5 ms·

Definitely annoying when "official" looking stats are not just a bit out of date but sometimes are completely wrong. Then a tonne of the uneducated take these t

Definitely annoying when "official" looking stats are not just a bit out of date but sometimes are completely wrong. Then a tonne of the uneducated take these things as gospel and either tout them to their friends or perhaps start breaking switches by turning everything off all the time. How much to replace a broken socket? Probably about £100 by an Electrician! The guy from the energy savings trust also mentions "Even the standby light on your TV takes up energy"! An LED? Seriously? So the crazy advice is switch the TV off at the wall instead. What happens then? Every time you switch it on you have to wait for longer for the PC inside it to boot properly and connect to the network, using much more energy than a tiny LED!

vrfvr 5y ago
When your PC doesn't need to boot when not disconnecting it from the grid, it means that there is more than a LED being powered
sdflhasjd 5y ago
That LED by itself obviously doesn't use a lot of power, but it does imply more usage by other components. Last year I got a thermal camera to help with some insulation in my old house, but it's great for literally seeing this standby usage. The TV, when off, is around 2 degrees above ambient. The author of the article laughs at the mention of the VCR, but doesn't acknowledge that we might be wasting more energy with newer gadgets. Every "IoT" device that runs of mains power seems to have very inefficient power supplies, because they are regularly warm to the touch even when doing nothing: Smart bulbs (by far the worst), Alexa, Fire/Google TV sticks, etc.
- ericpauley 5y ago
  > Fire/Google TV sticks Are you saying the device itself is warm or the power supply? If it's the device it can't be caused by an inefficient power supply because the supply is external.
  - sdflhasjd 5y ago
    I mention power supplies in both the colloquial sense, but also the power supply circuitry - I know things like the ESP chips themselves don't use much power. As for my Google TV: the PSU is integrated, so I can't tell.
    - mordae 5y ago
      You have it right. Many cheap always-connected smart devices use Low-dropout regulator (LDO). What does wiki say about these? > The advantages of an LDO regulator over other DC-to-DC voltage regulators include the absence of switching noise (as no switching takes place), smaller device size (as neither large inductors nor transformers are needed), and greater design simplicity (usually consists of a reference, an amplifier, and a pass element). The disadvantage is that, unlike switching regulators, linear DC regulators must dissipate power, and thus heat, across the regulation device in order to regulate the output voltage. See for example: https://templates.blakadder.com/athom_PG01EU16A.html https://templates.blakadder.com/athom_PG01EU16A.html You can clearly see an AMS1117 (a super popular LDO), right next to the ESP board on lower right. Most LDOs have a chunk of metal tinned to the board so that the copper layer can help with heat dissipation. You can quickly identify them by this design. Here is another one on a dev board: https://hallroad.org/images/watermarked/thumbnails/465/465/detailed/24/Espressif_ESP32_WROOM_32D_Development_Board.jpg https://hallroad.org/images/watermarked/thumbnails/465/465/d... More on the topic: https://www.digikey.com/en/articles/use-adjustable-low-leakage-ldos-to-extend-battery-life-in-wearable-designs https://www.digikey.com/en/articles/use-adjustable-low-leaka...
    - PaulHoule 5y ago
      A device that's designed to handle 200W power peak might have a power supply which is 95% efficient, but it might be that you're paying that 5% all the time. (If, for instance, there is a big-ass transformer with a lot of iron in it you are making magnetic domains flip back and forth whenever the device is plugged in.) The standby mode might use 0.2W but the device burns 10W in the power supply. Old power supplies used ideas from analog electronics: if the voltage regulator was reducing the voltage from 10V to 9V then you are directing that extra volt (10% of the energy) into a resistor. More modern power supplies use digital ideas such that transistors are fully on or fully off. Because they waste less energy they have an easier time getting rid of heat, so they can be dramatically lighter and smaller... but more expensive. Other answers would be to add a second smaller power supply for the standby mode or run the standby mode off a battery or super-capacitor that charges intermittently. The cost of those improvements would definitely be noticed by the manufacturer and the consumer at the time of purchase. The consumer might not notice a small increase in the electric bill, even if it adds up to a lot over time, because it is against a high and noisy background.
- CaptArmchair 5y ago
  To my mind, advising people to "turn them completely off" is only part of the story. The other part is questioning why they are that wasteful in the first place. One explanation might be that it's a side effect of designing them to make their development and production cost effective. In that regard, I wonder to what extent consumer protection regulations would be helpful e.g. clamping down on the production of consumer electronics that consume certain wattage per hour without any proper justification.
- eropple 5y ago
  > Smart bulbs (by far the worst) Maybe some. On the other hand, the Hue sitting here is cold to the touch when not emitting light and is drawing about 0.4W right now--I checked.
  - mordae 5y ago
    0.4W is a lot. WiFi-connected ESP32 draws about 140 mA. esp8266 will be similar. They are both 3.3V devices. 3.3 * 0.140 = 0.46 Your light bulb's esp8266 (or a similar chip, zigbee is 2.4 GHz as well) is fully powered. With a switching power supply and proper light sleep you should see about 1/10 of that. API reference: https://docs.espressif.com/projects/esp8266-rtos-sdk/en/latest/api-reference/system/sleep_modes.html https://docs.espressif.com/projects/esp8266-rtos-sdk/en/late... Sleep impact: https://github.com/chaeplin/esp8266_and_arduino/tree/master/_48-door-alarm-deepsleep https://github.com/chaeplin/esp8266_and_arduino/tree/master/...
    - eropple 5y ago
      Huh--I was certain it could be better, though I didn't realize it was that high! On the other hand, my electricity is hydro and it costs like $0.009 per kilowatt-hour.
masklinn 5y ago
There is something to be said for disconnecting or hibernating devices, either because they do consume a fair bit, or because lots of device consume a lot more than you’d expect on standby. The proper thing to do is to check them with a kill-a-watt or similar, and make the decision on a case-by-case basis. Plus it’s fun checking your various devices. E.g. a desktop PC might clock in at 100W when unused depending on the components, so if it’s not necessary hibernating it at night is a significant gain (if it has a working hibernate mode) for very little inconvenience. There are also devices which you only use once a month but sleep in at a few watts, completely unnecessarily.
- nicoburns 5y ago
  > E.g. a desktop PC might clock in at 100W when unused depending on the components 100W when unused! What on earth causes that? My laptop draws 100W under full load.
  - Y-bar 5y ago
    My 9700K desktop draws 45-50W (excluding monitor) when idle. Main usage is: - GPU at about 11-15W - Fans on motherboard and case 7-10W - Disks (1 HDD, 2xSSD): 5-10W - Peripherals (keyboard, mouse, headset, ...) 2-5W - CPU 5-15W I can imagine a scenario with additional peripherals, beefier GPU, more PCI cards, more disks could absolutely reach upwards to 100W.
    - ascagnel_ 5y ago
      What power level is the system at? That sounds reasonable for a system that's in its active mode (ACPI S0) but not under load, but would be extremely high for any other sleep mode (S1-S5).
      - Y-bar 5y ago
        I have never been able to consistently get Windows 10 to enter deep sleep for more than about 30 minutes, there is always something waking it up. So I have given up and try to remember to shut the computer off when not using it.
everdrive 5y ago
It's easy enough to plug a power strip into wall socket and simply turn off the power strip when not in use. I'm not suggesting that this means the passive power draw of these devices is enough to warrant the effort. However, it's also effortless just to turn off the power strip when not watching TV. We only watch TV once or twice a week, and have a pretty unusual setup: - large CRT - VCR - DVD player Given the ancient appliances, and how infrequently we watch TV, it does actually pay off to disconnect them via a surge protector.
woliveirajr 5y ago
This. Some things have a kind of "hidden cost" that aren't considered when doing those savings maths. The cost of turning up some "microCPU" vs. the cost of the led + sleeping sycles. The pollution cost of slowing down, stoping, driving again vs. stoping on bumpers. The costs of keeping the same old refrigerator vs. energy savings during 5 years.
pbhjpbhj 5y ago
If you think that a PC maintains a network connection only by keeping an LED lit then either your analytic abilities aren't up to this discussion and/or you're being purposefully disingenuous. >How much to replace a broken socket? // So, waste electricity because the 5¢ socket switch might break? What's wrong with turning things off when you're not using them.

