I'm looking at various driver spec sheets at the mo, and notice that very few of them quote power handling in electrical RMS watts. I'm aware (from the past) that there have been various ways of quoting power (max, peak, music, peak music etc) that can appear to double and double again, so I don't want to get caught out. Just looking at a few of the manufacturers I see:
Peerless - rated 'noise' power
Scanspeak - 100hr RMS noise test, and also Long-Term 'Max' power
SEAS - Short-Term power handling, and also Long-Term power handling
How do these various terms relate?
Thanks,
Kev
Peerless - rated 'noise' power
Scanspeak - 100hr RMS noise test, and also Long-Term 'Max' power
SEAS - Short-Term power handling, and also Long-Term power handling
How do these various terms relate?
Thanks,
Kev
I'm not saying that I want to 'use' the driver at full power, just that I want to understand what the power ratings actually are. In your example 1/4 or 1/10th of '???'
My primary concerns are things like distortion and x-max at the SPLs I'm designing for, but I would like the reassurance that I'm not going to be over-doing the power either. Probably its my inexperience, but I was assuming that its possible for a driver to be designed with good enough xmax and cone distortion compared to its heat dissipation that in some applications (e.g. a low sensitivity driver in a small sealed box) the power handling may not be completely negligible?
Cheers
kev
The terms don't always relate. There are many different ways to rate power handling, each manufacturer seems to choose it's own. Not all of them relate to music in any way.
One point to consider is that a lot of the "noise tests" are bandwidth limited noise of a certain rms voltage. Often a manufacturer will choose the bandwidth such that the impedance of the driver is above nominal and less than the expected power is dissipated. (I'm thinking Rod Elliot had an article about this, but not 100% sure.) This inflates the measurement.
Some, like Pyle, Pyramid, and others, just outright lie. The outright lie is usually only from the bottom tier though...
One point to consider is that a lot of the "noise tests" are bandwidth limited noise of a certain rms voltage. Often a manufacturer will choose the bandwidth such that the impedance of the driver is above nominal and less than the expected power is dissipated. (I'm thinking Rod Elliot had an article about this, but not 100% sure.) This inflates the measurement.
Some, like Pyle, Pyramid, and others, just outright lie. The outright lie is usually only from the bottom tier though...
Thanks, yes if they would specify that then it would be a lot clearer to me. We are talking pretty big differences between some of the quoted values, big enough to catch me out even were I trying to use only a fraction, so it would be reassuring to know.RMS Long Term is the most reliable measurement. Music Peak can be anywhere up to 4times RMS and will only last for a second or so before the speak coil cooks.
ScanSpeak's "100hr RMS noise test" is the only one I've so far seen thats kind enough to be specific about RMS (though I don't understand the actual test). In one sheet I'm looking at, its about 1/5th of their "Long term max power" rating, which seems far more of an increase than could be accounted for by simply measuring the peak/max (rather than RMS) amplitude, so perhaps that includes some sort of duty cycle averaged over an assumption of some kind of typical music use.
SEAS "long term power handling" sounds similar, but says 'power handling' not 'max power handling'..
I've no idea what the Peerless "rated noise power" is, I guess it could be white/pink noise measured at either RMS or peak amplitude. It sounds more like a measure of acoustic energy rather than anything electrical, though.
Cheers
Kev
Ah, most interesting, many thanks. I hadn't even considered that they'd measure at unrepresentative frequencies, but of course I can see why that may look better on paper.
Okay then, whilst the power rating itself would be a useful check it seems like the spec sheets could be pretty meaningless without an awful lot more detail.
Sooo... looking at it another way: Taking a worst case of fairly uniformly/continuously loud music sufficient to cause a high percentage of excursions around Xmax (but not significantly beyond), with say a lowish sensitivity driver in a small sealed enclosure... 'in practice' would power be something I ever need to worry about or is it just not an issue?
Cheers
Kev
I've been reading up on the IEC268-5 power tests; it seems the 100hr noise test is indeed the closest figure to the one that I understand/expect as a measure. Apparantly the standard specifies a pink noise signal of defined crest factor (to approximate music use); after 100hrs at the rated RMS power the driver shouldn't show "appreciable damage"... This 100hr RMS power figure is a small fraction of some of the other figures I've seen quoted, and even then "appreciable damage" is clearly not something I want to be getting near to, so it demonstrates to me just how much I shouldn't trust some of what the spec sheets seem to suggest.
