[SOLVED] Red light flashes next to barrel-port

[FreeRangeRon]

>further suggestions.
Again, pg4&5, data sheet
2,3,4 = 5V
12 = 3.3V (reference)
18 = 0.5 to 1.6 V
These are  input, but not 12
My guess is that the short has damaged it (but only a guess),
too bad you didn't check if it was really shorted (continuity)
If 14 is out of 2.025-2.1 V range, charging is shut down (pg22)
If the battery is too low it may fault too, pin 15 (should be 3.6 or >),
may slow charge down to 3.0

Thanks for the info wdt. I agree I should've checked for continuity before 'fixing' the 'short'. I have since soldered some through-hole resistor legs onto my meter's probes so that I can make good contact with the ICs pins.

2,3,4 = 5.1v
12 = 3.3v
18 = 1.25v
15 = 3.75v (battery voltage)
10 = 2.02v

The readings from the other pins look fine, and should allow charging to happen. But not when pin 14 is only reading 1.8v.

I don't know if there's any action I can take to change the value of FB (14). There's a resistor network that sets the battery voltage (R49, R51) but I'd need to get them off the board to test them, and I don't have a rework station which makes that pretty challenging. The same goes for the resistors (R52, R54) that set the temperature limits on TS (10).

When I was taking some readings I managed to cause a short and break the trace between R43 and C122, so that's had a scrapy repair job done and will have nullified any chance of sending the unit back to Pine for repair/replace.

It might be time to call it quits. I can probably either:

1) Buy a rework station and replace the BQ24171 (expensive)
2) Try and buy only the PBP mainboard from Pine (Best option, for the right price)
3) Build another charging circuit and squeeze it into the case (whole new project)

Cheers,
Ron

[wdt]

I assume that it never did not blink (when charging)?
I would guess that 14 can not stand 3.5+V
Send photo showing short when you order new mb,
maybe you will get better price

[FreeRangeRon]

Correct. It has always blinked when 'charging'. 

When support finally get around to replying to me, I'll share the picture of the short. Support seem to reply once every two or three days, from the other stories I've read I'm not really expecting much help from them but that's for another topic.

Thanks for your help with this wdt, and all other contributors.

Cheers,
Ron

[moonwalkers]

So looks like the root cause of the red light flashing (the battery not charging) can vary from case to case. But in my case I'm pretty sure it is related to power draw - just now I caught an exact balance where I pause download for single file and the light doesn't flash, but as soon as I resume the single download thread it start flashing. I pause it again and it stops flashing. Immediately. Now, I had cooling pad plugged in into one of the USB ports, with fans turned on. As soon I turned off the fans I could have one and even more download threads without red light flashing. And the reaction is pretty much immediate.

[dsimic]

Here's another insight...

While building Manjaro's Linux kernel package, which pegs the CPU to 100% over a prolonged period of time and generates a lot of heat, I've noticed that the red LED next to the barrel plug starts flashing, and the battery starts to drain rapidly.  After a brief investigation, everything points into the direction of the battery charging IC (i.e., TI BQ24171) overheating as a result of increased internal temperature of the laptop.

Interestingly, if I turn the laptop upside-down, the red LED doesn't flash while bulding the kernel package.  Surely, this confirms that the root cause, in this case, is the BQ24171 IC overheating as a result of the heat generated by the SoC/CPU, in which case the BQ24171 IC stops charging the battery.

I'll try to affix a small heatsink to the BQ24171 IC, which might help a bit, and post the results back here.

[dsimic]

Unfortunately, the red LED starts flashing after about half an hour of the compilation described in the previous post, with the laptop in upside-down position.  However, it's pretty safe to assume that the root cause is heat soak from the SoC.

[xyzzy]

Unfortunately, the red LED starts flashing after about half an hour of the compilation described in the previous post, with the laptop in upside-down position.  However, it's pretty safe to assume that the root cause is heat soak from the SoC.

Could it be battery temperature?  The STAT led will blink if the battery temp sensor is out of range.

It's not clear to me if the STAT led will blink on thermal shutdown, from §9.3.20:

As added level of protection, the charger converter turns off and self-protects whenever the junction temperature exceeds the TSHUT threshold of 150°C

It sounds like maybe it just turns off, rather than blink.  It does limit charge current at 120°.  Perhaps that causes blinking?

Also, I don't think the SoC can get to 150.  It should shutdown well before that.  If the SoC is at 90°, there's no way to can raise the temperature anywhere else higher than that.

The MOSFETs for the charger are internal to the BQ24171, and the power supply design of the PBP forces the BQ24171 to regulate the 5V input down to the battery 4.2V even when the battery is fully charged.  A more sophisticate design would allow the BQ24171 to bypass the battery when fully charged and directly connect the DC input.

This means the QB24171 is dissipating not insignificant amounts of power itself and might be hitting its thermal limit without help from the SoC.

[wdt]

From
https://www.ti.com/lit/ds/symlink/bq2417...52FBQ24171
,,STAT pin,,
It indicates various charger operations:LOW when charge in progress.HIGH when charge is complete or in SLEEP mode.Blinking at 0.5 Hz when fault occurs,including charge suspend,input over voltage,timer fault and battery absent

[xyzzy]

From
https://www.ti.com/lit/ds/symlink/bq2417...52FBQ24171
,,STAT pin,,
It indicates various charger operations:LOW when charge in progress.HIGH when charge is complete or in SLEEP mode.Blinking at 0.5 Hz when fault occurs,including charge suspend,input over voltage,timer fault and battery absent

Does the bq241717 reaching the 120°C thermal slowdown count as "charge suspend, input over voltage, time fault, or battery absent?"

If you read TI's board for the bq24171, it seems a lot of things can trigger fault flashing, and it can be hard to tell what is the cause.

[wdt]

I guess if it really reaches 120, you wouldn't be able to touch it, even a few minutes later