How much power?

[Jayrosc]

Can anyone put the following power output description into simple layman's terms? These are the specifications for an old Sony HP-610A:

Music power output,EIA: total 66W both channels at 8ohms

Dynamic power output IHF: total 50W at 0.3% THD

Continuous power output: 20W/20W rms at 0.3% THD

Being a layman, I don't know if this means this stereo, which sounds very nice, puts out 66 watts per channel or 20 watts per channel...or 50 watts per channel or...some combination of all three depending on how loud I have it.
Trying to figure this out because my speakers (PSB T5's) need a minimum of 20 wpc.
Thanks to anyone ready to educate.

 

[BOUXY]

Can anyone put the following power output description into simple layman's terms? These are the specifications for an old Sony HP-610A:

Music power output,EIA: total 66W both channels at 8ohms

Dynamic power output IHF: total 50W at 0.3% THD

Continuous power output: 20W/20W rms at 0.3% THD

Being a layman, I don't know if this means this stereo, which sounds very nice, puts out 66 watts per channel or 20 watts per channel...or 50 watts per channel or...some combination of all three depending on how loud I have it.
Trying to figure this out because my speakers (PSB T5's) need a minimum of 20 wpc.
Thanks to anyone ready to educate.

This unit is rated at a TRUE 20 watts RMS per channel not rated at today's peak watt ratings. Heavy bass and solid sound throughout.:thmbsp:

 

[Jayrosc]

er, thanks for the quick response, but I'm afraid I don't even know what "rms" stands for. I think you are telling me that these deliver just enough power so as not to starve my speakers,yes?

 

[RichPA]

EIA is one channel driven, 1000 Hz, 1% distortion - in other words, easiest case for an amp.

I forget the details, but IHF is brief peak power

Continuous power is the most meaningful measure, especially if the bandwidth is specified (it's generally harder for an amp to produce power at lower frequencies). RMS stands for Root Mean Square - a method of averaging that discounts brief extremes (a little Googling will find you a more technical description).

 

[Jayrosc]

as for bandwidth, I think it is specified, if by bandwidth you mean frequency response.
it says 30 Hz-50kHz at 1 W output.

 

[wmgwizard]

The figure that is most important is the RMS(averaged) power. The others are just various test methods that offer different results.

Typically you would read 20 wpc RMS from 20hz-20000hz @ .01% distortion.

 

[JoeESP9]

The only power spec that actually means anything is the RMS power. RMS means root mean square otherwise known as continuous power.

If you're asking this because of wattage ratings for your speakers, be advised, speaker wattage ratings are pretty much worthless and useless.

 

[stonedeaf]

1W figure

I sold these back in the day -wouldn't recommend them for a really power hungry speaker - but better than a lot of their compact stereo competition.
None of these ratings are really stringent by today's standards - what you'd like to see is RMS power from 20 hz to 20kHz - both channels driven at a staed max. distortion. Even today a lot of creative spec waffling occurs - none of these specs tell you anything like that. Probably safe to conclude originally something between 10 and 20 watts per channel. But that's when it was new (forty -45 years ago?) -if unrestored with all original and probably pretty tired power supply capacitors - probably a bit less than that before distortion becomes noticeable. And that's really the way to deal with this -listen for clipping.And turn it down when you hear that.

 

[ken kantor]

I really don't get why people feel that RMS is the most important spec on an amp. It >surely< is not. Almost all knowledgeable amp designers strive for headroom, which is the exact opposite of "RMS" power!

-k

 

[Jayrosc]

Thank yu all...now can someone tell me what "clipping" sounds like?:scratch2:

 

[RichPA]

I really don't get why people feel that RMS is the most important spec on an amp. It >surely< is not. Almost all knowledgeable amp designers strive for headroom, which is the exact opposite of "RMS" power!

-k

I understand that, but I think it's because RMS is the only one that seems to be consistently measured, especially back when there were truly ludicrous "peak power" claims. And I don't think 1kHz is where you need peak power. I think that complete power specs would include power under different conditions into different loads, with bandwidth specified. I'd be interested to know what set of specs you think would best characterize an amp for a consumer - it surely wouldn't be a single number?

 

[Chris Brown]

I really don't get why people feel that RMS is the most important spec on an amp. It >surely< is not. Almost all knowledgeable amp designers strive for headroom, which is the exact opposite of "RMS" power!

-k

I personally prefer to see wattage measured under a number of different cases/variables, rather than just one.

