Question about speaker specs: Easy vs. hard to drive speakers

[NewHTbuyer]

I have a basic understanding that a speaker rated 4 Ohms requires more current than an 8 Ohm speaker. Also, a speaker with lower sensitivity requires more power to reach the same volume as a speaker with higher sensitivity.

My question is, how much does the amount of bass a speaker can reproduce add to the load the amp sees? Everyone says that producing low and ultra-low bass requires the most power, which is why subs come with very powerful amps, often 1000 watts + for top of the line units. So, wouldn't it follow that a tower that can dig deep will put a bigger strain on an amp than a bookshelf or a tower with less bass?

Speaker A)
bandwidth: 28Hz-32kHz (+/-3dB), nominal impedance 8 Ohms, sensitivity of 88dB(2.83V @ 1meter)

Speaker B)
bandwidth: 42Hz-20kHz (+/- 3dB), 4 ohms nominal, 6 ohms average, sensitivity also 88 dB (2.83V @1 meter)

Speaker C)
bandwidth: 50Hz-24kHz (+/- 3dB), 4 ohms, sensitivity of 86 (2.83V)

So, does speaker B need a better amp b/c of the lower Ohm rating, or speaker A because it can dig 14 hz lower?

How about speaker C, very low impedance and sensitivity, but much less bass?

Also, I sort of understand that the sensitivity measurement is not directly comparable because the impedance is different. If I am right, for speaker B, 2.83V = 2 watts while for speaker A it equals 1 watt. So does that mean A is actually more sensitive?

Thanks for the info! Always trying to learn a bit more about this hobby.

 

[avengineer]

Impedance specs are often misleading because they are quoted as a single number, but in reality impedance changes with frequency, often radically. "8 Ohms" is a "nominal" rating, but that same speaker could dip to below 4 ohms and peak higher than 16 ohms depending on many aspects of the design. As a result, the demands a speaker places on an amplifier also change with frequency.

Here's a plot of an "8 ohm" speaker:



Separate from impedance is sensitivity, which can more often be thought of as a single number, even though it too changes with frequency. The goal of flat response, though, sort of makes a sensitivity as a single number more usable. However, sensitivity should always be quoted as dB SPL @ 1 meter for 1 watt input. That way, regardless of impedance, speakers can be compared. Quoting sensitivity as an SPL at 1 meter based on a voltage input is quite an indirect way to get there, since you'd have to back into the required power (as you have done).

But to answer your question about speakers with extended bass capability being more demanding on an amp...not necessarily. Again, the speaker is a complex load, and what happens to its impedance at the low frequency end is what would tell the tale. You need an impedance plot to know.

Comparing a powered subwoofer to a tower, especially in the aspect of amplifier demands, is probably not very worth while. A subwoofer is a very special case of a speaker, and to get extended low frequency response at high SPL it may be designed with an integrated amplifier capable of very high power. That requirement could be partially based on an inefficient design, but one that has advantages in overall performance. It's not necessarily true that any speaker with extended low end has a lower impedance that would be harder for an amp to drive. Here's a plot of a speaker with extended response, but not a terribly low impedance at any point:



You can probably find examples of any combination of impedance, sensitivity and frequency response.

 

[Steve81]

It's not necessarily true that any speaker with extended low end has a lower impedance that would be harder for an amp to drive.

Here's another example:
Pi Speakers - three Pi loudspeaker performance data

You can probably find examples of any combination of impedance, sensitivity and frequency response.

Within reason at least. I could power this thing to blistering levels with just a spare channel off my 7 ch amplifier. 115dB @ 2 meters outdoors with a 3dB down point of ~32Hz and a 21.53V input (~116W into the nominal 4 ohm load). I'll take two (over my wife's dead body)

 

[Stanton]

I think you got some good answers, but I also think you get the most important concepts:

1) You have to pay attention to speaker impedance when buying/matching an amp (or the amp may not work for long).
2) Speaker sensitivity will affect how much you have to turn that amp up to achieve a given volume.

Of course, a "good" amp should mean you don't have to worry about either one in the long run (but that costs $$).

 

[NewHTbuyer]

Thanks a ton for the answers. So, if I am looking at an impedance plot to try to analyze how hard the speaker is to drive, should I pay particular attention to the impedance at the lower frequencies?

Also, what measurements do you look at to see if an amp can handle a certain speaker? How can you tell if it has enough juice for a low impedance speaker, besides the manufacturers claim that a unit is "4 Ohms rated"?

