> For example, right now my laptop is at 63% battery and currently charging. It’s drawing 36W at the moment. When the battery charge is lower, it’s drawing more power from the outlet, and the higher the battery charge is getting, the less power it’s drawing from the outlet.

This is because Li-Ion charging logic is known as "CC-CV", or constant current followed by constant voltage. You limit the charging current to some value (say 1A) until the cell attains the target voltage (almost always 4.2V, though some chargers limit it to 4.1V to prolong cell life), and then you hold it at that voltage until the current diminishes significantly (most chargers cut the cell off and indicate charge complete when the current draw drops to 10% of its max (during the CC phase) charge current, i.e. 100mA here).

Ages ago I measured how much power it took for the Start menu to open in Windows 7 on a Dell desktop that was fairly average at the time. In my somewhat crude measurement it was 20W for about 2 seconds.

Brilliant! Thanks for measuring this - I know it may be crude, but it’s also the best measurement I have ever heard of for this!

Assuming you like that kind of thing, maybe you can also test the power drain from displaying seconds in the taskbar in Windows 11. I know Raymond Chen posted an article about it, but I’d be interested whether you can spot a difference. If it really is on the order of 5 mW, then I assume you can’t detect it.

One of the downsides of only using a laptop is that you can’t see this level of detail because the battery acts as a buffer.

Imagine what the draw is for opening some bloated electron monstrosity like Teams or Discord these days.

I like those monitors for finding weird, surprising (to me anyway) things - like when I charge my Framework laptop from a USB port on my work laptop (because I don’t have another power socket handy to plug them both into the wall) the Framework laptop draws twice as much power when it’s asleep as when it’s awake. The opposite way around to what I need!

From memory, 5W when running (not enough to prevent the battery slowly draining), 10W when in standby.

I love my AMD framework, and I don't think my numbers are as terrible as yours (7-9 watts just chilling watts idle. However, the 'sleep' use is still, bad. I hunch it's a linux thing. I don't even bother, I just turn the thing off.

The analogy I always use is the filling a water bottle one. In the beginning when the bottle is empty you can go full power and fill it up with high pressure. At the end you need to reduce the pressure to not spill the water. I know it doesn’t work like this with batterie cells but close enough. I had the same aha moment when realizing this. It’s one of the things no one normally thinks about in a world where everything is a given.

Or the opposite function of an audio potentiometer (logarithmic potentiometer).