I often push my amps hard, so RMS power is one of the ones I like to see, because I feel it's valuable to know what an amp is capable of when approaching something of a worst case scenario. I feel it's also particularly valuable in the face of an unrealistic wattage claim being advertised for a unit. Typically, seeking out the RMS power output is the easiest way to judge just how badly the advertised wattage claim has been exaggerated.

 

[Chris Brown]

Thank yu all...now can someone tell me what "clipping" sounds like?:scratch2:

THD tends to rise sharply as you approach clipping. The distortion is not subtle. If you are listening to music where most of the music isn't causing clipping but occasional strong bass notes or similar are clipping, it will be noticeable at first as a loss of definition in the bass notes. Bass will go from feeling like you're getting hit by a jackhammer, to feeling like you're getting hit by a wet sponge, to feeling you are being hit by block of foam. It should be pretty noticeable.

 

[RolandA]

rms refers to "root mean square" which basically means the equivalent power that you'd get if you ran a DC voltage into the speaker load at the 0.707 times the maximum continuous voltage the amp can put out. I bet that helps.
In other words, the rms power is the maximum continuous power the amp can produce at a reasonable distortion level for a long period of time. If you short your phono input so it hums loudly, and turn it up, that is showing you the rms power it can do. (Don't do that!)

The idea was that music is not continuous noise, it contains short peaks of high power mixed in with lots of lower power stuff, so you don't really need lots of rms power. This is good from the designer's point of view because he can use a wimpier power supply that produces a lot of voltage for a short time, but as soon as things get hot and heavy drops off. This economizes on the mains transformer and the rectifier diodes, and also means he can ignore any voltage regulating components. Of course it also means that the power supply voltage changes the louder you play it, so the operating set points of the various transistors or tubes also vary, so the thing doesn't work the way you thought it would.
Nowadays better amps tend to use large toroidal transformers so they suffer less from this issue.

A more important point is that it is the current draw that causes the power supply to droop, which means that the current the amp can supply the speakers is inherently limited. Since dynamic speakers are inductive loads around their resonant point (in the bass), this means the amp can't handle the speaker very well - it essentially current limits either explicitly or by virtue of the power supply dying. This makes the bass loose and boomy, and increases intermodulation distortion.

So the bottom line is that all those power ratings are worthless, other than the rms one, and that one should be taken with a grain of salt. The fact that the amp *does* have those other ratings is really a bad sign - they are trying to snow you into thinking it is better than it really is.

 

[stonedeaf]

power of the era

The OP's piece goes back to a era where power output was rated with about the same inherent accuracy as political TV adds. I am always interested in alternative amp topologies that can at least make a argument for providing bass control with high peak power. Class G is interesting ? The various "digital" amps are probably the wave of the future and sure as heck don't sound the same as each other?
But - the very best I've ever heard - big stupid straight ahead class A amps. Hot , heavy and double power down to say 2 ohms. Most of these don't have much headroom - but they will punch current into a real world loudspeaker all day long. And have a talent for making even ridiculously modest speakers shine.

 

[ken kantor]

rms refers to "root mean square" which basically means the equivalent power that you'd get if you ran a DC voltage into the speaker load at the 0.707 times the maximum continuous voltage the amp can put out. I bet that helps.
In other words, the rms power is the maximum continuous power the amp can produce at a reasonable distortion level for a long period of time. If you short your phono input so it hums loudly, and turn it up, that is showing you the rms power it can do. (Don't do that!)

The idea was that music is not continuous noise, it contains short peaks of high power mixed in with lots of lower power stuff, so you don't really need lots of rms power. This is good from the designer's point of view because he can use a wimpier power supply that produces a lot of voltage for a short time, but as soon as things get hot and heavy drops off. This economizes on the mains transformer and the rectifier diodes, and also means he can ignore any voltage regulating components. Of course it also means that the power supply voltage changes the louder you play it, so the operating set points of the various transistors or tubes also vary, so the thing doesn't work the way you thought it would.
Nowadays better amps tend to use large toroidal transformers so they suffer less from this issue.

A more important point is that it is the current draw that causes the power supply to droop, which means that the current the amp can supply the speakers is inherently limited. Since dynamic speakers are inductive loads around their resonant point (in the bass), this means the amp can't handle the speaker very well - it essentially current limits either explicitly or by virtue of the power supply dying. This makes the bass loose and boomy, and increases intermodulation distortion.

