Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[John Roberts {JR}]

I wish the term "peak power" would go away. It is misleading and wrong. Power must be calculated with RMS voltage over at least one complete cycle to have any relevance to audio. For 20Hz, that's 50 ms. I prefer "tone burst power" Tone means audio, at least one complete cycle. Burst means time, how many cycles at what frequency. Power is the area under the curve which relates to the heating effect of the waveform.

Amplifier power can't be expressed with just one number these days. They are designed to have a short term power to get thru a Thump and a long term power to get thru EDM. A curve of power versus time will show both values and if the amp can transition smoothly between them without overshoot or oscillation.

In my bench testing I measured the K10 to do 3600W at 2 ohms for 1.5 seconds, then smoothly dropping to 1260W long term. Test conditions were 50Hz, both channels driven, 230VAC mains. Behavior was good but unfortunately Powersoft consistently overstates their power. I am sure their marketing people say "but everybody does it", and quite a few do, but I don't in my amps. http://www.speakerpower.net/comparative-performance.html

I have given this topic a great deal of thought over decades and the reality is music is the proverbial square peg that does not fit neatly into simple (round) power measurements, while consumers demand and get those simple numbers.

While few make this connection amp power output capability needs to mimic loudspeaker power handling capability, and to some extent, over time modern amps have evolved to meet market needs.  Unfortunately there is not one universal musical genre so there will always be a continuum or range of product capability.

I probably sound like a broken record but smart powered speakers offer the capability to protect transducers from amplifiers capable of breaking them if abused.

JR

PS: I am always reluctant to ASSume any brand intentionally and habitually over-states product specs. Market feedback should correct any egregious outliers.

[Brian Jojade]

Anyways, I didn't mean to sound like I had anything against Powersoft or their products, I'm just still confused how they can make a claim that their amps make more power than they consume.

The reason they can make that claim is that the power flowing out is not at the same rate as the power flowing in.  There are capacitors that store power.  What you need to look at is how many joules are going in and coming out of the amp.  Watts over a period of time will get you how many joules of energy you have.

If the amp has huge storage capacitors, you could have a very low continuous current pull from the wall, yet have huge reserves available for short burst which is what audio is.  e.g., if your amp draws 100 watts for 10 seconds and you could output 1000 watts for 1 second, then wait for it to charge back up. Rinse and repeat. (of course assuming 100% efficiency, but that's another topic)

Better amps will have larger reserve capacity and therefore will seem to have lower power demands from the wall. It's still getting the same amount of power, but the demand can be spread over time.  Cheaper amps with less reserve will pull from the wall when needed.  This isn't about efficiency, it's about storage reserve.

Class D amplifiers are very efficient can theoretically approach 100% efficiency.  Current amplifiers can be in the high 90's for efficiency.  Class AB amplifiers run in the 50-75% efficiency range. Therefore, using a class D amplifier will allow you to get more power to your speaker from the same circuit instead of just converting the power to heat within the amplifier.

[Ivan Beaver]

PS: I am always reluctant to ASSume any brand intentionally and habitually over-states product specs. Market feedback should correct any egregious outliers.

SHOULD is the key word.  But certainly not ALWAYS

Just sayin'-----------------

[John Roberts {JR}]

SHOULD is the key word.  But certainly not ALWAYS

Just sayin'-----------------

I thought ASSume was the key word... As I've always said lets name names... perhaps Powersoft will enlighten us to their scientific basis for peaks power specs. I am not aware of an industry peak duration standard (back in that day I used 20 mSec of 1 Khz tone burst, so seconds of burst power is a lifetime compared to that.)

  JR

[Ivan Beaver]

I thought ASSume was the key word... As I've always said lets name names... perhaps Powersoft will enlighten us to their scientific basis for peaks power specs. I am not aware of an industry peak duration standard (back in that day I used 20 mSec of 1 Khz tone burst, so seconds of burst power is a lifetime compared to that.)

  JR

I am not allowed to name names-but let's just say that some of the biggest names in the industry are "pulling these types of misconceptions".