Cheers
Kev
Cheers
Kev
Hi, yeah car audio has been guilty for years of inflating specs 
Only in recent times have most of the good ones started to be more honest.
Look for tests that include the lowest f they can be used at, usually fs.
Here's a couple of nice links
https://en.wikipedia.org/wiki/Audio_power
http://www.jblpro.com/pub/technote/spkpwfaq.pdf
Only in recent times have most of the good ones started to be more honest.Look for tests that include the lowest f they can be used at, usually fs.
Here's a couple of nice links
https://en.wikipedia.org/wiki/Audio_power
http://www.jblpro.com/pub/technote/spkpwfaq.pdf
The 100hrs noise test is probably the best we an hope for but for tweeters it still helps to read the small print.
Pretty much all of them use a crossover or band limited noise, the difference is in the low cut off point. These should be stated but can vary dramatically!
The lowest stated off the top of my head is Vifa with 700Hz (at least on some models) butI've seen as high as 4000Hz which in many cases makes it an almost useless number.
Pretty much all of them use a crossover or band limited noise, the difference is in the low cut off point. These should be stated but can vary dramatically!
The lowest stated off the top of my head is Vifa with 700Hz (at least on some models) butI've seen as high as 4000Hz which in many cases makes it an almost useless number.
Depends on the music.
To take my (sub-optimal) subwoofer as an example...
I have a pair of JBL GTO1214 drivers in one of those 35L pre-fab sealed cabinets. They're sharing that volume, so estimate ~16L each (once the driver displacement is accounted for). This is a very small sealed cabinet, so even with a fairly chunky amplifier (half a kilowatt), cone excursion stays reasonably benign.
I once played a 5Hz sine wave through this system. It felt weird in my small room. I couldn't actually hear much, but the cones were definitely moving. The amplifier wasn't clipping, and yet after around 30 seconds, I started to smell the warm driver smell. I immediately stopped the signal generator, took one of the drivers out and let them run essentially free-air to get them to cool a bit.
On the flip side, even the most demanding movie effects I've tried so far haven't managed to get the warm driver smell.
Things like the hospital explosion in the Dark Knight; the massive shootout in the Matrix when they enter the compound to save Morpheus; the helicopter explosion shortly after that.
The amplifier can clip and yet the drivers don't start to smell: the clipping is brief so that, while the delivered power can be rather silly, its in very short bursts.
Its the prolonged application of power that kills drivers, and, seeing as you're unlikely to run any compressors in your system, nor are you likely to run your amplifier into square waves; I'd say that as long as you make sure nothing sounds particularly unhappy, you're likely to be fine: even the 6dB dynamic range found in some modern recordings won't stick enough prolonged power to kill speakers. Within reason.
Thanks Chris, maybe I'm worrying too much then. I wasn't intending to electronically extend the response or play anything like a continuous wave through them, probably the worst they'll get is some rock at higher SPLs if I'm in the room next door..
Cheers
Kev
Very good point; I've been looking at the x-over point in relation to frequency response, but haven't so far checked how that compares to the power band. If this can begin at up to 4khz then I definitely need to give it some attention!
Cheers
Kev
More specifically the old Vifa D27TG-35-06 is measured from 700Hz, the XT25BG60-04 from 500Hz and the XT25TG30-04 from 400Hz!
The Peerless/Scan-Speak D2608 from 4kHz. Which I find odd because it seems to be generally regarded as quite robust even when crossed low. I haven't used it myself though so far.
The Peerless/Scan-Speak D2608 from 4kHz. Which I find odd because it seems to be generally regarded as quite robust even when crossed low. I haven't used it myself though so far.
I'm surprised that no one else has pointed it out, but technically RMS can't be applied to power, it is a reference to voltage. But, the term "RMS" has come to be associated with Continuous Power, and so is accept as a reference term for this type of power.
Let me use amps to explain the difference between 'RMS' and Music Power. RMS power rating are based on a sustained tone. It is not pulse or peak or transient. It can still be measured relatively quickly, but the fact is that a continuous sustained tone especially at low frequencies can drain the power supply and result in low Power ratings.
Onkyo is a good example, in the USA, their power rating are very straight forward, but in the EU and UK they are a bit sketchy. I suspect they are either using a different power supply or the power supply is not able to deliver the same SUSTAINED power with 50hz.
