T2 power consumption

[art]

Hi, I'm new here, I have a new T2 and naturally have a few questions ...

There are lots of references through these forums to the power consumption in "deep standby" being less than 1W.

Does that figure come from measuring the 12V DC consumption? Or the 240V AC consumption?

Measuring the AC consumption of my box with a (good quality) clamp meter shows the following results:

• running ~21W
• standby ~19W
• deep-standby ~10W
• rear-panel power switch off ~9W
• wall outlet switch off 0W

Notice that the external power brick is wasting about 9W even when the back panel hardware switch is turned off (ie the DC consumption is zero).

I've been trying to optimise (ie minimise) the power usage but in practice deep-standby doesn't save as much as I expected.

Do I have a peculiarly in-efficient power brick? Or is this normal behaviour?

[IanB]

Your brick is possibly faulty (but see below).

IEC energy compliant switch mode power supplies (SMPS) are supposed to be less than 0.5 watt at no load and greater than 87% efficient between 10% and 90% of full load.

Shitty cheap capacitors that run hot on their way to an early grave are usually to blame for the energy waste.

Could also be confusing your power meter. To achieve very low standby consumption many SMPS do buck regulation at idle, this mode has a fairly high VA value but a near zero power factor. To unconfuse your meter add a small static restive load to the circuit, eg.. a 15watt incandescent lamp, and check the delta power between the various states. e.g. 15.0, 15.5, 15.8, 34, 36 [meter was deceived] or 15, 24, 25, 34, 36 [meter was correct]

[prl]

If the power brick is really 9VA with power factor very close to 0, and the meter is measuring VA rather than watts, I'd expect a 15W resistive load in parallel with the power brick to show a total 17.5VA (with PF 0.85). But it's a long time since I've needed to do a phase diagram

Also, if the brick really is acting as a 9W resistive load to the mains with only a small DC power out, I'd expect it to be quite warm. Hot, even.

[art]

To achieve very low standby consumption many SMPS do buck regulation at idle, this mode has a fairly high VA value but a near zero power factor. To unconfuse your meter add a small static restive load to the circuit, eg.. a 15watt incandescent lamp, and check the delta power between the various states. e.g. 15.0, 15.5, 15.8, 34, 36 [meter was deceived] or 15, 24, 25, 34, 36 [meter was correct]

This looks like the explanation. I did a few experiments with parallel resistive loads and can see the near 90deg phase shift at zero load. Behaviour is not linear, but I haven't the time or tools to go further. Thanks for the info on these power supplies, I hadn't looked closely at them before.

[IanB]

Power factor effect for sinusoidal voltage and current.



Red instantaneous power PF=1 (0 to 1 100Hz)
Blue current PF=1 (-1 to 1 50Hz)
Orange instantaneous power PF=0 (-0.5 to 0.5 100Hz)
Green current PF=0 (-1 to 1 50Hz)

[prl]

Behaviour is not linear, but I haven't the time or tools to go further. Thanks for the info on these power supplies, I hadn't looked closely at them before.

The behaviour is linear, but only in the complex plane (or equivalent vector representations)

The complex power is V.Iresistive + j.V.Ireactive (j = sqrt(-1), engineering convention, if you are a mathematician feel free to use i instead). In that form you can simply add the power components for parallel connections.

The true power is then V.Iresistive (real part of the complex power), the VA = sqrt((V.Iresistive)^2 + (V.Ireactive)^2) (norm of the complex power), the power factor PF = V.Iresistive/VA and the phase angle 𝜃 = arccos(V.Iresistive/VA) (phase angle between the voltage and the current).

[art]

When I said "not linear" that was too vague. Using your suggestion of a parallel load and measuring the complex current I get enough data to calculate the resistive and reactive currents, phase shift / power factor and real power in the various running states. The power factor improves with load, but the reactive current surprisingly (?) increases with load.

Anyway, this has all gone a bit deeper that I intended, you've explained my concern about deep-standby power consumption ... the "problem" that I was seeing was of my own making by ignoring the power factor. FWIW I see a no-load phase shift of about 88deg improving to about 78deg in deep-standby state.

[IanB]

And of course all this Power Factor, Reactive and Resistive currents are a bit moot, as the current wave forms here are anything but sinusoidal. (typically a short spike when in the idle/buck mode, and not every cycle)

[art]

And these few mA are on the same power circuit as my pool pump