On Thursday The Guardian published information linking Samsung to the current Volkswagen emissions fiasco. Samsung is accused of installing a ‘defeat device’ on some televisions that uses less energy during official testing conditions than would be found during real-world use.
“The apparent discrepancy between real-world and test performance of the TVs is reminiscent of the VW scandal that originated in the US last week,” wrote [Arthur Nelson] of The Guardian. This report was based on an unpublished lab test by the research group ComplianTV which found discrepancies between real-world and test performance when measuring power consumption. According to ComplianTV, this is due to the ‘motion lighting’ setting included in some Samsung TVs. Samsung vehemently denies this ‘motion lighting’ saying that it is not a method of cheating the consumption tests.
Not one to let a good controversy go to waste, the BBC reports a Samsung TV will reduce its power draw shortly after the start of the test. A graph of the power draw of a TV – not explicitly a Samsung television – demonstrating this functionality was found in a PDF of a ComplianTV workshop from last year labeled as, “Typical results recognized during testing” with a decrease in power consumption being a recognized behavior when the appropriate test video was found.
This is not the first time ComplianTV tested a Samsung TV equipped with a ‘motion lighting’ setting. Earlier this year, ComplianTV measured the power consumption of the Samsung UE55H8090 television, and found this TV was compliant with energy regulations. Incredibly, all Samsung TVs listed on the ComplianTV database were found to be compliant with the relevant energy directives.
Samsung’s rebuttal to the Guardian article states the ‘motion lighting’ technology is an ‘out of the box’ feature, active in both the lab and at home. Unlike Volkswagen’s ‘defeat device’ for their diesel engines which is only active during emissions testing, the ‘motion lighting’ technology is active whenever it is enabled in the TV’s settings menu.
Anyone in the US who has shopped for a television in the last four years will have noticed cost-per-year estimates for operating the appliance. This is only an issue if the televisions don’t actually meet that advertised benchmark. Until we see a published study we’re raising our eyebrows at The Guardian, easily one of the most trusted journalistic institutions on the planet, and reserving judgement for Samsung.
Looks like prefetch for me: 100% CPU load etc to buffer (load and maybe decode) some period of video.
Rather a boot-up I presume.
Nah, that’s a power graph, not CPU usage. The ARM types used in TVs aren’t gonna suck up an extra 30W at full usage.
Well, If Samsung says there is no rigged software in there, they should know. I mean, they have a lot of experience with that sort of thing..
Samsung is being accused of lying, and your answer is to just believe them, because they have experience with this? Because you can trust liars to admit to their lying, right?
Would be more believable if they included similar graph from SAME TV playing another video clip in exactly same circumstances (time from cold boot etc)
I had a look at the report in the Source link below the graph. As a retired video editor who started out as an engineer, I have a “sort of professional” interest in this story, plus I own a Samsung HD TV. The report in question was commissioned to examine the regulations around TV for sale in the EU. It mainly covers how wording could be improved, as well as how testing and reporting can be done.
The graph shown seems to have nothing to do with the story. It is given as an example saying “this is what it might look like”. And it is indeed showing the power consumption during boot-up.
As an aside, my experience of TV’s goes all the way back to the early 60’s, and I am rather amused that they now take as long, if not longer, to start up as the old valve ones did. That’s “tube” to those of you on the left side of the Atlantic!
I have a Sony Bravia which I’m fairly sure is running on Linux. From observation it seems to have two modes of shutdown, one is like sleep and the other like shutdown. So the first time you turn it on it takes forever.
I wish these were modular and I could just upgrade the computer driving the thing.
…and that’s where the hacking begins.
Like in ‘left side’ when looked at the Atlantic from above earth with Antarctic at the bottom of the view?
;-)
I set up Samsung’s on a daily basis for several years now. The increased power draw shortly after power up is due to the background light, which automatically gets dimmed down after a couple seconds with standard settings. When you made it through the first time installation, there appears a notification at the top side of the screen that informs you about the Eco Mode being switched on, which will regulate the background lighting according to the brightness in the room. Of course you will find that the brightness will be overly low, especially in a brightly lit room. You will have to turn off the Eco Mode so that the brightness will actually stay on the level you set it to in the Picture Menu, but then of course the power draw will be increased and the whole advertised energy label will be void.
Samsung even puts a Typical Power Rating (which is the “good” one you find on the Energy Label, and is true only with out-of-the-box settings) and a Maximum Power Rating (which is the “realistic” one with full backlight brightness) on the type label of every TV set.
This is not exactly the same thing like what VW did. It’s more like an A+++ rating for the 60°C setting of your washing machine, while it is only able to achieve 45°C in reality…
Are we talking about trusting Samsung’s word who were caught – and proven to be – rigging mobile phone benchmarks?
Payola much?
In a future where it becomes common for people to have rooftop solar offsetting their power bills, there may be a lot less concern about saving power on all our stuff. Some may even enjoy leaving their incandescent lights on all the time!
I would’nt do that. Due to planned obsolescence they only last 1000 hours each, and due to stringent regulations you’d run out of replacements pretty soon! }:-)
You can still buy them from China, just not as a “lightbulb”, but rather a “hotbulb” or something like that… Once it’s supposed to make heat, they become pretty efficient, so the bureaucrats have no power here ;-)
The “lightbulb conspiracy” that people quote is just a myth based on a propaganda documentary.
In reality the efficiency of a regular incandecent lightbulb is inversely proportional to its lifetime, so the “cartel” was really about standardizing light output per Watt.
Reason being that manufacturers were one-upping each other by promising longer and longer lasting bulbs, which meant they were either a) lying, b) selling crappy dim lighbulbs, c) both.
That was a race to the bottom and they were risking the loss of consumer trust, because consumers had no means to measure the actual light output to any meaningful accuracy and the whole thing was threatening to turn into a lemon market: https://en.wikipedia.org/wiki/The_Market_for_Lemons
Once everyone assumes your lightbulbs are going to be crap and you’re probably cheating on them, competing with quality becomes impossble and the cheaters win. Hence the need to standardize so that people can trust a 60 Watt lightbulb is going to be roughly the same brightness from all manufacturers – and the most direct way to do that was to standardize the lifespan of the filament.
You also have to remember that the standard evacuated tungsten filament lamp wasn’t the only lamp available.
For example, Nernst lamps operated by a ceramic rod, which when heated above a certain temperature became conductive and incandecent, and would operate in open air. In fact, if the pre-heater wire had gone, you could light one with a match.
The minimum temperature to conduct also meant they had some certain minimum luminous efficiency because of the black body radiation spectrum at that temperature. They were twice as efficient as the best carbon filament lamps, but less efficient than the tungsten filament lamps which didn’t lose so much heat.
So the tungsten filament lamp manufacturers risked losing the market entirely if they let the cheaters sell less efficient but longer lasting bulbs. Eventually someone would have measured the output, concluded that the bulbs are less efficient than their competition, and that would be lights out for the industry.
Thanks for this. I thought I knew the history of lightning fairly well, but I definitely learned something new today! ;)
The Nernst lamp was directly obsoleted by the tungsten filament bulb once they figured out how to make ductile tungsten wire. Before that, they had sintered tungsten wires and tantalum wires, but the tungsten was too brittle and the tantalum couldn’t reach temperatures high enough to compete with the non-evaporating ceramic glower of the Nernst lamp.
It would have been irony if a market failure had put the better product out of the market. The real irony is that the means by which the failure was prevented is spun into a cynical narrative of the sinister nature of capitalism.
Maybe.. but then only during day, for night time energy wasting you’d have to prop up a battery system and the more wastage you want the bigger and more expensive it needs to be.
No free lunch.
Batteries have a related concern called ESOEI, which is energy stored on energy invested.
It might not surprise you to know that batteries cost energy to make and use. Over its lifespan, a lead acid battery will consume half the energy you put through it, so you have to install 50% more solar panels to compensate.
…which also consume energy to manufacture and use. It’s like the Tsiolkovsky rocket equation, where by the end of it you have fuel to lift the fuel to lift the fuel to lift the payload and at the end only 5% of the mass of the rocket ends up in orbit.
Not to mention that when you have to start using solar power to actually make batteries, your ESOEI goes way down because you need to convert it to other forms to manufacture the necessary chemicals.
There isn’t a magic box that takes electricity in one end and pops out new solar panels at the other. This is the white elephant in the room that renewable energy discussions completely omit – how to actually run our societies on the forms of power that we’re building.
i think you missed the joke…
I’m not joking – I’m making a point. The reason why solar panels and batteries are cheap because they’re being made with coal, oil and gas, that all cost around 1-2 cents a kWh in their raw form.
With that, we get “renewable” energy that costs 23 cents a kWh which is how much subsidies and tax breaks are being paid to solar power in the US to keep the whole show going, and another 5 cents the producers get by selling it to utilities, or about 12 cents for net metering schemes where utilites are forced to trade solar power 1:1 with their own supply.
When the energy output is vastly more expensive than the energy input, the system is not self-sustaining. If you actually did try to use solar power to make more solar panels, the price would increase every time through. It’s the same problem with wind power, wave power… etc. every form of renewable energy we have are utterly dependent on there being cheap fossil fuels to build them.
So the point I’m making is that it’s all a huge scam.
It’s not a scam when you stop externalizing the true costs of burning fossil fuels and coal. If dirty power producers are forced to clean up their emission products before they release it for the rest of us to choke on, then solar costs are actually lower.
“Promotional articles for this journalistic piece were provided by Apple”
(or LG? they have a history together.)
“easily one of the most trusted journalistic institutions on the planet”
Bahahahahahahahaha!!!!!
…Right.
Hey don’t laugh, Journalists are experts in technology … :)
It’s true, but it’s evidence of the sad state of (english language) journalism.
This is such a non-story, if the feature exists in day to day use how is it trying to defeat a test?
Like I keep saying, somebody is whining that an unrealistic test gives unrealistic data…
::sigh:: …UK media following the lead of US media. Sensationalism and hype to generate click-revenue for their online sites.
Sure- these tests are meant to provide a guideline but they are almost always under conditions that are completely unrealistic in real-world circumstances. Until they start factoring in production facility efficiency, realistic component lifespan, realistic replacement part availability and overall repairability I’m not compelled to give a shit.
Its like CCFL bulbs and the mandates against incandescent bulbs… I’m skeptical that cheap Chinese CCFL bulbs really save all that much energy once the reduced lifespan and rarity of them actually being responsibly recycled gets factored in to offset increased production and material costs. But the current test data says one thing so the President throws his weight behind it because he’s the ‘Green’ President. Its like arming Syrian rebels that months later make up the bulk of the ISIL fighters…. by the time the truth becomes apparent its too late to do anything except shut-up about it and hope everyone else does the same.
Clearly businessmen are perfectly willing to break rules. Duh..
Samsung (and LG) were caught doing the same with their refrigerators: http://www.consumerreports.org/cro/magazine-archive/2010/february/appliances/energy-star-glitches/overview/energy-star-glitches-ov.htm
“The LG LFX21975[ST] and the LG-made Kenmore 7973[7] used about 50 percent more energy in our tests than the numbers on their labels. The Samsung RFG298AA[WP] was off by 33 percent, and the Samsung-made GE Profile PFSS9PKY[SS] by 39 percent. All closely matched their labels’ numbers when tested under the DOE protocol. Some of these tests were done by an outside lab.”
Step 1) Dig out your Kill-a-Watt, attach
Step 2) watch exactly 2 hours of your favorite show.
Step 3) report back here. Brand, model, power usage.
Kill-a-Watts aren’t nearly accurate enough and they’re easily fooled by power factor and line noise.
The kill-a-watt isn’t accurate at both high or low loads, nor does it measure inductive/capacitive loads correctly, and it’s too slow to measure rapid current drain peaks, so even PFC devices which draw loads of power in short spikes are a hit-and-miss.
Ironically, it’s marketed for tracking “phantom loads”, which are invariably very small loads with loads of inductive/capacitive and PF issues. In reality it’s most suitable for checking things like whether your electric water heater works up to the specs.
Any good recommendations for a mains power measurement device that can accurately measure PF (I should add low cost, if possible, as I have trouble justifying much more than the Kill-A-Watt)? I’m imagining this is similar to the difference between a “AC” and a “True RMS AC” multimeter
Well, many high end multimeters have the capability to measure and log power.
There isn’t a low cost alternative. All the kill-a-watts and the like are basically the same. Your electric utility co. usually has proper measuring devices they will lend you for the purpose.
Inaccuracies of the kill-a-watt aside…
Samsung
LN32B360C5D (32″ LCD)
0.18 KWH x 2h @ $0.125 = $0.02
…a penny an hour. I think I can live with it. Not to mention that my actual rate (minus availability charge) is a third less than the rate I did the math at. Actual is $0.85/day + 0.08394/kWh.
It’s the sort of thing that only matters on a statistical scale, when Samsung sell so many millions of TVs, and it all adds up.
I dunno what “motion lighting” is, but I’d guess it’s something where the LED backlights of the TV can dim, or switch off, depending on how bright or dark the pixels in front of them are supposed to be. To increase the dynamic range to more than that of a normal LCD. Of course, only works on LCDs. Then the usage would depend largely on the pictures being shown.
I’m not sure HOW you’d “pull a Volkswagen”, since testing conditions for TVs are surely sat in a room displaying video, just like actual usage conditions. How could a TV reduce power without affecting the picture? Volkswagen’s trick only worked because test conditions were nowhere as demanding as real roads were.
Sounds a bit fishy… And also a bit like a desperate attempt by a “journalist” to ride on the coat-tails of the coat-tails of a bandwagon that’s still fairly big.
The Internet’s really done a lot of damage to journalism. Or at least provided an opportunity for media companies to let the damage happen. Click-bait, churnalism, and aggregating sites written entirely by computer scripts infest the Internet, and bloody Facebook brings them customers. Or “impressions” is probably a better word.
Motion lighting sounds like dynamic overdrive to improve pixel response time. It’s basically driving up the signal amplitude (and using more power) to get the liquid crystals to flip quicker. Each pixel is basically a capacitor that has to be charged and discharged to change state, and the quicker you do that, the more power you use.
Detecting a certain standard test video or a pattern would allow the manufacturer to program the TV to turn off overdrive and generally slow things down because they’re not testing picture quality as such.
LG beat Samsung and VW to ‘defeat devices’ in consumer goods by several years. Some of their fridges went into energy saving mode when used in conditions similar to those they would be tested under, resulting in 20% better efficiency, occasionally spoiled food, and a fine by the Australian government.
Citation needed.
I tried a quick search but found no links.
Ref:
https://www.accc.gov.au/media-release/lg-australia-provides-accc-with-undertaking-over-energy-efficiency-claims
Yep, noted that in a reply above. Seems like BOTH LG and Samsung were doing it…