But music power, that is, the amp functioning under normal dynamic music conditions, performs very nicely because they don't have the drain of the sustained tones dragging on the power supply reserves.
For the record, I have two Onkyo stereo amps that I'm very happy with.
If we look at Pink Noise or other similar noises in general, you tend to get all frequencies at all levels playing randomly. It sounds like the static hiss between FM radio stations. In a sense, it is a form of random noise. However NAMED noise like Pink Noise, Brown Noise, White Noise have consistent and predictable characteristics to them. Because of that they can be used to simulate music in a consistent and predictable way.
While Pink Noise may sound like a constant HISS to us, it is actually very dynamic. Because it is dynamic, it gives the power supply time to rest and re-charge, which means it is able to recover more quickly.
Now let's apply this to speakers. First, and I'm sure people are tired of hearing this but -
It is never overpowered or underpowered amps that destroy speakers; it is always the guy running the volume control.
The best advice I can give is - don't be that guy.
RMS power ratings on speakers involve sustained tone sine waves. Music power or other power ratings might use Pink Noise or by some other means try to simulate actual music conditions.
Sustained tones cause heat to build up much faster in the voice coil and as such will lead to pretty early failure of the driver.
A simulated Music tone like Pink Noise, again is dynamic, always changing. This gives the voice coil time to rest and time for heat to dissipate resulting in a much longer time to failure.
However, as has been implied, an underpowered amp can still damage the speakers. It is about sustained volume (or voltage) over time.
The next aspect is Excursion. One very large speaker can sound significant and can reach high SPL very easily. To get the same SPL (sound pressure level; loudness) from a smaller speakers, the only solution is to push that smaller speaker harder.
So, besides electrical meltdown, the other way to damage speakers is to over drive them. This is based on pure VOLUME or Loudness. To push them beyond their mechanical limits. Typically what happens is you crank the bass to the max, then crank the volume near the max, and you push the speaker so hard, you push it forward out of the magnet.
Under these conditions, the cone is so physically distorted that the voice coil can't and doesn't neatly fall back in to the Gap. More likely it slams into the frame of the magnet and is physically damaged. The alternate is that rather than moving forward, the coil and cone move so far back that the smash into the structure of the frame and magnet and again are physically damaged.
Many people use 100w speakers on 150w and 200w amps with no problems. How and Why? Well they simply understand the limits of their equipment and stay within those limits. Good sound is not necessarily loud sound. Even at a house party, it is possible for the sound to get too loud. You can usually tell when it is too loud because people start moving out of the room.
And many people use 100w speakers on 50w amps an go through them like confetti on New Years Eve.
Continuous or 'RMS' imply a sustained tone over a sustained period of time, but that time period is not indefinite. Even at very low levels is a tone is sustained without relief it can damage a driver.
Music and other such higher rated powers are an attempt to imply how the speaker will function under real work conditions. Though they can't allow for every circumstance, such as the circumstance in which the local village idiot has a beer in one hand and the volume control in the other.
Peak Power is likely the near instantaneous failure point.
Time is a factor, a low level sustained for a long time can build up enough heat to cause damage.
Heat is a factor, how quickly it builds up and how quickly it is allow to or is able to dissipate determines how long your speaker will survive.
Mechanics are the 3rd factor. If you stretch a rubber band too far or blow up a balloon too much, they break. If you mechanically push a speaker too far, it breaks.
For what it is worth.
Steve/bluewizard
Let me use amps to explain the difference between 'RMS' and Music Power. RMS power rating are based on a sustained tone. It is not pulse or peak or transient. It can still be measured relatively quickly, but the fact is that a continuous sustained tone especially at low frequencies can drain the power supply and result in low Power ratings.
Onkyo is a good example, in the USA, their power rating are very straight forward, but in the EU and UK they are a bit sketchy. I suspect they are either using a different power supply or the power supply is not able to deliver the same SUSTAINED power with 50hz.
But music power, that is, the amp functioning under normal dynamic music conditions, performs very nicely because they don't have the drain of the sustained tones dragging on the power supply reserves.
For the record, I have two Onkyo stereo amps that I'm very happy with.
If we look at Pink Noise or other similar noises in general, you tend to get all frequencies at all levels playing randomly. It sounds like the static hiss between FM radio stations. In a sense, it is a form of random noise. However NAMED noise like Pink Noise, Brown Noise, White Noise have consistent and predictable characteristics to them. Because of that they can be used to simulate music in a consistent and predictable way.
While Pink Noise may sound like a constant HISS to us, it is actually very dynamic. Because it is dynamic, it gives the power supply time to rest and re-charge, which means it is able to recover more quickly.
Now let's apply this to speakers. First, and I'm sure people are tired of hearing this but -
It is never overpowered or underpowered amps that destroy speakers; it is always the guy running the volume control.
The best advice I can give is - don't be that guy.
RMS power ratings on speakers involve sustained tone sine waves. Music power or other power ratings might use Pink Noise or by some other means try to simulate actual music conditions.
Sustained tones cause heat to build up much faster in the voice coil and as such will lead to pretty early failure of the driver.
A simulated Music tone like Pink Noise, again is dynamic, always changing. This gives the voice coil time to rest and time for heat to dissipate resulting in a much longer time to failure.
However, as has been implied, an underpowered amp can still damage the speakers. It is about sustained volume (or voltage) over time.
The next aspect is Excursion. One very large speaker can sound significant and can reach high SPL very easily. To get the same SPL (sound pressure level; loudness) from a smaller speakers, the only solution is to push that smaller speaker harder.
So, besides electrical meltdown, the other way to damage speakers is to over drive them. This is based on pure VOLUME or Loudness. To push them beyond their mechanical limits. Typically what happens is you crank the bass to the max, then crank the volume near the max, and you push the speaker so hard, you push it forward out of the magnet.
Under these conditions, the cone is so physically distorted that the voice coil can't and doesn't neatly fall back in to the Gap. More likely it slams into the frame of the magnet and is physically damaged. The alternate is that rather than moving forward, the coil and cone move so far back that the smash into the structure of the frame and magnet and again are physically damaged.
Many people use 100w speakers on 150w and 200w amps with no problems. How and Why? Well they simply understand the limits of their equipment and stay within those limits. Good sound is not necessarily loud sound. Even at a house party, it is possible for the sound to get too loud. You can usually tell when it is too loud because people start moving out of the room.
And many people use 100w speakers on 50w amps an go through them like confetti on New Years Eve.
Continuous or 'RMS' imply a sustained tone over a sustained period of time, but that time period is not indefinite. Even at very low levels is a tone is sustained without relief it can damage a driver.
Music and other such higher rated powers are an attempt to imply how the speaker will function under real work conditions. Though they can't allow for every circumstance, such as the circumstance in which the local village idiot has a beer in one hand and the volume control in the other.
Peak Power is likely the near instantaneous failure point.
Time is a factor, a low level sustained for a long time can build up enough heat to cause damage.
Heat is a factor, how quickly it builds up and how quickly it is allow to or is able to dissipate determines how long your speaker will survive.
Mechanics are the 3rd factor. If you stretch a rubber band too far or blow up a balloon too much, they break. If you mechanically push a speaker too far, it breaks.
For what it is worth.
Steve/bluewizard
One additional point, for purposes of another post in this forum, I tested the output of my system with the volume control at 60% which is about 2 o'clock on the volume dial. 1 o'clock is about 100db, so I'm guessing 2 o'clock is about 105dB. That is LOUD, especially in my small room.
What do you suppose the average long term highly smoothed highly averaged power was at that volume level?
It was in the range of 2w to 3watts. That's all. Though again that we very averaged long term sustained power.
Just though you might like to know.
Steve/bluewizard
What do you suppose the average long term highly smoothed highly averaged power was at that volume level?
It was in the range of 2w to 3watts. That's all. Though again that we very averaged long term sustained power.
Just though you might like to know.
Steve/bluewizard
Thanks for the extensive post! You seem to be supporting some things I've been discovering whilst researching this, which is reassuring.
Now that I understand what the noise in the 100hr noise test is, I can see why thats a better measure than an unrepresentative RMS value. I like how the crest factor is a part of the 'noise' used in the test too, which probably relates to your point about the difference between peak power/volume and the much lower average power.
I also now understand how the use of compression in recordings and/or clipping at playback both start to reduce the crest factor, effectively closing the gap between average power and peak power, increasing the chances of damaging drivers through excess power even with a modestly powered amp.
So I'm now going to just use the 100hr noise test figures as a very rough check on how stressed the drivers are likely to get playing actual music at my intended SPLs. I'm also going to start making much more distinction between this and the peak power of the amps, which will be significantly higher to avoid clipping.
Thanks
Kev
Now that I understand what the noise in the 100hr noise test is, I can see why thats a better measure than an unrepresentative RMS value. I like how the crest factor is a part of the 'noise' used in the test too, which probably relates to your point about the difference between peak power/volume and the much lower average power.
I also now understand how the use of compression in recordings and/or clipping at playback both start to reduce the crest factor, effectively closing the gap between average power and peak power, increasing the chances of damaging drivers through excess power even with a modestly powered amp.
So I'm now going to just use the 100hr noise test figures as a very rough check on how stressed the drivers are likely to get playing actual music at my intended SPLs. I'm also going to start making much more distinction between this and the peak power of the amps, which will be significantly higher to avoid clipping.
Thanks
Kev
I'm trying to get a handle on some of this with my own speakers. The driver I'm working with is a SB Acoustics SB12PFC25-4, which is specced at "30W". It's well and truly run in - I've been listening to lots of music on this speaker for months. It's a nice little inexpensive 4" mid-bass, that I use in a small TL enclosure with a bit of EQ to do ~40Hz, albeit at reasonably low level (noiseUnit - transmission line computer speakers.).
I've just built a new amp to drive this speaker, and at the moment it's capable of putting around 12W into a 4Ω load before clipping. I ran the speaker with REW this morning to test the whole system (active crossover, power amps, speaker) and drove it pretty hard, to the point where it was obviously distorting below ~100Hz. Wondering if it was the speaker or 12W amplifier that was the limiting factor, I put a CRO across the speaker and had a look.
I was surprised at just how low the level needed to be before the voice coil was literally hitting it's stops - with just 5.4V RMS (7W) at 60 Hz. I guess the impedance will vary at these low frequencies so it's not necessarily 7W, but still.
It's making me re-evaluate the power needs for them, at any rate. If I really did put 30W into this poor thing it'd just melt.
I've just built a new amp to drive this speaker, and at the moment it's capable of putting around 12W into a 4Ω load before clipping. I ran the speaker with REW this morning to test the whole system (active crossover, power amps, speaker) and drove it pretty hard, to the point where it was obviously distorting below ~100Hz. Wondering if it was the speaker or 12W amplifier that was the limiting factor, I put a CRO across the speaker and had a look.
I was surprised at just how low the level needed to be before the voice coil was literally hitting it's stops - with just 5.4V RMS (7W) at 60 Hz. I guess the impedance will vary at these low frequencies so it's not necessarily 7W, but still.
It's making me re-evaluate the power needs for them, at any rate. If I really did put 30W into this poor thing it'd just melt.
It has been several years since I worked on this and so I'm I bit rusty, but yes the peak travel of the coil can easily limit things in some cases; initially by becoming non-linear and then by reaching mechanical limits.
There is a huge difference between instantaneous power (of the peak) and more sustained/average power. The former doesn't cause much heat but may hit excursion problems, the latter may be 'much' lower in magnitude and not cause much excursion at all but its (considerably) more relentless nature can cause over-heating.
By way of example, both were factors in my build. Using/misusing a full-range driver just as a mid driver meant excursion capabilities were far more than would be needed, so sustained power handling was the limitation to consider there (especially as a lot of power is present in my chosen mid-range part of the spectrum). The bass drivers could handle lots more power and so wouldn't be troubled by that when matched to the mids, but the low frequency peak excursion could be reached relatively easily (for high power use, I elected to use a rumble filter to constrain low-end excursion).
There is a huge difference between instantaneous power (of the peak) and more sustained/average power. The former doesn't cause much heat but may hit excursion problems, the latter may be 'much' lower in magnitude and not cause much excursion at all but its (considerably) more relentless nature can cause over-heating.
By way of example, both were factors in my build. Using/misusing a full-range driver just as a mid driver meant excursion capabilities were far more than would be needed, so sustained power handling was the limitation to consider there (especially as a lot of power is present in my chosen mid-range part of the spectrum). The bass drivers could handle lots more power and so wouldn't be troubled by that when matched to the mids, but the low frequency peak excursion could be reached relatively easily (for high power use, I elected to use a rumble filter to constrain low-end excursion).
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.