vrfvr 5y ago

When your PC doesn't need to boot when not disconnecting it from the grid, it means that there is more than a LED being powered

sdflhasjd 5y ago

That LED by itself obviously doesn't use a lot of power, but it does imply more usage by other components. Last year I got a thermal camera to help with some insulation in my old house, but it's great for literally seeing this standby usage. The TV, when off, is around 2 degrees above ambient. The author of the article laughs at the mention of the VCR, but doesn't acknowledge that we might be wasting more energy with newer gadgets. Every "IoT" device that runs of mains power seems to have very inefficient power supplies, because they are regularly warm to the touch even when doing nothing: Smart bulbs (by far the worst), Alexa, Fire/Google TV sticks, etc.

ericpauley 5y ago
> Fire/Google TV sticks Are you saying the device itself is warm or the power supply? If it's the device it can't be caused by an inefficient power supply because the supply is external.
- sdflhasjd 5y ago
  I mention power supplies in both the colloquial sense, but also the power supply circuitry - I know things like the ESP chips themselves don't use much power. As for my Google TV: the PSU is integrated, so I can't tell.
  - mordae 5y ago
    You have it right. Many cheap always-connected smart devices use Low-dropout regulator (LDO). What does wiki say about these? > The advantages of an LDO regulator over other DC-to-DC voltage regulators include the absence of switching noise (as no switching takes place), smaller device size (as neither large inductors nor transformers are needed), and greater design simplicity (usually consists of a reference, an amplifier, and a pass element). The disadvantage is that, unlike switching regulators, linear DC regulators must dissipate power, and thus heat, across the regulation device in order to regulate the output voltage. See for example: https://templates.blakadder.com/athom_PG01EU16A.html https://templates.blakadder.com/athom_PG01EU16A.html You can clearly see an AMS1117 (a super popular LDO), right next to the ESP board on lower right. Most LDOs have a chunk of metal tinned to the board so that the copper layer can help with heat dissipation. You can quickly identify them by this design. Here is another one on a dev board: https://hallroad.org/images/watermarked/thumbnails/465/465/detailed/24/Espressif_ESP32_WROOM_32D_Development_Board.jpg https://hallroad.org/images/watermarked/thumbnails/465/465/d... More on the topic: https://www.digikey.com/en/articles/use-adjustable-low-leakage-ldos-to-extend-battery-life-in-wearable-designs https://www.digikey.com/en/articles/use-adjustable-low-leaka...
  - PaulHoule 5y ago
    A device that's designed to handle 200W power peak might have a power supply which is 95% efficient, but it might be that you're paying that 5% all the time. (If, for instance, there is a big-ass transformer with a lot of iron in it you are making magnetic domains flip back and forth whenever the device is plugged in.) The standby mode might use 0.2W but the device burns 10W in the power supply. Old power supplies used ideas from analog electronics: if the voltage regulator was reducing the voltage from 10V to 9V then you are directing that extra volt (10% of the energy) into a resistor. More modern power supplies use digital ideas such that transistors are fully on or fully off. Because they waste less energy they have an easier time getting rid of heat, so they can be dramatically lighter and smaller... but more expensive. Other answers would be to add a second smaller power supply for the standby mode or run the standby mode off a battery or super-capacitor that charges intermittently. The cost of those improvements would definitely be noticed by the manufacturer and the consumer at the time of purchase. The consumer might not notice a small increase in the electric bill, even if it adds up to a lot over time, because it is against a high and noisy background.
CaptArmchair 5y ago
To my mind, advising people to "turn them completely off" is only part of the story. The other part is questioning why they are that wasteful in the first place. One explanation might be that it's a side effect of designing them to make their development and production cost effective. In that regard, I wonder to what extent consumer protection regulations would be helpful e.g. clamping down on the production of consumer electronics that consume certain wattage per hour without any proper justification.
eropple 5y ago
> Smart bulbs (by far the worst) Maybe some. On the other hand, the Hue sitting here is cold to the touch when not emitting light and is drawing about 0.4W right now--I checked.
- mordae 5y ago
  0.4W is a lot. WiFi-connected ESP32 draws about 140 mA. esp8266 will be similar. They are both 3.3V devices. 3.3 * 0.140 = 0.46 Your light bulb's esp8266 (or a similar chip, zigbee is 2.4 GHz as well) is fully powered. With a switching power supply and proper light sleep you should see about 1/10 of that. API reference: https://docs.espressif.com/projects/esp8266-rtos-sdk/en/latest/api-reference/system/sleep_modes.html https://docs.espressif.com/projects/esp8266-rtos-sdk/en/late... Sleep impact: https://github.com/chaeplin/esp8266_and_arduino/tree/master/_48-door-alarm-deepsleep https://github.com/chaeplin/esp8266_and_arduino/tree/master/...
  - eropple 5y ago
    Huh--I was certain it could be better, though I didn't realize it was that high! On the other hand, my electricity is hydro and it costs like $0.009 per kilowatt-hour.

