Mitch, @almarg is pretty much dead on.
There are a couple of other indicators. A square wave into a simulated speaker load, but that's usually only provided by Stereophile, as well as the damping factor. A high damping factor is an indicator of low output impedance, and, by inference, high current capabilities.
But these really are just indicators.
My personal view is that very few speakers justify being designed to be so difficult. Electrostatics one of the few exceptions. Otherwise, I'd say steer clear of "discerning" or difficult to drive speakers as they usually have some other underlying issue which a simpler design would avoid.
Best,
Erik
|
Clarification!
Sorry, I forgot. Yamaha used to, and may still, provide square wave signals in their manuals, indicating very good frequency and current capabilities into difficult loads.
Others may as well, but in the consumer market I haven't seen any manufacturer provide that level of performance detail.
Best,
Erik
|
Hi gdhal,
If you imagine a sine wave, you can measure it as peak to peak or RMS. P-to-P is just that, looking at an oscilloscipe, you look at the peak and valley and measure the difference. RMS (root means squared) tries to calculate the area under the curve, and is what amplifier watt ratings are usually based on.
So what the maker is saying is that if you had a sine wave, at maximum you'd get 22.5 amps in the + direction, and 22.5 in the negative direction.
I'm not aware of any particular standard for measuring current, so this spec is just as good as any. :)
Of course, what matters in high capacitance loads is the current at the zero voltage. That's why you will read reviewers talk about phase angle. The worst conditions for an amplifier are high phase angle (high capacitance) along with low impedance.
Best,
Erik
|
Sadly, these impedance issues don't often discouarage sales. Instead they are sold at a premium for being more "discriminating." with the idea that somehow they will reveal more in the music.
I did a thorough electro-acoustical analysis of a Focal Profile 918. And by this I mean I measured each driver electrically and acoustically as well as reverse engineered and simulated the crossover.
To simplify it, perhaps too much, the bass section had an array of 8 power resistors (8 x 10 = 80 watts) to deliberately leave the midbass with a very low impedance (3 Ohms).
Replacing these resistors and associated parts with a coil left the complex frequency response identical but raised the impedance to similar amounts as the Thiel you mentioned, around 4.5 ohms.
Since then I've noticed a wide variety of Focal speakers, regardless of
woofer or woofer count seem to share the same impedance profile, though not as blatant an approach.
Best,
Erik
|
@gdhal
So low impedance = low ohms.
8 is really high impedance, 4 starts getting lower. What you want to look at is how well it drives 4 ohms. It should be close to double the 8 ohm spec. So, 220 x 2 = 440 Watts at 4 ohms.
But more importantly, what is your amp and what are your speakers? :)
Best,
Erik
|
@unsound Take a look at my blog post about it, and scroll down to the "Bogus Woofer Cap" section. The original minimum impedance for this speaker was 2.6 Ohms. http://pqltd.blogspot.com/2016/01/focal-profile-918-ultimate-upgrade-guide.htmlThere is 1 cap and 4 resistors that are simply completely unnecessary. They are the same type of conjugate network Thiel used, but instead of being for the buyer, are for the seller. It's designed to waste up to 40 watts of power in an area where it's easy to hear. Best, Erik |
It's worth pointing out though that damping factor is not just one number. It can rise considerably in higher frequencies, where the Triton has oddly low impedance numbers.
AMT tweeters do not inherently have low impedance by the way. Cheap one's do. :)
Best,
Erik
|
@unsound
Yes, one could argue it. :) That is not a position I ascribe to very often any more though.
I can see a variety of choices in drivers, and integration between them mandating certain impedance profiles, but once I saw crossovers deliberately "juiced" to be low impedance, I stopped believing in them. That is, I stopped believing that these low impedance designs were in fact the best the manufacturer could do, but that instead they were making speakers deliberately for the crowd that feels a speaker that can tell the difference between a low and high current amplifier was more "revealing" and therefore better for music listening, which is nonsense. In fact, the less "revealing" speaker (i.e. higher impedance) design would play better under all circumstances and was only a better speaker if you wanted to hear amplifier differences.
Best,
Erik
|
@unsound
Thanks so much for using a very specific example, it makes discussion a lot easier. The example you’ve shown for a Thiel is quite benign and an example of thoughtful design.
