Power measurements

[VoxUnius]

Here are my observations.

OS: factory Debian.
OS storage: MicroSD.
State: LightDM off, screen brightness 0, CPU governor - ondemand.
Additional equipment: USB Ethernet adapter.
Task: mainline kernel compilation.
Parameters: -j6, storage - NVMe.
Power supply: 4A unit.

[pfeerick]

Earlier this day, I measured all three with a 1 liner
Copy/paste removed the tabs.

Show off! And code tags probably would have kept the tabs.

That's some pretty major movement... 4.35v when 100%, then down to 4.20 (still 100%?), and then suddenly 3.97v (91%)... would have to be a pretty major current spike... more than I would expect to see on a SBC... maybe the charge IC throttled down a bit because it was warming up? Either way, both your and VoxUnius's results suggest 3A isn't keeping up when the PBP is loaded up.

Are current measurements not working atm? This is the sort of thing I get for the A64 Pinebook... negative when the battery is discharging, positive when the battery is charging... possibly negative if the external PSU couldn't keep up with the load. I was expecting to see something more interesting than zeros! 

Code:

-------------------------------------------------------------
|    date    |   time   | Voltage | Current | SoC | Elapsed |
|            |          |         | (mA)    | (%) | Time    |
--------------------------------------------------|----------
| 2019-11-22 | 19:11:30 | 3.97    | -773    | 90  | 0:00:01 | battery only, no artificial load
| 2019-11-22 | 19:12:31 | 3.83    | -2105   | 90  | 0:01:01 | battery only, cpuburn running
| 2019-11-22 | 19:13:31 | 4.07    | 223     | 90  | 0:02:01 | external 2.7A capable supply, cpuburn running
| 2019-11-22 | 19:14:31 | 4.07    | 204     | 90  | 0:03:01 | " "

[VoxUnius]

Earlier this day, I measured all three with a 1 liner
Copy/paste removed the tabs.

Show off!  And code tags probably would have kept the tabs.

That's some pretty major movement... 4.35v when 100%, then down to 4.20 (still 100%?), and then suddenly 3.97v (91%)... would have to be a pretty major current spike... more than I would expect to see on a SBC... maybe the charge IC throttled down a bit because it was warming up? Either way, both your and VoxUnius's results suggest 3A isn't keeping up when the PBP is loaded up.

Are current measurements not working atm? This is the sort of thing I get for the A64 Pinebook... negative when the battery is discharging, positive when the battery is charging... possibly negative if the external PSU couldn't keep up with the load. I was expecting to see something more interesting than zeros! 

Code:

-------------------------------------------------------------
|    date    |   time   | Voltage | Current | SoC | Elapsed |
|            |          |         | (mA)    | (%) | Time    |
--------------------------------------------------|----------
| 2019-11-22 | 19:11:30 | 3.97    | -773    | 90  | 0:00:01 | battery only, no artificial load
| 2019-11-22 | 19:12:31 | 3.83    | -2105   | 90  | 0:01:01 | battery only, cpuburn running
| 2019-11-22 | 19:13:31 | 4.07    | 223     | 90  | 0:02:01 | external 2.7A capable supply, cpuburn running
| 2019-11-22 | 19:14:31 | 4.07    | 204     | 90  | 0:03:01 | " "

A tiny correction: even 4A is not enough. Either that, or some internal logic limits the current from the power adapter.

[hmuller]

I would expect the voltage to be constant perhaps with some small swings. Perhaps you observed small swings?

There were small fluctuations in voltage_now during all tests. During -j6 test the maximum recorded value was 4.2, minimum of 4.1, with an average of 4.2.

[Der Geist der Maschine]

Earlier this day, I measured all three with a 1 liner
Copy/paste removed the tabs.

Show off!

It's not as elegant as your code, of course

And code tags probably would have kept the tabs.

But then you get scroll bars and who scrolls them down?

That's some pretty major movement... 4.35v when 100%, then down to 4.20 (still 100%?), and then suddenly 3.97v (91%)... would have to be a pretty major current spike... more than I would expect to see on a SBC... maybe the charge IC throttled down a bit because it was warming up? Either way, both your and VoxUnius's results suggest 3A isn't keeping up when the PBP is loaded up.

Are current measurements not working atm?

I see current measurements working when solely running from battery. I interpret my data that there is a huge current spike between two measurement points and current is always is 0 on the actual measurement points.

This is the sort of thing I get for the A64 Pinebook... negative when the battery is discharging, positive when the battery is charging... possibly negative if the external PSU couldn't keep up with the load. I was expecting to see something more interesting than zeros! 

Code:

-------------------------------------------------------------
|    date    |   time   | Voltage | Current | SoC | Elapsed |
|            |          |         | (mA)    | (%) | Time    |
--------------------------------------------------|----------
| 2019-11-22 | 19:11:30 | 3.97    | -773    | 90  | 0:00:01 | battery only, no artificial load
| 2019-11-22 | 19:12:31 | 3.83    | -2105   | 90  | 0:01:01 | battery only, cpuburn running
| 2019-11-22 | 19:13:31 | 4.07    | 223     | 90  | 0:02:01 | external 2.7A capable supply, cpuburn running
| 2019-11-22 | 19:14:31 | 4.07    | 204     | 90  | 0:03:01 | " "

---

I assume if we disconnect the battery, then the Pinebook Pro will shut down on load. Basically, we can't disconnect the battery.

[pfeerick]

A tiny correction: even 4A is not enough. Either that, or some internal logic limits the current from the power adapter.

Er no... I don't think so... as I mentioned before...

the lithium charger IC is limited to 3A input current to protect the power supply

... so if you provide more than 3A... it won't doing anything other than ensure that the PSU isn't overloaded.

---

But then you get scroll bars and who scrolls them down?

Me! Data is groot! :-P And IIRC it'll only scroll if you do more than 10 or so lines.

I see current measurements working when solely running from battery.

Damn... maybe the power monitoring on the PBP isn't as good/detailed as on the PB? Due to the different PMIC and CPU combination.

I assume if we disconnect the battery, then the Pinebook Pro will shut down on load. Basically, we can't disconnect the battery.

I suspect so... if it's having to use the battery, then it's obviously not letting enough come in from the external supply.