How to read a spec sheet?

[LiveFreeAndRoam]

Take any spec sheet, and there are some things I understand, some I think I understand and others I don't have a clue about. As I read each spec, I like to ask myself two questions, "So what?" and "What does this mean to the end-user?" Maybe some knowledgeable folk can fill in the blanks and correct me where I'm off-target?

(Apologies up front, as this post is a little on the long side. Don't feel you should reply to all the points. Any contributions are appreciated.)

Take the RF-62 as an example. I've bulleted each spec below and given my brief interpreration of that spec. If I'm off-base, please don't hesitate to set me straight.

FREQUENCY RESPONSE 38Hz-23kHz ± 3dB

Frequency response describes the operating range of the speaker, and that it is flat (± 3dB) within that range. However, knowing how flat would presumably be helpful. Also, the slope of the roll-off may be of interest when fine-tuning crossovers (I think). Why not publish the curves, e.g. look at the curves in this HT-Labs review? BTW, I've always felt that ± 3dB is a huge range. That's up to 6dB delta, which is very signifcant change in volume and also the power variation (4x) to drive it. Does anyone ever publish ± 1dB figures?

POWER HANDLING 125W RMS / 500W Peak

Presumably the higer these figures, the better? I've read that this may help you select minimum power of the amplifier. One article suggested it should be at least 3dB more than the Amp's output. Other's suggest that if you never drive your amplifier to its clipping point, then this figure is not relevant. I found this site to be interesting reading.

SENSITIVITY 97dB @ 2.83V / 1m

This is possibly the most important specification and describes the efficiency of the speakers. The larger the better and the horn-tweeter largely contributes to Klipsch's high sensitivity. A speaker with sensitivity of 97dB requires half the power compared to a driving a speaker with 94dB sensitivity. This also gives the amplifier greater dynamic range as it more easily copes with sudden volume changes - e.g. an explosion. Knowing this helps to size the power-rating of the amplifier. Is there anything else I should know about this?

NOMINAL IMPEDANCE 8 ohms compatible

I believe impedance varies over the frequency range and that it can drop quite low (maybe in the 3 ohm range) at certain frequencies. Some amplifiers are not able to drive very low ohmage equipment. So what does "Nominal Impedance" mean to the end-user? Can any amplifier rated for 8 ohms drive this speaker - if not what should I look for in amplifier specs? Why not publish the curves of how ohmage varies with frequency?

HIGH FREQUENCY DRIVERS 1" (2.54cm) Titanium diaphragm compression driver mated to 90°x 60° square Tractrix® Horn

I have lots of questions on this one spec. This describes some characteristics of the tweeter. Though I don't understand the signifcance of it being Titanium. Is a compression driver better than a non-compression driver? What is a compression driver? The horn is square, but what is 90°x 60°. What is Tactrix (apart from being registered to Klipsch)?

HIGH FREQUENCY CROSSOVER 1800Hz

I think that this means that the internal (comb?) filter will send frequencies lower than 1800Hz to the woofers and higher than 1800Hz go to the tweeter. What does it mean to the end-user if the crossover point is higer or lower?

LOW FREQUENCY DRIVERS Dual 6.5" (16.5cm) Cerametallic cone woofers

This describes some charactersitics of the woofer. What is "Cerametallic"? I presume this has something to do with the stiffness and assists with clarity? The cone woofer is common enough. What other type of woofer is there?

ENCLOSURE TYPE Bass-reflex via dual rear-firing ports

I believe reflex means that when the driver pulls, that air is also expelled from the ports, and this somehow helps to improve the bass-response in both volume and frequency response? I think reflex systems are tuned, so that the response especially affects certain frequencies. What does it mean to the end-user if the enclosure is sealed, as is the case for the RC-64?

Thanks for your patience!

LFaR.

[Canadian_Dude]

LOW FREQUENCY DRIVERS Dual 6.5" (16.5cm) Cerametallic cone woofers

This
describes some charactersitics of the woofer. What is "Cerametallic"?
I presume this has something to do with the stiffness and assists with
clarity? The cone woofer is common enough. What other type of woofer
is there?

I believe these are the same thing... A bit of a big read for some people like me.


http://www.infinitysystems.com/home/technology/whitepapers/cmmd.pdf

[damonrpayne]

Hey this is really a fantastic question!

Take any spec sheet, and there are some things I understand, some I think I understand and others I don't have a clue about. As I read each spec, I like to ask myself two questions, "So what?" and "What does this mean to the end-user?" Maybe some knowledgeable folk can fill in the blanks and correct me where I'm off-target?