Please note the minimum impedance is around 4.5 ohms (by eyeball). The Thiel CS3.5 isn’t lowering the impedance to 3ish ohms, but rather it’s lowering what would otherwise be pronounced impedance peaks around 12 Ohms or much higher to be more "normal." This type of network, while drawing a little more power, makes the speaker very tube friendly, while maintaining good performance with solid state amps. The frequency response of tube amps often track the speaker impedance so by flattening the impedance above the woofer's resonance Thiel was making sure the speakers would behave similarly regardless of amplifier used. Good for Thiel for thinking of this in advance.
That’s not the kind of charlatanism I’m talking about though. In fact what I'm describing is the opposite, where impedance is brought to 3ish ohms or less, and reviewers, like buyers, are amazed that the speakers can sense such a difference in amplifiers. The Thiel CS3.5 on the other hand would play very nicely with a variety of amplifiers.
Best,
Erik
|
Ahhhh.....
The GE Tritons are electrically and acoustically complicated beasties and exactly the kind of speaker design I've learned to avoid. Make yourself happy of course, though.
As for the amp, I would say that since they are excluding 4 ohm measurements and damping factor they are not particularly proud of it's output into low impedances.
Not the best match, surely.
Erik
|
Again, I defer to @atmasphere 's experience. If he says my speaker kit is more tube friendly than the Thiel CS, who am I to argue? :)
|
@atmasphere Thanks for the clarification. If you don't mind, I may quote you when I recommend the kit. :)
Since I don't build tube amps, I'll defer to your mastery of the subject.
By the way, the LM-1 is a free-to-build kit in honor of my hero, Dr. Marshall Leach Jr. All the design docs are online and I make no money at all when someone builds the kit.
Best,
Erik
|
Damping factor is usually measured as (8 Ohms / output impedance). So an amp with 100 df has an output impedance of 0.08 ohms. 20 has around 0.4 Ohms.
The higher the apparent output impedance (i.e. the lower the damping factor) the more the Voltage response will track the speaker impedance chart, instead of reamining flat.
By the way, adding feedback essentially improves damping factor. The normal way to improve df without feedback is to use massive banks of parallel output stages.
Best,
Erik
|
Very generally:
High DF = good = low output impedance.
Low DF = bad = high output impedance
Where did some one (maybe me) say high DF was bad? I may have meant to say high output impedance is bad.
|
@unsound That doubling/halving is exactly dependent on output impedance. :)
So, more feedback = less deviation = less aparent output impedance. :)
And, that's what you see in the two charts. The tube amp has more frequency dependent variability than the solid state amp.
This is the sort of thing you get into when you think about speakers that are more or less tube friendly. :) The higher, and flatter the impedance of the speaker, the flatter the amplifier's output.
Best,
Erik
|
@unsound You probably already know this, but take a look at my LM-1 speaker measurements. It is NOT tube friendly like the Thiel. Notice the 24 Ohm peak in the bass and 16 Ohm impedance peak where the woofer and tweeter response meet. http://speakermakersjourney.blogspot.com/2016/05/lm-1-bookshelf-measurements.htmlThese peaks are what the Thiel crossover was trying to avoid. On the other hand, a solid state design would play beautifully with the LM-1 Best, Erik |
Al,
Probably true! :)
Erik
|
@unsound Feedback improves DF = lowers output impedance = less variability with output voltage. :)
Output impedance can be thought of as a resistor at the outputs of the amplifier. With a perfect amplifier, this resistor is zero, and the voltage at the speaker is always the voltage at the amp.
However, amplifiers are not perfect. Batteries are like this too. Looking at a AA battery, you model it as a 1.5V source, but in reality it has a rather large output resistance. So whether your flashlight sees 1V or 1.3V has to do with how much current is drawn and how big that output impedance is.
The more old-school your flashlight, the more current, and the more V will be lost in the output impedance. But modern LED bulbs sip current, so most of the batterie's V will appear at the LED terminals.
Confused yet? :)
Erik
|
@unsound
With very high output impedance the frequency response of the amplifier looks like the speaker impedance. I.e. not flat! :)
With very low output impedance the frequency response of the amplifier stays flat.
The missing step:
Increasing feedback reduces the amplifier output impedance. That and lowering distortion are why most amplifiers have at least some feedback incorporated into the design.
|
|