masklinn 5y ago

There is something to be said for disconnecting or hibernating devices, either because they do consume a fair bit, or because lots of device consume a lot more than you’d expect on standby. The proper thing to do is to check them with a kill-a-watt or similar, and make the decision on a case-by-case basis. Plus it’s fun checking your various devices. E.g. a desktop PC might clock in at 100W when unused depending on the components, so if it’s not necessary hibernating it at night is a significant gain (if it has a working hibernate mode) for very little inconvenience. There are also devices which you only use once a month but sleep in at a few watts, completely unnecessarily.

nicoburns 5y ago
> E.g. a desktop PC might clock in at 100W when unused depending on the components 100W when unused! What on earth causes that? My laptop draws 100W under full load.
- Y-bar 5y ago
  My 9700K desktop draws 45-50W (excluding monitor) when idle. Main usage is: - GPU at about 11-15W - Fans on motherboard and case 7-10W - Disks (1 HDD, 2xSSD): 5-10W - Peripherals (keyboard, mouse, headset, ...) 2-5W - CPU 5-15W I can imagine a scenario with additional peripherals, beefier GPU, more PCI cards, more disks could absolutely reach upwards to 100W.
  - ascagnel_ 5y ago
    What power level is the system at? That sounds reasonable for a system that's in its active mode (ACPI S0) but not under load, but would be extremely high for any other sleep mode (S1-S5).
    - Y-bar 5y ago
      I have never been able to consistently get Windows 10 to enter deep sleep for more than about 30 minutes, there is always something waking it up. So I have given up and try to remember to shut the computer off when not using it.

everdrive 5y ago

It's easy enough to plug a power strip into wall socket and simply turn off the power strip when not in use. I'm not suggesting that this means the passive power draw of these devices is enough to warrant the effort. However, it's also effortless just to turn off the power strip when not watching TV. We only watch TV once or twice a week, and have a pretty unusual setup: - large CRT - VCR - DVD player Given the ancient appliances, and how infrequently we watch TV, it does actually pay off to disconnect them via a surge protector.

woliveirajr 5y ago

This. Some things have a kind of "hidden cost" that aren't considered when doing those savings maths. The cost of turning up some "microCPU" vs. the cost of the led + sleeping sycles. The pollution cost of slowing down, stoping, driving again vs. stoping on bumpers. The costs of keeping the same old refrigerator vs. energy savings during 5 years.

pbhjpbhj 5y ago

If you think that a PC maintains a network connection only by keeping an LED lit then either your analytic abilities aren't up to this discussion and/or you're being purposefully disingenuous. >How much to replace a broken socket? // So, waste electricity because the 5¢ socket switch might break? What's wrong with turning things off when you're not using them.