(Apologies up front, as this post is a little on the long side. Don't feel you should reply to all the points. Any contributions are appreciated.)

Take the RF-62 as an example. I've bulleted each spec below and given my brief interpreration of that spec. If I'm off-base, please don't hesitate to set me straight.

FREQUENCY RESPONSE 38Hz-23kHz ± 3dB

Frequency response describes the operating range of the speaker, and that it is flat (± 3dB) within that range. However, knowing how flat would presumably be helpful. Also, the slope of the roll-off may be of interest when fine-tuning crossovers (I think). Why not publish the curves, e.g. look at the curves in this HT-Labs review? BTW, I've always felt that ± 3dB is a huge range. That's up to 6dB delta, which is very signifcant change in volume and also the power variation (4x) to drive it. Does anyone ever publish ± 1dB figures?

POWER HANDLING 125W RMS / 500W Peak

Presumably the higer these figures, the better? I've read that this may help you select minimum power of the amplifier. One article suggested it should be at least 3dB more than the Amp's output. Other's suggest that if you never drive your amplifier to its clipping point, then this figure is not relevant. I found this site to be interesting reading.

SENSITIVITY 97dB @ 2.83V / 1m

This is possibly the most important specification and describes the efficiency of the speakers. The larger the better and the horn-tweeter largely contributes to Klipsch's high sensitivity. A speaker with sensitivity of 97dB requires half the power compared to a driving a speaker with 94dB sensitivity. This also gives the amplifier greater dynamic range as it more easily copes with sudden volume changes - e.g. an explosion. Knowing this helps to size the power-rating of the amplifier. Is there anything else I should know about this?

NOMINAL IMPEDANCE 8 ohms compatible

I believe impedance varies over the frequency range and that it can drop quite low (maybe in the 3 ohm range) at certain frequencies. Some amplifiers are not able to drive very low ohmage equipment. So what does "Nominal Impedance" mean to the end-user? Can any amplifier rated for 8 ohms drive this speaker - if not what should I look for in amplifier specs? Why not publish the curves of how ohmage varies with frequency?

HIGH FREQUENCY DRIVERS 1" (2.54cm) Titanium diaphragm compression driver mated to 90°x 60° square Tractrix® Horn

I have lots of questions on this one spec. This describes some characteristics of the tweeter. Though I don't understand the signifcance of it being Titanium. Is a compression driver better than a non-compression driver? What is a compression driver? The horn is square, but what is 90°x 60°. What is Tactrix (apart from being registered to Klipsch)?

HIGH FREQUENCY CROSSOVER 1800Hz

I think that this means that the internal (comb?) filter will send frequencies lower than 1800Hz to the woofers and higher than 1800Hz go to the tweeter. What does it mean to the end-user if the crossover point is higer or lower?

LOW FREQUENCY DRIVERS Dual 6.5" (16.5cm) Cerametallic cone woofers

This describes some charactersitics of the woofer. What is "Cerametallic"? I presume this has something to do with the stiffness and assists with clarity? The cone woofer is common enough. What other type of woofer is there?

ENCLOSURE TYPE Bass-reflex via dual rear-firing ports

I believe reflex means that when the driver pulls, that air is also expelled from the ports, and this somehow helps to improve the bass-response in both volume and frequency response? I think reflex systems are tuned, so that the response especially affects certain frequencies. What does it mean to the end-user if the enclosure is sealed, as is the case for the RC-64?

Thanks for your patience!

LFaR.

Frequency Response:

+/- 3db for a speaker is not a huge delta. My studio mic, which in some ways has to do far less than a speaker, has +/- .5db and I consider that to be very impressive. Considering the room you listen in can easily add +/- 20db or more, 3db from the speaker is not something to worry about. The other value of the +/-3db measurement is that it has emerged as a de facto standard usable to compare different makes and models to each other as long as the other parameters of the test are constant.

Sensitivity:

Some (not me) also value sensitivity because it means tubes or SET amplifiers can somewhat adequately drive very efficient speakers.

Nominal impedence:

Impedence curves, not for the average buyer.

HF Drivers:

Tractix horns are a type of horn, as compared to exponential horns. The HF driver materials mean some things to some people, for example some people consider titanium to be "harsh", which I don't agree with. I cannot speak to the better-ness of compression drivers but I can tell you the basics of what they are. A driver pushes air, a compression driver starts out with a wider space and pushes air through a narrower space, the result is higher pressure as the same air flowing through a wide pipe must pass through the narrower pipe. Someone else will have to fill in the "so what". I have some articles comparing tractix to exponential horns, I should get around to reading them.

