T i m wrote:
On Fri, 31 Jul 2020 19:00:16 +0100, alan_m
wrote:

On 31/07/2020 15:00, T i m wrote:
On Fri, 31 Jul 2020 11:24:31 +0100, Pamela
wrote:

Is this sort of plug-in power meter suitable for measuring a desktop PC with
a switched mode power supply?

I bought something similar a long time ago but found it was giving misleading
readings. Are these better now?

https://www.amazon.co.uk/Consumption.../dp/B07FZZ17ZY
I don't think they are particularly sophisticated
According to Big Clive it has a dedicated power analyser chip with a 100
page data sheet.

Cool.

I guess what I meant was 'I don't think they are particularly
sophisticated' ... compared with professional and more sophisticated
kit that costs many times more and probably comes with a 100 page
instruction book and a certificate of calibration ...'. ;-)

If they were as good for that price that would be nice to know.

Because the 'cheap stuff' is made down to a price rather than up to a
spec (or a much higher price g), they often fail / become less
accurate at the extremes, or when measuring the more awkward stuff.
All down to the laws of diminishing returns on the cheaper gear.

Cheers, T i m

They typically use dual sigma-delta converters, with maths
in firmware. And the maths in firmware, have been slightly
tweaked over the years for better accuracy. (This is what
happens when your privately held maths are exposed to
public scrutiny.)

Yes, the meter is good, and fit for intended purpose.
With better quality analog components used inside,
the accuracy might be improved slightly. But not in
a way that affects the gross improvement these meters
bring (7W reading on the above meter, versus 100W
approximate reading with separate regular meters).

Using an ordinary digital meter to do volts,
a digital meter to do amps, the calculation of VA resulting,
the 100W, is dead wrong. You can validate the 7W number,
against the known loads inside the (sleeping) PC at the time.

Vetting a PC in S0 run state, that's harder to do.
That's why I recommend checking PC in sleep state,
then compare to any other method you want to cook up
for sleep state, to do the same task.

The refresh power for DRAM chips is known and on the
datasheet. You can compare that load, to what the
meter indicates.

You can also apply DC loads to USB ports, then figure out
roughly what kind of efficiency exists in the +5VSB converter
(a separate converter, separated from the main converter).
In the old days, +5VSB was only 50% efficient (terrible).

Paul