A Question About Impedance - trust the graph or trust the nominal rating?

[LiveJazz]

Ok - so I know that "nominal" impedance is a fuzzy topic, and the impedance varies wildly by frequency. In general though, I would think that a speaker with a reported 4 ohm impedance would have lower average impedance that a speaker that reports 6 or 8 ohm.

But looking at speaker measurements (courtesy Stereophile), I'm seeing some serious discrepancies. For example, how to you explain this:

The graph shows the impedance of the impedance/frequency of the Wharfedale Diamond 10.1, a speaker with an 6 ohm nominal imp:


And this is the same graph for the EPOS Epic 2, a speaker with a 4 ohm imp:


Now I might be crazy, but those look like they'd average out pretty close to the same, and in fact the Wharfdale dips lower than the Epos does. So, which speaker would actually be harder on an amp? Is there something else going into it? Does it even matter if you aren't playing at high levels and/or the frequency doesn't sit in a low impedance zone for long?

 

[anamorphic96]

It more or less comes down to how the speaker company wants to report it. Some are more cautious than others in how they do this. In my opinion all manufactures should state the minimum and maximum impedance. Some speakers have been known to have serious dips all the way down to 2ohms even though the manufacture rated the speaker at 8ohms. This is not very commonplace but does happen. Hope this helps.

 

[TLS Guy]

What a speaker manufacturer says the impedance of a speaker is means nothing. The best rough guide is a manufacturer quotes the minimum impedance is minimum impedance + 10%.

In general speakers that are 8 ohm in the power range below 400 Hz are not much good.

In my view a device not comfortable handling a 4 ohm load is not worth 2 cents.

 

[Pyrrho]

Some manufacturers are honest when it comes to impedance ratings, and some are just liars. But most likely, you will have no idea about a particular company. So you should trust a measured graph, not a rated nominal impedance. But you don't "average" the graph; if a speaker has a minimum impedance of 4 ohms and it is from 40-500 Hz, you have a properly nominally 4 ohm speaker. If the minimum were at 20,000 Hz, then it would likely be trivial, as there is very little in the way of music at that frequency, and also very little power is needed to drive most tweeters anyway.

With your particular example, Epos told the truth about that particular model, but Wharfedale lied about that model. So never, ever trust Wharfedale to tell you the truth about anything, because you know they lie. With Epos, they might be honest, but one should not judge a company (or an individual) as honest from one example.

Judging from the curves, I would guess that the Wharfedale would be slightly more difficult to drive, if everything else were the same with them (which is almost certainly not the case). It would also depend on the sensitivity of each speaker; if one speaker is far more sensitive than another, it will be easier to drive (if all else is equal).

 

[fmw]

Based in the graphs I would say giving the speakers a 4 ohm nominal rating is quite conservative and a 6 ohm rating is reasonable. 8 ohms would be getting into the area of lying. If the amp is comfortable with a 6 ohm impedance, it should drive either of the speakers to reasonable home theater volume levels without much of a problem. At full rated power, however, the amp would be in trouble with either one. I don't know about you but I never get to full rated power in my system.

 

[Swerd]

As others have already said, trust the impedance curve. And as Pyrrho pointed out, the minimum impedance in the frequency range of roughly 100-400 Hz generally is the source of most difficult loads for amplifiers. Only an impedance curve can show you that. A single number impedance rating, whether honest or not, tells you nothing about that.

Those curves from Stereophile are considered good. I pay attention to their speaker measurements, but little to their verbal reviews of speakers. If I recall, the measurements are done by different people than those who write the verbal review.

A good impedance curve can also reveal something else, possibly undesirable, about a speaker. In both curves you provided, there are small impedance bumps between 200 and 300 Hz. They can be caused by unwanted resonances from the woofer and its frame, or by the cabinet itself. They may or may not be audibly noticeable, but the impedance curve can alert you to their presence.

Did the reviews provide a plot of cabinet resonance? Stereophile usually does that too. It will look like a 3 dimensional waterfall curve, a plot of frequency (X axis) vs. SPL (Y axis) vs. time in milliseconds (Z axis). Instead of measuring response from the speaker with a microphone, they use an accelerometer taped to the cabinet wall, which measures cabinet vibration. If there is an irregular bump in the impedance curve between 200 and 300 Hz, look for a ridge at that frequency in the 3D cabinet resonance curve. The longer the ridge lasts, the more the cabinet "rings" at that frequency. This can color a speakers sound.

 

[Swerd]

I looked up these two reviews. The problem is not speaker impedance, but the good possibility of audible cabinet resonances. These measurements suggest that so strongly that I wouldn't bother listening to either of them.

There are abundant other choices available, keep looking.

Wharfedale Diamond 10.1

"Investigating the cabinet's vibrational behavior with a plastic-tape accelerometer did uncover a resonant mode at 234Hz, as well as others nearby at 184 and 262Hz. The 234Hz mode was strongest on the top panel, but the strongest mode on the side panel was higher in frequency, at 563Hz (fig.2)."

EPOS Epic 2

"Investigating the enclosure's vibrational behavior with a plastic-tape accelerometer revealed a strong resonance at 258Hz on the side panels (fig.2), which I could hear through a stethoscope as I played the half-step–spaced toneburst track from Editor's Choice (CD, Stereophile STPH015-2)."