LF Drivers:

Ceramettalic is a material/treatment Klipsch uses to stiffen the cone material.

"How to read a spec sheet", that sounds like an article waiting to be written by some saucy bloggers or writers for Klipsch fan websites.

[DrWho]

I think it's important to keep in mind that the spec sheets are posted

by the marketing department. They will be as accurate to the

performance of the speaker as they choose, but will always make sure to

publish specs of the speaker performing its best. With that in mind,

the specs are still able to provide us with at least some kind of

useful information. However, I think it would be beneficial to cover

some of the specs in the order of priority that PWK lived by.

Sensitivity:

PWK always believed that the efficiency of the speaker be the most

important priority in a speaker. From a broader perspective of all

feats of engineering, it always seems to be the case that the things

that work well, are the things that are simple and efficient at their

function. Loudspeakers are terribly inefficient. The reason PWK liked

horns is because they made the speakers less inefficient. I say less

inefficient because even with the best of horns you're converting most

of the electrical energy into heat (I think it's somewhere around 85%

is wasted on a good horn and like 99% on a crappy driver). Anyways, the

sensitivity of the speaker gives you a good idea as to the efficiency

of the system. There is actually more at play, but it'll only confuse

the issue.

The important thing about efficiency isn't so much that it takes less

power to achieve the same SPL. Instead, it's the general trend that

higher efficiency generally means things aren't working as hard, which

usually means less distortion (this would include both the amp and the

speaker).

Distortion:

In fact, the second thing on PWK's list is distortion. While it

partially relates to the efficiency of the system, there are also other

ways to reduce distortion that don't necessarily improve the

efficiency. PWK was a big advocate for the importance of IMD

measurements over the more typical THD measurements. IMD is pretty much

distortion resulting from the doppler effect where you have a high

frequency wave being modulated by a lower frequency wave being played

by the same driver. Basically, the more the cone moves, the more IMD

you're going to have. The nice thing about horns is that they reduce

the cone motion for the same SPL, which as a result directly lowers the

IMD of the system. IMD is also a function of bandwidth, which is where

posted crossover specs might come into play. Basically, the more

frequencies being played by the same driver, the more IMD it is going

to have. This is one of the arguments for going with multi-way systems

(less passband for each driver).

Controlled Polars:

The third thing on PWK's list is the polar response, or coverage

angles. The 90x60 spec of the tractrix horns is indicating how widely

the speaker sprays the sound into the room. Think of it like a

spotlight where you can change the focus to shine over larger or

smaller areas. PWK noted that about 90% of the sound we hear at the

listening position is actually contributed by the reflections in the

room. As a result, he advocated that the frequency response stayed the

same at every angle the speaker sprays sound into the room. If you walk

behind your speakers, you're going to notice that the high frequencies

go away, but the low frequencies are still there. This would be an

example of uncontrolled polars where the LF energy is being radiated

everywhere, but the HF energy is only being radiated over a small area.

The result of this is that the reflections in the room have more bass

information than treble information which will make the speaker sound

bass heavy. Speaker engineers are away of this fact and will voice the

speaker accordingly (drop the on-axis frequency response a bit) so as

to help offset the perceived influence of the reflections.

Also, ignoring reflections for a second - controlled polars give you a

wider sweet spot too. If the polars were something narrow, like 10x10,

then you'd have to have your head in a vice in order to heard direct

sound. Likewise, you want the polars to have the same frequency content

at every angle so that multiple people can listen to the same speaker

at the same time and still hear everything.

One of the arguments against going with multi-way systems is that you

can get better polars with fewer drivers. This has to do with the fact

that drivers take up space and will never be able to be located in the

exact same physical location in space. Also, more crossover transitions

means more distortion too.

Frequency Response:

The fourth thing on PWK's is the frequency response. The interesting

thing about the frequency response is that if you take care of all of

the above, then you usually end up with a speaker that has a pretty

normal and easy to work with frequency response. If anything funky is

happening in the frequency response, the loudspeaker manufacturer can

easily compensate for it with EQ in the speaker's crossover. I think

some people might be surprised by how often this is done and how much

EQ is being implemented (like up to 15dB sometimes). The only

significance of the on-axis frequency response is that it gives you an

idea of what notes the speaker is able to reproduce. You can pretty

much always assume that the lower limit of the frequency response is

6dB down from the rest of the spectrum (same for the higher limit too).

While certain trends in the shape and roll-off rates and all that can

influence the sound of the speaker, I would argue that they are rather

insignificant in the sense that you should never be trying to stretch

the system to its limits. For instance, if your music has heavy 40Hz

content, you probably don't want a speaker that is rated +-3dB down to

40Hz. Instead, try looking for something that digs down to 30Hz. If you

want 30Hz, then target 20Hz. The reason I say this is because the

speaker is introducing all sorts of distortions and weird things at the

limits of its response. The only downside to this is actually knowing

what kind of frequency response one needs for their music...but that's

a totally different topic.

As far as some of the other things you mentioned:

Power handling is like bragging how much gas your car burns. It's

actually a totally meaningless spec. I think the reason it still exists

is because you can assume a perfect linear model and get an idea as to

the max SPL of the system. This calculation, however, will yield a

number higher than the actual max SPL because all speakers start to

exhibit power compression. With that in mind, one can get an idea how

much headroom they have before power compression might become an issue -

and it also helps in choosing an amp if you're trying to take full

advantage of the speaker's performance.

The impedance response of the speaker is actually a pretty big deal and

is interestingly something rarely published. I think the reasons for

this is because there is a lot of misinformation regarding impedances.

There is also a lot of research that needs to be done in the realm of

amplifier to speaker interactions (which can be quantified through the

impedance relationships). One thing to note is that the "8 Ohm

compatible" rating on the Klipsch Reference lines is there because

people won't buy a speaker that says 4 ohms when their receiver says 8

ohms. Really the speakers should probably be rated as 4 ohms, but

marketing gets into play pointing out that most of the impedance band

isn't very low...there's just a few dips here and there.

Specs involving the type of drivers and horns and cabinets and whatever

are indicating that there are certain sonic attributes associated

(distortions) with

the different choices. Some people find some of these attributes more

annoying than others so it really becomes a personal decision as to

what kinds of materials and box alignments and whatever people prefer.

I think one of the most important things to do is use your ears. If

something sound good or bad, measurements (specs) are there to help

quantify what is being heard. While it's difficult to draw strong

conclusions from the typical advertised spec sheets, they can always

provide at least some indication of what to expect.

[Canadian_Dude]

So what happens if your amp or receiver can only go as low as 8 ohms and you have a speaker that goes below that? Does the the signal not get amplified properly at the frequency responses that dip too low for the amp? If it makes a difference in sound would it be noticeable to the average person?

[DrWho]

The signal gets amplified the same. Thinking of amplifiers in terms of power ratings just makes it more confusing, but ultimately lower impedances cause more current to be drawn from the amp for the same voltage level. As long as the amp can supply the extra current there is nothing to worry about. Ultimately it just means you can't turn the receiver up all the way.

[jmdore]

If distortion is the second most important priority to PWK, then why is it generally not listed on speaker spec sheets (even for Klipsch)?<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" />

Given the 125W RMS/ 500W Peak, 8 Ohm specs of the RF-62, what minimum specs do you recommend an amplifier have that would be used with this speaker?

In looking at the KS-525-THX specs, the power handling is rated at 100W max continuous (400w peak). Is this presentation consistent with the forgoing (125W/500W) or different? If different, how so?

- Segue -

The KA-1000-THX has the following specs:

FTC Rated Power: 500W into 8 ohms, 1000W into 4 ohms, 20Hz-1kHz @ <1.5% THD / Dynamic Power*: 2050W into 8 ohms, 3025W into 4 ohms @ 5% THD / *Dynamic power indicates the actual power this amplifier w (the "w" is not a typo on my part as it is taken from the web page, yet seems akin to a dangling particible).

The Outlaw 2200 M-Block has the following specs:

Power Output: 200 watts @ 8 ohms, 20 Hz - 20 kHz, <0.05% THD; 300 watts @ 4 ohms, 20 Hz - 20 kHz, <0.05%, Power Consumption: 600 watts maximum/<3 watts standby.

The 1000 has a direct/inverse relationship (proof 500x8 = 1000x4) with respect to the power rating, whereas the 2200 does not (which seems to be more common). Is this good, bad or a non-factor (please try to explain)?

How do the dynamic power and max power correlate, if at all? In addition, is the ratio of the dynamic and/or max power to the rated power (i.e. 2050/500 or 4.1) of any significance?

[DrWho]

I think the main reason distortion specs aren't listed is because the

marketing department provides the specs. Showing levels of distortion

implies that there is distortion...not to mention that it's very

difficult to interpret distortion plots (in fact, it's probably best to

just use your own ears)

As far as power specs...

Think of "max power" as maxing out on the bench press and think of

"continuous power" as running a marathon. The max power is determined

by the maximum instantaneous voltage the amp can deliver at its

outputs. The continuous power is determined by the maximum voltage that

the amp can deliver over long periods of time. The fundamental reason

that amps can't deliver their peak voltage for long periods of time is

because a lot of current flows - and current creates heat. So

basically, the continuous power is a thermal limitation. The maximum

voltage is a limit of the voltage rails supplied by the power supply.

Since amplifiers are voltage amplifying devices, the actual power

output is directly related to the load of the speaker. Take for

instance an amp that can output a peak of 16V. If you connect an 8ohm

load (speaker) and crank through Ohms Law you end up with 32W of

amplifier power output:

I = V / R = 16 / 8 = 2A

P = V x I = 16 x 2 = 32W

If you put that same 16V onto a 4 ohm load, you've got 64W:

I = V / R = 16 / 4 = 4A

P = V x I = 16 x 4 = 64W

However, if the heat sink on your amp is only able to dissipate the

heat corresponding to 3A, then your 4ohm load is going to be limited to

36W:

V = I x R = 3 x 4 = 12V

P = V x I = 12 x 3 = 36W

In other words, your maximum voltage output into a 4ohm load decreases

because if it went above 4 ohms, then it would end up deliving more

than 3A of current. However, the peak voltage output at 4 ohms would be

64W because the amp could theoretically produce 16V at the output. It

just can't sustain it because 16V would result in 4A of current, which

would slowly heat up and eventually melt the amp.

However, even though the amplifier can dissipate 3A, it can't provide

that 3A into the 8ohm load because the voltage is limited to 16V. In

order to get 3A into the 8ohm load, you would need 24V at the output of

the amp, but it can only deliver 16V.

I know I'm just crunching a bunch of numbers, but I hope this explains

why amps don't always double their power ratings when the impedance is

cut in half.

Another thing I didn't mention is that the power supply must always be

able to provide more power than the amp puts into the speakers. In our

above example, we could have a 30W power supply which would limit the

power output at 8ohms and 4ohms to both 30W. With a 60W power supply,

we would be limited to the above numbers. In this case, undersizing the

power supply is arguably

a bad design, but when you're talking high power amplifiers, generating

something like 3000W starts to get insane. If you think about it, the

power provided by the power supply also needs to be

provided by the wall outlet. With a 120V outlet and a 20A breaker,

you're limited to a continuous power draw of 2400W! So if you're going

to put out 3000W for a short period of time, then you're going to need

a way to store up energy before it's needed. Since music is transient

in nature, capacitors can be filled during the dead times with the

energy required to supply shorter duration transients. This stored

energy is often the reason why these amps can continue playing music

for a while even after the plug is pulled from the wall.

[michael hurd]

Continuing on Dr.Who's ramblings.... different types of amplifiers have differing "efficiency" ratings.

This link has an explaination of different amplifier classes. http://www.caraudiomag.com/specialfeatures/0111cae_anatomy_power_amplifier/index.html

[jmdore]

Very interesting to say the least. Very helpful as well. Thank you.

[LiveFreeAndRoam]

Just back from a 3-day camping trip. Guys, thanks so much for the info. There's been some really interesting discussion and well-made points.

To Canadian_Dude: That was an excellent paper. Thanks for the pointer.

To damonrpayne:

> "How to read a spec sheet", that sounds like an article waiting to be written by some saucy bloggers or writers for Klipsch fan websites.

Now I would love to see such an article. I hope an expert is listening.

To DrWho: That was aptly put, and especially appreciated knowing more about PWK's point of view. I'll be reading your post a few more times over to digest it all.

To michael hurd: Are you any relation to Mark [] (sorry I couldn't resist). Your pointer to the Anatomy of the Power Amplifier is much appreciated. More reading and lots of techonology to digest.

I'll likely have more follow up questions too.

Cheers!
LFaR.

[getech]

I sometimes use my reading glasses...sometimes the spec sheet will be in a larger font and I don't have to. Sometimes I just look at it and read from left to right.

Cheers!

[LiveFreeAndRoam]

>I sometimes use my reading glasses...sometimes the spec sheet will be in a larger font and I don't have to. Sometimes I just look at it and read from left to right.

Wow, the simplicity of it all. Talk about cutting to the chase! []

LFaR.