Here is a real world situation, my Ultra towers and HK3490 receiver.

Here is the impedance plot for the Ultra Towers (from the S&V review):

Test Report: SVS Ultra Tower Surround Speaker System | Sound and Vision Magazine

The blue trace is the tower. It drops below 5 Ohms between 200-500 Hz and then again between 60-150 Hz.

Here are the measurements for the HK3490:

harman/kardon HK 3490 Measurements and Analysis | Audioholics

# of CH	Test Type	Power	Load	THD + N
2	CFP-BW	> 100 watts	8-ohms	0.1%
2	1kHz Psweep	110 watts	8-ohms	0.1%
2	1kHz Psweep	116 watts	8-ohms	1%
2	1kHz Psweep	167 watts	4-ohms	0.1%
2	1kHz Psweep	178 watts	4-ohms	1%
2	Dynamic PWR	140 watts	8-ohms	1%
2	Dynamic PWR	227 watts	4-ohms	1%

<tbody>
</tbody>

So, would these units be a good match? Which of the HK measurements should I focus on?

Thanks!

 

[ski2xblack]

I would look at the HK's 4 ohm ratings since the towers present that in the power hungry lower registers. (My guess is the HK is a fine choice for those speakers.)

Nobody has mentioned phase angles, which can contribute greatly to how demanding to drive a speaker can be.

Understanding Ohm's Law, Impedance And Electrical Phase 101 | Audioholics

 

[Irvrobinson]

I think most people worry too much about amplifier power and whether speakers represent a difficult load or not. If you have very capable speakers in a large room worrying about the relative difference in how difficult speakers are to drive can be interesting, but for most speakers in most rooms I don't think it is. If you choose an amp or a receiver rated into 4 ohm loads you'll probably be just fine with it. There are extraordinary cases, like Gene with his Titus 8Ts where he's thinking 450w/ch isn't enough, but most speakers in rooms of less than 3000 cu-ft just aren't going to be that power-hungry.

I can't recommend enough getting a good sound level meter and figuring out how loud you like to listen. If you're listening at average levels of 85db or so, with peaks of less than 100db, worrying about amplifier power is probably unwarranted, unless you have weird speakers, like some old Apogees or Legacys, or those silly Wilson WATTs. On the other hand, if you have a large room and you want to hear clean peaks of 105db or more, like Gene, well, my advice is just to go for overkill in an amp, like 300 watts+/ch, and just worry about what speakers you like. High power doesn't have to be all that expensive these days, with products like Crown and Emotiva around.

Subs are a different story altogether, because of movies with sound effects. As someone posted a couple of years ago, the right amount of power for a subwoofer is what it takes to push the driver to maximum excursion, which can be a couple of thousand watts or more. The resulting output levels, like 120db at 40Hz, look good in Audioholics tests, but I wonder how many people push their subs that hard. I know I don't, but I have a music-only system. Bass also has subtlety that higher frequencies don't, because you feel bass as much or more than you hear it, and while hearing is logarithmic sense of touch isn't, it's linear. The nice thing about subs is that you can buy a lot of bass power for under $1K these days.

 

[avengineer]

Pretty much over thinking this whole thing. Any responsible speaker manufacturer won't build something that the average amp can't drive. The concept of "pairing" is best reserved for wine and foods. And the concept of frequency ranges that are power hungry is largely fiction, especially with full range speakers (subwoofers are a special case, as mentioned before).

But if you want to over-analyze, look at the impedance plot and see if it dips below the minimum load rating of your amp over any significant range of frequencies (it doesn't, by the way), and if not you're good to go. You're not going to over-heat or burn out anything under normal use. If you rock out into solid amp clipping, that's not normal.

The only thing worth taking a quick look at is speaker sensitivity/efficiency vs available power. If your system can play at 75-80dB SPL with 20dB of headroom before the onset of distortion, you're pretty much fine. If you want to hit the THX spec, you want to be able to hit 105dB SPL on short peaks without distortion to play movies at reference level (115dB SPL for the sub). But that would be a high performance system, and typical home viewing is done at 75dB average, with 20dB of headroom to 95dB.

Gather your figures and use one of the on-line home theater SPL calculators.