So the bottom line is that all those power ratings are worthless, other than the rms one, and that one should be taken with a grain of salt. The fact that the amp *does* have those other ratings is really a bad sign - they are trying to snow you into thinking it is better than it really is.

Gosh, I don't know where to begin... Just.... No.

-k

 

[ken kantor]

I understand that, but I think it's because RMS is the only one that seems to be consistently measured, especially back when there were truly ludicrous "peak power" claims. And I don't think 1kHz is where you need peak power. I think that complete power specs would include power under different conditions into different loads, with bandwidth specified. I'd be interested to know what set of specs you think would best characterize an amp for a consumer - it surely wouldn't be a single number?

Yes, I understand the history of amp ratings. (I designed power amps for a number of years and contributed to the FTC standard that is currently in use.) The stringent adoption of RMS was a clear over-reaction to "When struck by lightning," power claims of the past. It brings with it several problems and weaknesses. Any power spec that ignores the signal characteristic of what it is purporting to measure is doomed to failure. An anal allegiance to RMS rating yields amps that are over-priced, inefficient and potentially sonically compromised, and the RMS power of an amp tells the user relatively little about how loud the amp can play.

I wouldn't say that the cure was >worse< than the disease, but close!

-k

 

[ken kantor]

I personally prefer to see wattage measured under a number of different cases/variables, rather than just one.

I often push my amps hard, so RMS power is one of the ones I like to see, because I feel it's valuable to know what an amp is capable of when approaching something of a worst case scenario. I feel it's also particularly valuable in the face of an unrealistic wattage claim being advertised for a unit. Typically, seeking out the RMS power output is the easiest way to judge just how badly the advertised wattage claim has been exaggerated.

I agree with your first point, but not the last sentence. While true that RMS can expose unrealistic or fraudulent power claims in a bad amp, it can just as easily obscure the excellent performance of a >good< amp.

-k

 

[charles 1973]

For all intents and purposes the "Continuous power output: 20W/20W rms at 0.3% THD"
is the most legitimate. To be totally legit it should include the bandwidth (from 40 - 20,000 Hz for example), an impedance (typically 8 ohms), and would be driving both channels simultaneously. It should all be included in the power spec. Bandwidth as a separate spec may not be the same.

Peak power in a stereo amp is Very Important. The IHF spec is a peak rating, but unfortunately it is only measured over 10 - 20 Ms. Short term peak power sustained for 200 Ms would be more useful, but is seldom mentioned. It is the peak power of one amp that makes it sound much more powerful than it's RMS rating would suggest. Large power supplies and plenty of heat sinking relative to its output in a class A/B amp are an indicator of such reserves.

It should be enough to drive those PSB T5's to satisfying SPL's without clipping for the most part. If your trying to impress your friends with how loud those speakers will go, then you might want 50 - 100 wpc. Myself, I seldom use more that 10 wpc with speakers of comparable sensitivity, and they play plenty loud then.

A word of caution: If you are using the tone controls to add bass or treble, back them off to neutral before running the amp near full power. Adding 6 db's of bass would require about 80 wpc instead of 20. Loudness is best used at lower volumes, and should be turned off at higher volume. Often loudness is linked to the volume control, so as we approach full power, the loudness compensation is reduced to 0. But I wouldn't test it.

 

[BOUXY]

For all intents and purposes the "Continuous power output: 20W/20W rms at 0.3% THD"
is the most legitimate. To be totally legit it should include the bandwidth (from 40 - 20,000 Hz for example), an impedance (typically 8 ohms), and would be driving both channels simultaneously. It should all be included in the power spec. Bandwidth as a separate spec may not be the same.

It should be enough to drive those PSB T5's to satisfying SPL's without clipping for the most part. If your trying to impress your friends with how loud those speakers will go, then you might want 50 - 100 wpc. Myself, I seldom use more that 10 wpc with speakers of comparable sensitivity, and they play plenty loud then.

A word of caution: If you are using the tone controls to add bass or treble, back them off to neutral before running the amp near full power. Adding 6 db's of bass would require about 80 wpc instead of 20. Loudness is best used at lower volumes, and should be turned off at higher volume. Often loudness is linked to the volume control, so as we approach full power, the loudness compensation is reduced to 0. But I wouldn't test it.

I answered in the 1st Post to reply "This unit is rated at a TRUE 20 watts RMS per channel not rated at today's peak watt ratings. Heavy bass and solid sound throughout."! No mention of any other rating was answered just the Power and that is what I replied period:scratch2: