PC on Sleep

[Radio Fixer]

Was quite surprised and impressed with the PC running W10 and using the "Sleep" option rather than the less convenient "Shut Down".

Power drops to 6W on Sleep against 106 to 130 something depending what Apps I have open.

Gary

[PerdioPal]

Some PCs have a sleep for so long, quick start using a bit of power, and then hibernate, no power. Great for laptops.

[Sinewave]

I want to measure the power used by my computers. I have a few which would have in the past been on all the time. A desktop, laptop, a small format thingy bob for running FT8 - they all talk to a synology server to access the same files if needed, I wonder what power that is using too.

[Mark Hennessy]

What are you using to measure power?

I've got a Brennenstuhl PM230 that I bought from Maplin some 15+ years back. I have reason to doubt its accuracy at these low power levels, so would like to try something else. I think this is the 3rd time I've asked this question recently on this forum, but no "bites" so far

For as long as I can remember, I've put my PCs to sleep rather than powered down fully. In the XP days, a power cut during sleep meant the PC would start from cold when you next switched them on, and probably nag you about not shutting them down properly IIRC.

With W10, power cuts are not a problem - aside from having to hit the front panel power button rather than mashing the keyboard or clicking the mouse to wake them up, they just come back exactly as they were. So they must dump the RAM to HDD or similar - if it's keeping the RAM powered (and refreshed) then I'd expect a similar cold start followed by a nag after the outage.

I have less success with my laptop. If I put it to sleep, the battery does run down reasonably quickly - I suspect that it powers itself up and installs updates (sometimes I look at the event log to see, but normally I'm running about setting up a gig, so don't have time to investigate). Occasionally, even when completely shut down, the battery will still run down. Things improved after a BIOS update which fixed a bug that was introduced by - guess what - a BIOS update! It would probably be fine if I used it at home regularly, but it mostly lives in a flight case awaiting the next gig. I check on it roughly once a fortnight, and allow it to update, etc, at the same time.

When it full "sleep", I'd expect them to use about the same amount of power that they take when plugged in but not switched on. On other words, the main PSU is not running, and just the standby PSU is alive. The standby PSU is normally rated at 10 watts or so, but what is taken from the wall socket will depend on what the motherboard takes. When fully shut down, it only needs to respond to the power switch and wake-on-LAN if enabled (which it usually is by default, but turning that off is well worth doing). When in sleep, it additionally needs to keep the keyboard and mouse powered, but not much else - as mentioned, I don't think it needs to keep the RAM powered and refreshed, else it wouldn't survive power cuts.

You'd hope that the one watt initiative applies to PCs. Or half watt initiative, as it now is. All my PCs are ancient, and don't give specs on standby power consumption. My DIY machines have decent PSUs that claim high efficiency, but I've not been able to measure it reliably. Which is where I came in...

[Radio Fixer]

I use a "Plugin Power & Energy Monitor" possibly also from Maplin years ago.
Measures: Volts, Amps, Watts, VA, Hz, PF, KWH. How accurate? Somewhere there maybe a booklet?
Fine for what I do and for comparisons.

Gary

[Mark Hennessy]

Classic GVR thread-drift

I think that discussing how to measure very small operating powers, such as a PC in sleep, is just about on-topic. But talking about measuring incoming power to a large installation is definitely off-topic for this thread, so I'll move those posts to a new thread. I'll also move this thread to the Modern Technology forum.

[jimmc101]

There are quite a few Chinese modules which I believe use a chip designed for smart meters.
I use this one PZEM-022

 
 Although claimed to read down to 0.2W, current measurement only works above 20mA.
I boxed it with a 13A socket and a switch to select 1 or 10 turns on the current transformer to give better sensitivity.

On the 10turn setting:
100k resistor across the mains gave 242V, 2.5mA, 0.585W, 0.97pF
200k                                                242V, 0.0mA, 0.294W, 1.0pF

A USB power supply with no load gave 242V, 2.9mA, 0.229w, 0.31pF

So it looks reasonably accurate at low powers.

Jim

[Mark Hennessy]

Thanks, Jim - that does look interesting.

Some years back I bought a D52-2048 module, which fits a DIN rail and, and did plan to build it into a box with a socket, but as the best current resolution is only 10mA, I haven't bothered yet. It might be better suited to a workshop consumer unit or similar... The PZEM module looks more useful. One question - you're saying the supplied transformer has two tappings, or have I misunderstood?

Thanks again,

Mark

[jimmc101]

The current transformer is supplied with a 1000 turn secondary only, for the primary I used 1 turn of 1.5mm plus 9 turns of 1.25mm flex which was about the maximum that the aperture in the core would comfortably take (see below).

   




The meter was built into an ABS Verobox 180x110x56.5. A Crabtree GB6:4304/D socket was fitted with a 16mm back-box behind the ABS to give a mounting surface for the C.T.




   




Jim

[Mark Hennessy]

Hi Jim - thanks for the extra details, that makes perfect sense. I'll investigate further when things ease up for me.

Cheers,

Mark