 

[PENG]

# of CH	Test Type	Power	Load	THD + N
2	CFP-BW	> 100 watts	8-ohms	0.1%
2	1kHz Psweep	110 watts	8-ohms	0.1%
2	1kHz Psweep	116 watts	8-ohms	1%
2	1kHz Psweep	167 watts	4-ohms	0.1%
2	1kHz Psweep	178 watts	4-ohms	1%
2	Dynamic PWR	140 watts	8-ohms	1%
2	Dynamic PWR	227 watts	4-ohms	1%

<tbody>
</tbody>

So, would these units be a good match? Which of the HK measurements should I focus on?

Thanks!

You should probably focus on the 4-ohms 0.1% THD+N but as others have suggested you should really get a SPL meter and find out what is the loudest average SPL you normally listen to and the maximum you would go or tolerate and then report back with room demensions before we can comment on how much power would be enough for you. It is hard to have too much power but easy than you think to have enough power plus adequate head room. For example, the 3490 will be enough for my speakers in my room for my taste, though all my receivers and amps are more powerful than that.

 

[avengineer]

....should really get a SPL meter and find out what is the loudest average SPL you normally listen to and the maximum you would go or tolerate and then report back with room demensions before we can comment on how much power would be enough for you. It is hard to have too much power but easy than you think to have enough power plus adequate head room. For example, the 3490 will be enough for my speakers in my room for my taste, though all my receivers and amps are more powerful than that.

SPL meter is not necessary. Nearly all AVR's power output is within a 3dB window, between 90 and 180 watts. Everyone listens, on average, at 8 to 10dB below reference (85dB SPL), that puts us at 75-77dB SPL. Listening at "reference" nearly always seems too loud, except in a very large room. There are no home theater speakers made that, when connected to any modern AVR, wouldn't produce that adequate SPL in an average size room. We run into trouble only when the room, and specifically, the distance from speaker to listening position, is very large, and the speakers are very inefficient, and we are trying to play at reference or higher. Remove any two of those exceptions, and things will work out just fine.

Here's a good calculator:

Peak SPL Calculator

Here's a particularly bad example:

Speakers sensitivity: 88dB
Amp power: 90 WPC
Distance to speakers: 10'

Max SPL: 102.6

As you can see, that combination is on the edge of adequate, but if you increase your amp power to the high side of the typical AVR power, like 140 watts, you only barely hit the THX recommendations.

So, again, it's not amp/speaker "pairing" at all, it's speaker, listening distance, and room size. And, if you don't plan to listen at "reference" the entire picture changes. But in any case, your average power is barely 1 watt...yes, 1 watt..with peaks rarely passing 10 or 15 watts. So no problem with any amp driving any load at that level.

 

[NewHTbuyer]

I get about 100-104 dBs using that calculator. I appreciate the extra info. Nice to know that most of the time it is not necessary to worry too much if the amp is decent quality and the room is not too big.

Thanks for all the help interpreting the measurements!

 

[Irvrobinson]

SPL meter is not necessary. Nearly all AVR's power output is within a 3dB window, between 90 and 180 watts. Everyone listens, on average, at 8 to 10dB below reference (85dB SPL), that puts us at 75-77dB SPL. Listening at "reference" nearly always seems too loud, except in a very large room.

"Everyone" doesn't listen at 75-77db. I know people that listen louder to some types of music, and some who think they listen loudly, but listen at lower volumes than that. An SPL meter is a cheap tool to learn something about one's system and habits. Nonetheless, I'm with you on the importance of the first watt. Depending on the genre of music, I sometimes listen at mid-80s db averages. I know rock and roll junkies who listen even louder. I'm not one of those people, but I know some.

 

[avengineer]

"Everyone" doesn't listen at 75-77db. I know people that listen louder to some types of music, and some who think they listen loudly, but listen at lower volumes than that. An SPL meter is a cheap tool to learn something about one's system and habits. Nonetheless, I'm with you on the importance of the first watt. Depending on the genre of music, I sometimes listen at mid-80s db averages. I know rock and roll junkies who listen even louder. I'm not one of those people, but I know some.

Perhaps "everyone" wasn't the best...but on the average, in the home, it's 75-77dB (with 20dB of peak headroom), with occasional averages up to 85dB. It's a bell curve, and there are always extremes, but that's not the point. Cheap SPL meter just give people more ways to obsess about things they can't fix anyway, and if you consider they are of questionable calibration, sometimes off by 2dB, what's the point? On the other hand, amp specs and speaker specs are usually measured with calibrated gear, so the calculator is good enough.

The SPL meter is a good learning tool, though, so long as people understand the limitations of cheap meters.