I am citing specific examples (without using names but using screen captures of their data) in the paper I am working on (JR you were correct-I did not finish it over the Christmas break).

I don't think there is a standard for the length of the burst-but around 80ms seems to pop up a lot-which is very short

[John Roberts {JR}]

I am not allowed to name names-but let's just say that some of the biggest names in the industry are "pulling these types of misconceptions".

By all means scope photos are good, but before you call something a misconception show me the specification that it doesn't meet.

I am citing specific examples (without using names but using screen captures of their data) in the paper I am working on (JR you were correct-I did not finish it over the Christmas break).

keep plugging away...

I don't think there is a standard for the length of the burst-but around 80ms seems to pop up a lot-which is very short

I didn't want to go down this rabbit hole again but I designed a trick amp back in the '80s that used cap doublers in the amp PS rails so I could momentarily deliver 2x voltage which in theory "should" be 4x power... and perhaps it was 4x for a microsecond, but in the real world power rails start sagging as soon as you draw power.

Back then (like now) there was no industry spec for burst power so I had to use my own judgement. I gravitated to an even older hifi (IHF?) spec for burst power that consisted of 1 kHz tone burst that was on for 20 mSec and off for 480 mSec so repeated 2x a second..  My cute little 35W amp could put out close to the theoretical 140W momentarily, but as the mill-Seconds ticked off the un-clipped tone burst got smaller and smaller.  It turns out for me to conservatively specify a clean tone burst for 20 mSec I had to de-rate the burst power to only 100W. Not bad for a 35W amp but marketing would prefer to call it a tens of watts more (I didn't ask their opinion). ****

Of course my 100W for 20 mSec wasn't the full story, for longer than 20 mSec it could put out 60W for something like 15 sec before a thermal breaker would open and the power collapsed to the base 35W 24x7. In my experience my little trick amp could hold it's own against 120W conventional amps on dynamic music. Of course driving a woofer in a multi-way system would be a different story, and that is the rub... Different duty cycle is required for different tasks...

At the risk of repeating myself in the same thread...  I really like powered speakers because the design engineer who knows what he needs, and where, can apply power properly. This trying to understand a not clearly specified parameter really seems like to much mental gymnastics for civilians.

JR


PS: I have a new anecdote about the sheeple clinging to single number specs to buy products with. My old favorite was about vacuum cleaners ranked by line cord current draw (more must be better right?)... I recently had to replace my dishwasher when the old one started making end-of-service-life noises. The new metric to rank dishwashers by is Decibels.  I just bought a "55-Decibel" Whirlpool.  (less must be better... it's more expensive.)
   

**** If you think about what musical transients look like, they do not look like 20 cycles of 1kHz but more like one or two really loud that quickly decay... but that is only one kind of transient. Burst power to reproduce a heavily compressed bass line may have much less crest factor and be harder. If it was easy to come up with a universal spec somebody would have by now. 

[Steve Bradbury]

What you need to look at is how many joules are going in and coming out of the amp.  Watts over a period of time will get you how many joules of energy you have.

If you are measuring joules, that is a measure of work. A small amp at a long gig is likely to pass more joules in total than a larger amp at a short gig and thus do more work.

Power is the rate of doing work. It is equivalent to an amount of energy consumed per unit time. The unit of power (MKS) is the joule per second (J/s); more commonly known as the Watt.

The question that seems to be in dispute is what time frame you use to measure the number of joules coming in and out of the amplifier to define the power rating.

Is publishing the specifications showing the instantaneous power over a millisecond lying or not? It all depends on whose version of the truth you believe.

[John Roberts {JR}]

The question that seems to be in dispute is what time frame you use to measure the number of joules coming in and out of the amplifier to define the power rating.

Is publishing the specifications showing the instantaneous power over a millisecond lying or not? It all depends on whose version of the truth you believe.

Exactly...  Back a few decades ago our seminar guy was promoting his theory that amp headroom was defined by PS rail voltage. I don't think I ever convinced him to discard that notion but it's more complicated than that and rail voltage at best only defines the very first instant of peak power.

The crest factor of different musical genres varies widely for full range, not to mention compressed band-passed applications that can be several times more demanding.

This is why there are different amps sold since light duty customers don't want to pay for heavier duty products they don't actually need. The market generally sorts out which products work for which applications and which don't. That is exactly why we abandoned 24x7 power decades ago.

If the industry put numbers to this, amp designers would game the numbers and marketers would hype them (in a more is better chase), probably like damping factor and slew rate. While more is always better, paying for more than we actually need is not, so after a learning process the marketplace would re-stabilize, but not without even more confusion than now. 

YMMV

JR   

[John Schalk]

In my bench testing I measured the K10 to do 3600W at 2 ohms for 1.5 seconds, then smoothly dropping to 1260W long term. Test conditions were 50Hz, both channels driven, 230VAC mains. Behavior was good but unfortunately Powersoft consistently overstates their power. I am sure their marketing people say "but everybody does it", and quite a few do, but I don't in my amps. http://www.speakerpower.net/comparative-performance.html

Like the OP, I too have often wondered about the tremendous output that some amps claim to produce from a power cord with a 15 amp edison connector on the business end.  I have purchased 3 Powersoft amps (2 K2s, 1 M50q) with DSP in the past year, so after reading this thread, I did some research on Powersoft's website. I learned that the power output ratings they use are based on the EIAJ specification.  I've always understood that to be "1k @ 1% THD", but assumed that it was a steady state or continuous spec, like the FTC rating.  Despite my best efforts, the exact wording of the EIAJ spec has alluded me.  I did find the following information in a PDF that explains how to replicate the results for DIAGM amps:

1. Power RMS x Channel (1kHz, 0.5%THD):
• Maximum continuous output power delivered to the load for at least 5 seconds, then internal limiting processor reduces output voltage swing to _ maximum output.
2. Power RMS x Channel in stereo working:
• Maximum EIAJ standard test output power: average of RMS output voltage squared divided by load impedance.
• Test signal 1kHz tone burst 8 cycles 0dB, 24 cycles no signal, THD 1%.

I can't quite understand what these two tests are telling me.  Is point 1 saying that a K2 will output 1000 watts into 8 ohms for 5 seconds, but measuring only a single channel? I don't get point 2 at all. Can someone enlighten me please?  And is this the EAIJ specification?

I did find a published test result for a K8 by ABELtronics which shows the K8 exceeding its 3000 watt rating at 1k and 10k, but missing by 84 watts at 40 Hz.  They defined continuous as 5 seconds for their testing.  Here is the link:

http://www.abeltronics.co.uk/amptesting/amp-meet-three.php

Final comment; I am happy with my Powersoft amps.  I racked up a K2 and the M50q in a 2 space SKB style case and configured the DSP so that I have a biamped stereo system (K2 for lows, 2 channels of M50q for tops) and 2 monitor sends.  Now, when I walk into a breifcase gig, if the band's power looks dodgy, why I just bring in my amp "rack" in one hand!  I can't wait to ditch the copper snake.

[John Roberts {JR}]

Final comment; I am happy with my Powersoft amps.  I racked up a K2 and the M50q in a 2 space SKB style case and configured the DSP so that I have a biamped stereo system (K2 for lows, 2 channels of M50q for tops) and 2 monitor sends.  Now, when I walk into a breifcase gig, if the band's power looks dodgy, why I just bring in my amp "rack" in one hand!  I can't wait to ditch the copper snake.

The computer ate my first answer so I'll keep this brief.

AFAIK the 5 sec duration for power testing is arbitrary so not an industry standard.

How the amps work for you in your application is the real test.

Don't just take my word for it, but have some trust in the market to identify and punish any serial liar who over-promises and under-delivers on amp power. Back in the '70s it was so bad that the government got involved. Any simple government imposed answer to a complex issue like amp duty cycle will cause unintended consequences (like the old FTC specs did).

Research actual end user reports about specific products on forums like this. Trying to understand dynamic power specifications is complicated (please trust me about that at least). 

JR

[ProSoundWeb Community]