Old power specs vs new ones

[stevehertz]

As some will know, in my ever shifting focus of interest within the field of vintage radio, audio and video, my present 'thing' is late 70s hifi receivers. Anorak as I am (aren't most of us on here?) I am now well conversant with lots of models, their power rating etc etc. Some time ago I began to wonder how these vintage receivers would compare with today's multi-channel versions, many of them specified with equally impressive high power output levels. Ha ha, that is until you begin to dig. I've read comparisons between modern and old receivers, and in some cases the old ones have faired better wrt sound quality, not a surprise to me and perhaps many of us on here. But the biggest er, 'anomaly' is the power ratings that are bestowed upon the modern sets. You'll see (for example) 7-channel AV receivers rated at 150W per channel. How can this be true? surely these sets would be colossal? Some of my vintage stereo (that's just two channels!) c100W/ch receivers measure 565mm wide (nearly 2ft) and weigh up to 64lbs! You can literally, barely lift them! So I realised that something 'was going on', but then I read this article that put it all, very succinctly, into words. Taken from an Audioholics web page about 'Those high powered Pioneer vintage receivers:

The Age of the Receiver was upon us. And a Golden Age it was. Every year major manufacturers like Pioneer, Kenwood, Sony, Sansui, JVC, Marantz and Sherwood introduced new and better models, with more features, more power, and lower prices.

Still, no one had cracked the “magic” power figure of 100 watts RMS per channel. The 3-figure barrier seemed like an impenetrable wall, almost foreboding and sinister, as if something horrible would befall the party with the temerity to attempt it. It was almost like the fear that aviators had in 1947 about breaking the sound barrier before Chuck Yeager did it in the Bell X-1.

There was the safety of 40, 50, 55, 70 watts per channel. But no one dared to go to…..<gulp>….100. Remember, too, in those days 60 per side was considered more than enough. And since the reputable manufacturers in those FTC days were rating their equipment honestly and conservatively. It was a competitive badge of honor to see how far you could surpass your “published specs” in a review in a major magazine, then quote the reviewer in your next print ad saying, “The Kenwood XYZ easily surpassed its rated power specifications on our test bench, clocking an excellent……” 60 watts RMS was 60 really, really gutsy watts RMS.

Not like today, when most popularly-priced “100-watt x 7 channel” receivers sold at department stores are 100 watts at 1kHz with only one channel driven—not over the full 20–20kHz band with all (or even two) channels driven simultaneously (which really taxes the power supply, the output devices and the heat sinks), like they had to do in 1974. And too many of today’s receivers do it at a barely-hi-fidelity THD of 1% at 1kHz--not at the 1974 usual of .5% or .2%, over the full 20-20kHz range.

If you read the owner’s manuals of one of today’s popularly-priced “100 watt x 7” receivers very closely, you’ll likely find—in small print—a real rating of something like “60 watts RMS at .5% THD 20–20kHz, two channels driven simultaneously.” This is how they can still maintain FTC compliance.

Uh…..not exactly “100 watts x 7.” Hey, there’s no free lunch. Some lightweight 22-lb wonder selling for $299 is not going to miraculously deliver 100 watts RMS x 7 channels. The typical Kenwood or Pioneer 60-watt RMS x 2 integrated amplifier of 1978 weighed more than many so-called “100-watt x 7” midfi home theater receivers of today.

So, just like many car audio freaks boast that their systems are rated at '1000W/ch', there are lots of home theatre enthusiasts who think that they are getting (a true, meaty) '100W/ch' out of their 5, 7 and 9 channel systems. Nope! In the example above they suggest that a 100w/ch 7-channel unit may be rated at 60W/ch at .5% distortion. Well given that - in Pioneer's case - they used to rate their amps at .1% distortion, then I would de-rate that quoted 60W/ch even further, maybe 40-50W, who knows? No, there's 'old watts' and there's 'new watts', and ne'er the twain shall meet.

[ms660]

From the First Law Of Gullibility (FLOG)...

BS cannot be created or destroyed only its form can alter

EDIT: Actually I just made that up...

Lawrence.

[paulsherwin]

Many modern home cinema amps can indeed deliver a lot of power from a relatively small unit. There are two reasons for this. Firstly, they use SMPSUs so don't need whacking great mains transformers. Secondly, they are not rated with all channels driven, and the output capability of an individual channel is limited by the available DC rather than the chips or circuit design, so will be much higher if it's the only channel doing anything.

