in an environment you're familiar with (this way the room issues aren't part of the equation).

Agreed.

I got a little info recently from Matt King that bummed me out some. Have a 2x12(8 ohm) and 4x12(16 ohm). Had wanted to try running both cabs off 1 head, so I asked about replacing speakers in the 2x12 and wiring to 16 to match the 4x12. Then running both cabs off the Archon set to 8 ohms. I forget exactly what he said, but it boiled down to. Yes, you can and it'll be wired correct and safe but might not sound great. Something to do with the 4x12 being wired parallel/series and one cab probably would sound quieter. I'd have to dig thru my email to get it exactly right.

 

[Dusty Chalk]

It's probably because in addition to impedance, there's sensitivity -- how loud a signal will be based on the same electrical signal. This varies based on speaker, and when you wire up a cab with multiple speakers in series or in parallel, it varies even more. It's better if you have two 2x12s or two 4x12s wired up similarly, then you only have to deal with disparity between the speakers in the cabs.

The only way to determine without actually trying it -- which is easier -- is to do the math. Follow the amount of power coming out of each speaker based on the fraction of electricity they will be getting.

 

[Deleted member 5962]

Here is simple math. 50 watts into two speakers = 25 watts per speaker. Compared to 50 watts in one speaker, each speaker would be 3db quieter (half the power) than if it got all 50 watts. But since there are two of them, you double the output which= +3dB. So 50 watts into one is the same volume as 50 watts into 2. No, it doesn't sound the same, but same output.

There are other factors which can influence this such as cab resonances at certain volumes, speakers reaching max output levels, etc. One example being, if a speaker is hitting about max output, then running 2 instead of 1 will get louder because the one was maxing out and not using all the power it was provided. But the exact opposite could happen in some cases... you have a 1x12 that with X amount of watts is pushed so hard it induces a cab resonance which equals a boost at a certain freq. range, and suddenly the 1x can be louder than the 2x. So...There seem to be "exceptions" for almost everything audio...

There was some blind testing done at a forum I saw a couple years back, of the effect of adding insulation inside your guitar cab. Several guys immediately claimed the cab was not as loud. After testing with meters, it was as loud, but had less resonance and smear from waves inside the cab. After being proven that it was still just as loud, they all agreed it was "cleaner and clearer" sounding. But initially the resonance and smear from all the sound waves pounding the back side of the speaker cone were perceived as more volume and more bass. (I hope I didn't just open another can of worms. LOL)