Explain Like I'm 5: Speaker impedance/efficiency

Greg Nal

Audio Wizard Wannabe
Subscriber
I'm kinda new to the audio thing and trying to wrap my head around so many new facts and theories. One that has somewhat eluded me is the whole concept of speaker efficiency and what amps pair well with what speakers. Short of trial and error, is there a short course of discourse to help brighten the dimness of the light bulb over my head when it comes to speakers? Once I get through this one I'll have another on phono cartridges but that is for another day. C'mon AK'ers, give me your best shot!
 
Typically the more efficient the speaker the fewer watts are needed to drive it to the same sound pressure level(SPL).

So a speaker with 99db efficiency powered by a 10 watt amplifier can play as loudly or louder than a speaker rated at 89db powered by a 100 watt amp.

Impedance has no relation to efficiency, it represents the "load" the amplifier sees. The higher the impedance the easier it is on the amplifier but higher impedance reduces total available power. For example an 8 ohm speaker can get the full 100watts from an amp rated for 8 ohm speakers. With a pair of 16 ohm speakers the amp will put out only about 50watts. A 4 ohm speaker may get up to 200 watts from the amplifier provided it is rated to handle a 4 ohm load. Rarely does an amp actually double it's output into a 4 ohm load, but better quality amps can get close.
 
With respect to vintage speakers, how do you determine/where do you find the efficiency rating? I have been slowly acquiring a number of well thought of 60's-70's speakers and am not aware of a resource to look up the efficiency. is there a site like hifiengine for amps/receivers or fmtunerinfo for tuners, for speakers?
 
Sensitivity (efficiency) is measured in dB (decibels) at 1 watt at 1 meter.

SPL (sound pressure level) is logarithmic, not arithmetic. An increase of 10dB will double the perceived volume.
 
With respect to vintage speakers, how do you determine/where do you find the efficiency rating? I have been slowly acquiring a number of well thought of 60's-70's speakers and am not aware of a resource to look up the efficiency. is there a site like hifiengine for amps/receivers or fmtunerinfo for tuners, for speakers?

Have a look at any labelling on the back of the speaker. The "sensitivity" will give you the number so that you have something to compare.
 
With most vintage speakers I have acquired, they're one foot out of the grave. Rare to have labeling intact after 50 years. I am one-by one restoring the exterior, re-coning/capping as needed. I am interested in comments I read about pairing speakers with amps (SS vs Tube) as many of my speakers are early AR, Advent, KLH, Dynaco, Altec, Klipsch with some 70's JBL's and a few odd ducks thrown in for good measure. I am working from the top down-I inherited Marantz 7/8b duo from my dad and have been trying to do some comparisons on speakers/receivers with Pioneer, Harman Kardon, Marantz 22xx, some McIntosh SS and Fisher/Scott tube. I want to develop my "ear" and am trying to understand the electronics behind the music, bit by bit. I have some background--a long dormant EE degree/emphasis on digital circuit design that has not connection with my carreer of 30+ years. Truly clearing out the cobwebs as I move into this hobby/obsession. The WAF has been an issue as boxes continue to arrive.
 
Speaker sensitivity and impedance, it should be understood, are quantitative attributes which provide no indication of sound quality. They are useful in helping the owner determine how much power he may need for his application (room size, SPLs desired, etc) but reveal nothing about sound quality, per se. Luckily, we are born with the best meters for assessing SQ.
 
index.php
 
If I hook up set A (Spendor SP1/3, 8Ohms) set B (MB Quart MCS 390, 4ohms), can I do A+B? Will it fry my amp(Sansui AU-D717)?
 
What impedance is the minimum the amp can safely drive?
8 and 6 ohm A+B combo would be around 6 ohms.
 
So a speaker with 99db efficiency powered by a 10 watt amplifier can play as loudly or louder than a speaker rated at 89db powered by a 100 watt amp.

So, unless you find a 110dB sensitive speaker, it really won't get louder because you simply cannot throw X amount of watts to a speaker, there is a limit to watt it can handle.
 
Sensitivity (efficiency) is measured in dB (decibels) at 1 watt at 1 meter.

SPL (sound pressure level) is logarithmic, not arithmetic. An increase of 10dB will double the perceived volume.


Sensitivity and efficiency are related but not the same thing. Often used that way in discussion though, sorta like phase and polarity. Not the same but often (incorrectly) used synonymously.

Perhaps the OP cannot find the efficiency of the speakers because that is a rarely-published specification.

Instead, the OP should look for the sensitivity specification as you subsequently informed.

(FWIW, a 90dB speaker is only ~0.63% efficient.)
 
Last edited:
Indeed, there are factors for consideration but the point is a good one for conceptual understanding of how sensitivity and power can interact.

10dB higher sensitivity is, in basic concept, like having 10x more power.

So, back to that point, 99dB speaker with 10W and 89dB speaker with 100W would, theoretically, produce the same net SPL.
 
With most vintage speakers I have acquired, they're one foot out of the grave. Rare to have labeling intact after 50 years. I am one-by one restoring the exterior, re-coning/capping as needed. I am interested in comments I read about pairing speakers with amps (SS vs Tube) as many of my speakers are early AR, Advent, KLH, Dynaco, Altec, Klipsch with some 70's JBL's and a few odd ducks thrown in for good measure. I am working from the top down-I inherited Marantz 7/8b duo from my dad and have been trying to do some comparisons on speakers/receivers with Pioneer, Harman Kardon, Marantz 22xx, some McIntosh SS and Fisher/Scott tube. I want to develop my "ear" and am trying to understand the electronics behind the music, bit by bit. I have some background--a long dormant EE degree/emphasis on digital circuit design that has not connection with my carreer of 30+ years. Truly clearing out the cobwebs as I move into this hobby/obsession. The WAF has been an issue as boxes continue to arrive.