To be fair, it isn't very helpful to rate 7.1 amps with all channels driven, as it's highly unlikely that they'll be used in that way.

Output power ratings have always been a minefield.

[kalee20]

Well, an amplifier that puts 10V RMS of undistorted sine-wave into an 8Ω load, is producing 12.5W average of real power.

Of course, the peak voltage will be 14.14V, so peak power is 25W. But it's only instantaneous, twice each cycle. Good for specmanship, though.

A bit of calculus shows that if you wanted to define an RMS figure, it would be 15.3W. Not really relevant unless you are writing a spec sheet, where 15.3W RMS looks better than 12.5W true power.

But most audio amps can't drive a steady sine wave for long. The output devices may be up to it, but not the heat sinks or the power supply. Music peaks which last a second or so are catered for by thermal inertia of output devices, and by stored energy in reservoir capacitors. So the continuous power rating may be one figure, the brief music power another... then you could play the 'Average,' RMS,' 'Peak,' cards to get another three figures. 'Average peak music power,' might be acceptable to us lot, just about. The salesmen on the other hand might go for 'Peak peak music power.' Which is just a figure for the sake of it!

[M0FYA Andy]

These sorts of discussions always remind me of a Grampian amplifier I have as one of my 'relics'. Rated at 100 watts using four KT66's in parallel push-pull, this was quite adequate to provide 'music while you work' to loudspeakers throughout a five-story hosiery mill full of noisy knitting and sewing machines................
Andy

[bluepilot]

Originally Posted by kalee20

So the continuous power rating may be one figure, the brief music power another

I remember when a lot of manufacturers quoted a "music power" rating. Based on the assumption that music will contain a lot of frequencies which interfere destructively at times they seemed to think you could take an amplifier with say 10W continuous sine wave rating and call that say 20W music power. I guess the rating depended on what sort of music you liked.

[G6Tanuki]

A lot of the "AV receivers" and surround-sound stuff uses 'Class-D' digital amplifiers, which are basically a D/A converter mutantly hybridized with a SMPS. Which in a way makes sense when the signal-source is itself digital - rather than do a D/A conversion in the TV/CD/Blu-Ray-player that's driving it and then amplify as analog to the required power, take a digital signal from the source (via something like SPDIF or HDMI as the interconnect) then move the D/A conversion process to immediately before the speakers.

If you want crazy audio power-ratings look at the speakers you get to attach to a PC: somehow these can claim "100 Watts output" when powered from a humble 12V 1A wall-wart...

[Edward Huggins]

I find certain highly esteemed manufaturers just don't give specs today in the way that we were used to seeing them e.g. in forensic detail. For example BOSE, do not publish or make available, even upon telephone request, ANY specs for their range of audio equipment. Their Technical Helpline will say "all the power you will ever need", "very deep bass for the size", "no distortion even at full volume"....It's all probably Class D amplification and only the Boffins in the US will know the measurements.

[stevehertz]

That's true Edward. I had a Bose L1 portable PA system and you could never get a power rating for it.

[kalee20]

But that's not exactly new. Rolls-Royce cars, the power wasn't disclosed either, simply that it was 'sufficient.'

And the idea of rating amplifiers for more than they could sustainably deliver, well, Mullard did that with the 5-10, low-loading. It could give more power than the normal-loading, but it couldn't do it for very long. Music peaks drew more current from the reservoir capacitors as the thing moved heavily into Class AB, and if sustained, the voltage fell, the operating conditions changed, the power available started to sag, and the EZ80 rectifier got overloaded trying to keep up. But unless you liked loud organ music, it worked very well, and it was at least a technically justifiable circuit and spec tweak!

[mhennessy]

Everyone seems to forget impedance...

Power is the product of volts and amps, and amps depend on the applied voltage and the impedance.

But the impedance of a loudspeaker varies wildly with frequency, so the power it dissipates will strongly depend on frequency. There's more about this here: http://www.jmwa.demon.co.uk/How%20do...%20signals.pdf

So it would be far better to forget about power and quote output voltage. Then, you could correctly use the term "RMS" - of course, "Watts RMS" is a marketing term, but it seems to have stuck as "continuous average sine wave power" is a bit of a mouthful

Of course, power output tells you hardly anything about useful loudness anyway. There is that unfortunate matter of logarithms, that mean that going from 50W to 60W results in an insignificant and inaudible 0.8dB increase - that rather inconvenient truth is never quoted in reviews and marketing material.

As for the weight of modern AV receivers compared to older receivers, don't forget that real audio signals have a large PMR (peak to mean ratio), meaning that the heat sinks and power supplies aren't expected to deliver the full sine wave power into a dummy load. A typically assumed PMR is 20dB, which is a factor of 100 in watts - in other words a 100W amplifier that is only just clipping on the peaks will be delivering an average power of just 1W to the load. In practice, movies have much greater PMRs. So it's quite valid (and environmentally and commercially responsible) for manufactures to save cost/weight on the PSU and heat sinks. Those heavier receivers from the '70s are over-engineered. Naturally, in those heady days when people were pushing the limits of the current state of the art, a bit of over-engineering was a sensible commercial decision - attempting to save costs before the experience of how these things perform in the field would have been risky indeed. But that's the standard product development cycle: work out how to make it to a reasonable standard of performance and reliability, then work out how to make it as cheaply as possible without causing reputational damage. Having got past the first part of the cycle, over-engineering is as much of a sin - and perhaps arguably greater - as under-engineering.

Regarding class D amplifier, class D is not digital. At their core, they use pulse-width modulation, taking in an analogue signal as their source. You wouldn't call a switched-mode power supply "digital" just because the switches are either on or off. OK, some implementations might use digital circuitry to simplify things for mass manufacture, but it's not true to state that a class D amplifier is the DAC for an S/P-DIF signal. Especially when you consider the need for negative feedback.

[RojDW48]

And - to complicate the issue further - there is the issue of valve vs solid state power and 'measured performance' as in the famous Hi-Fi for pleasure blind test carried out in 1975, in which an old Radord STA25 (!) was judged the best of all the latest solid state amps by people who should be able to tell! - see here - https://www.stereophile.com/reference/70/index.html

[stevehertz]

Fascinating article. Well written and certainly food for thought!

[crackle]

Simple, look at the mains power rating.
That will tell you if the amp which claims to be 100 watts is likely to deliver it.
Take a Quad 303, 45w per channel into 8 ohms, the power input is 40 to 200 watts. It will deliver what it claims at the distortion levels it claims in the specs, and will deliver loud transients with no effort.

I have a solid state PA amp that claims to deliver 320 watt per channel into 4 ohm speakers, and the mains rating is 1280 watts max.
Now that is what I call an amp that should deliver what it claims.


Mike

[mhennessy]

More factors contributing to the lower mass of modern electronics:

Cooling fans:

The DC fans we have today - thanks in part to the computer industry - are simple to control, very cheap, and can be very quiet. In conjunction with a simple control circuit (perhaps just 2 transistors) and an NTC thermistor, you've got a fan that only runs up when needed. The ability of a given heat sink is increased massively when you introduce forced-air cooling, saving cost and weight. Some might think that a fan is the last thing you want in a high quality audio system - and in principle this is right - but these amps aren't class A, so the fan will only run when listening loud. And luckily, that isn't all that often (look back at my previous comments about peak-to-mean ratios).

Modern heat sinks:

Traditionally, heat sinks have been extruded aluminium. These are effective, but heavy and expensive. Modern heat sinks use a sheet of thick aluminium to help conduct the heat away from the devices, and bonded sheets of thin aluminium to form the fins. These are much lighter than the equivalent extruded model, and expose more surface area to the air, so can be more effective for a given physical size. If needed, copper can be used instead for the spreader (fairly common) or both (less common, but sometimes seen, especially with forced-air cooling).

Thermal switches:

When these were more expensive and less commonplace, you had to build your amplifier so it would not overheat. However, put these in, and you have an amplifier which simply shuts down with no harm done. Once it's cooled down, it works again as if nothing happened. Put some vague advice in the user manual about not blocking the ventilation and not over-driving it, and you've made it an SEP, while saving a decent amount of money and resources in the process. It's easy to see this as "cost cutting" but it's actually quite pragmatic because normal music with a wide PMR shouldn't cause the amp to get all that hot, but if someone uses it for a party, then the amp is protected against this form of "abuse" (manuals state that domestic hi-fi amps are not built for PA and similar).

Switched mode power supplies:

We all know that these are much smaller and lighter for a given power throughput. So there is much temptation to use these for audio - even if you are sticking with a class B amplifier rather than class D. As with any engineering choice, there are pros and cons, but from the environmental and commercial point of view, they make a lot of sense. From a technical POV, they have much to recommend them too.

As regulations get tighter, we will see more and more of these. Getting the standby consumption to below 0.5W is rather tricky unless you use a SMPSU, and tightening regulations around power factor correction will eventually see an end to the traditional 50Hz transformer and rectifier and large C supply. Just look at how PFC has been applied to all computer power supplies today, not just the large ones.

With careful design, switched mode power supplies can be made to run at any mains voltage - and if you are using PFC (very likely), then actually, auto-mains adjustment comes as part of the deal. So, no need to include a 115/230V selector switch - which is something that, as well as adding cost, could confuse the consumer or cause damage to the product.

SMT

We all know about surface-mount components. Stating the obvious here, I know, but it's easy to forget how these result in a real saving in size, hence weight (smaller PCBs means a smaller chassis)

Cumulatively, using lighter and smaller internal components means that the (smaller) chassis can be made from thinner materials for the same structural strength. It all contributes...



In summary, size and weight really is no indication of quality. You have to read the specification carefully to see behind the "headline" figures - obviously - but reputable manufacturers can't lie about their power ratings. It's simply not worth the risk.

Hi-Fi News always tests amplifiers, and I can't recall an instance of a modern solid-state amplifier that hasn't met its spec - indeed, the measured value is usually well in excess of the published specification. That's because manufacturers will test their designs and "round down" to the nearest "nice" number - hence why you buy 100W amplifiers, not 113.8W amplifiers. In the case of an amplifier with a conventional unregulated PSU, the test will often be done with the mains voltage wound down slightly to build in a safety margin, as you can't guarantee what the mains voltage will be at the reviewer's house (or you might state the input voltage for a given output power). As I say, no-one wants to get caught lying about this.

Obviously, I'm not talking about the bottom end of the market here. We have "comedy" 1000W PMPO PC speakers with 6VA wall-wart PSUs exist, but that sort of stuff seems OT for this thread.

[Edward Huggins]

A very logical and informative Post. Thanks, I learned from this.

[Radio Wrangler]

I was told by a designer at Linn that they'd hadtogo to switch mode supplies in their larger amplifiers simply because of regulations about harmonics in the consumed current coming in at above certain power ratings, so they had to have power factor correction in the supply to legitimately get a CE mark.

David

[Peter.N.]

The Quad, Leak and the like amps of the '50s produced more sound than you could stand with big speakers and relatively low power, makes you wonder about the modern specs.

Peter

[mhennessy]

It's not so much the amplifiers; rather the big change is the loudspeakers.

In the early days, watts were expensive and difficult, so making the loudspeakers as efficient as possible was the aim. This means lightweight paper diaphragms and simple crossovers. Such loudspeakers easily fill a large room with less than 10 watts...

The trouble with these speakers is that they are not exactly neutral. While fine for many applications, you couldn't really use them for quality monitoring in studios and similar. And if you did, you found that no two outwardly-identical loudspeakers sounded exactly the same. This lack of consistency is what made the BBC give up on commercial loudspeakers and design its own.

But the improvements to quality and consistency meant heavier diaphragms, with a resultant loss in efficiency. Especially if you wanted to equalise the drive units to give something approximating a flat frequency response - this requires a ~6dB bass boost, and the only way to achieve this in a passive crossover circuit is to cut everything else by 6dB.

Luckily, amplifiers with larger and larger output powers were appearing, and with other improvements (better adhesives, etc), it was possible to make speakers that could soak up the power without failing (mostly). The result is much better quality audio, but we need bigger amps. Not much of a price to pay in all honesty...

Of course, very efficient loudspeakers do exist today. They are made for the benefit of those who believe in the magic of single ended triodes and the like. I've heard a fair few of these setups, and there's no doubt they are very entertaining, but accurate they're not

[Radio Wrangler]

With the advent of the multichannel home cinema amps, the definition of the watt was changed. It used to mean the ability to raise a 1/(9.81) kg mass upwards at 1 metre/second and it was changed to use the power of an 1980s Italian illegal CB 'burner' amplifier as a reference of 1000W.

David