Many older speaker companies did not publish sensitivity specs and it takes a bit more searching to find them. Usually an old stereo magazine test of that model will have those ratings. Also when one Googles the brand/model a lot of hits come up for this site with the info needed.

A good example of this is a page I found on eBay about some speakers I own, the Imperial 6. It was a dealer promo printed up by Marantz of a review in a magazine.

s-l1600.jpg

s-l1600.jpg
 
So -- here's hoping that the OP's fairly clear on sensitivity. Sensitivity is a measure of how loud a loudspeaker will play, expressed in sound pressure level (in decibels, relative to some standard), at a given distance from the loudspeaker, with a given level of input power. Efficiency is a related concept -- how much of the electrical power input is converted to acoustic power output. Speaker drivers are typically very inefficient transducers of power -- from a few to several 10s of percent of electrical power are converted to acoustic power by most speakers.

A quick scan through this thread suggests no mention of the somewhat arbitrary nature of nominal impedance ratings of loudspeakers. Impedance varies as a function of frequency, and the two non-constant components of impedance (capacitive and inductive reactance) don't vary the same way as a function of frequency -- thus the other important, frequency-dependent parameter is phase. Objective (quantitative) tests of loudspeakers will usually present graphs of impedance and phase angle as a function of frequency.

The nominal impedance is usually taken from the relatively flat (invariant) region a couple of octaves above the loudspeaker's resonant frequency (Fs). This is often somewhere in the neighborhood of 100 to 200 Hz. The nominal impedance is not necessarily a measure of how easy a loudspeaker is to drive for any given amplifier!

A lot of the "magic" ascribed to the pairings of certain loudspeakers and amplifiers boils down to the parameters mentioned above, and how well they complement each other, I do believe.

HTH, as they say.
 
Last edited:
Here are a few not entirely randomly chosen loudspeaker impedance and phase curves from Stereophile magazine (one of the last remaining publications that actually publishes thorough quantitative analysis of most of the stuff they test).

The solid line is impedance, the dashed line is phase. The left Y-axis is the impedance value (in ohms), the right Y-axis is phase angle (in degrees).

This one has a minimum impedance of just about 4 ohms.


1107Focfig1.jpg


This one ("Neko Audio Acoustic Zen") dips below 2 ohms at high frequencies; I suspect these would be hard for some (maybe many) amplifiers to drive without issues.
stereophile_impedance1.png


This one has its impedance minimum at about 4500 Hz.
LL


This one has its minimum in the region of a few hundred Hz (a little less than 4 ohms).
1208SA7fig1.jpg


The main point is to illustrate how different these loudspeakers are as electrical loads for any given amplifier. The "nominal impedance" values for these (in fact, most!) loudspeakers really doesn't say much about this important property of speakers.
 
Indeed, there are factors for consideration but the point is a good one for conceptual understanding of how sensitivity and power can interact.

10dB higher sensitivity is, in basic concept, like having 10x more power.

So, back to that point, 99dB speaker with 10W and 89dB speaker with 100W would, theoretically, produce the same net SPL.
:)
 
its actually a failrly involved concept, to address the OP, and I understood it well, converting between domains at will when I was but a ute studying EE in college, but impedance Z is reistance R plus reactance (jX) and the j part is complex, some even say imaginary due to the use of 'j' in the equation.

suffice to say, resistance R is easy, hook up a DVM and read what its says. that was simple.

but something magical happens when a copper wire is coiled and current is passed thru it, and it gets even more magical when done in the presence of ferrous materials and/or magnets.

because of time shifts due to current starting or dying in the coil we get this whole imaginary portion and unless you were good at calc and ODEs, dont worry about it.

what to do? well manufacturers used to publish an impedance curve for a speaker or loudspeaker setup (more than one driver plus crossover) and it has some nominal (read: as close to average as we can figure) rating and you use that to match up to your amp.

whereas some curves show a dangerous dip of impedance for certain frequencies, that is only a 'holy crap moment' if your listening tastes run to continuing tones varying very slowly. sort of like a psycho hearing test. at any point in a recorded passage of <insert fav song here> there are many tones present and hopefully the crossover is routing them to where they need to go. Lpads (and other stuff) protect the tweeter thingies cuz

a) it takes very little power to produce 'n' dbs of a high freq and
b) it takes a lot of power to produce 'n' dbs of a low freq and
c) copper wire wire has a very calculable point in which it burns in half therefore

high freqs are sent to the tweeter thingies at low power and low freqs are sent to the woofer thingies at high power and everyone is happy. and each is acting as its own separate Z=R+jX, in fact, the tweeters are often 180* out of phase with the woofer so we reverse the wires from the crossover (yes, sometimes its that simple)

If you want to know, as a sidebar, why more or less power for lower or higher freqs, ask yourself: whats new?

its an old physics/chem class joke from my HS days but greek 'nu' (new) is 'C/l' or C (some constant) over greek lambda, which I dunno howto make in ubb script, perhaps someone can help.

at any rate, it divulges a relationship between wavelength and frequency, which holds for, well, just about everything. in my example, we used C as the speed of light for transmitted waves/light etc. longer waves = lower frequency and longer waves in a speaker means the cone must travel further and moving cones requires: power

there, was that 5th grade level enuf to understand?

ps: EE joke I came up with at CMU, in Wean hall I posted this: -jwL=0 (where w = greek lower omega) as a sign over what was originally on the door. EE's got it.

figure it out and you are part of the club.
 
Back
Top Bottom