Pogo pins power clarification - reading schematics

[bokomaru]

The wiki seems unclear to me. Wiki - Pogo Pins [edit: link to an old version of the wiki page before changes that were made as a result of this thread. See also the old diagram image ]

There's an image describing the pinout, and there's a breakout board design which seems to match.

Code:

INT  SDA  SCL
VBUS VBAT GND

Below that, the wiki says

The VBUS pin is powered by USB and is 5V at up to 500mA. The second power pin is VBAT, which connects to the battery voltage.

I wonder what "connects to battery voltage" means...

Looking at schematics and datasheets...

I don't see the pogo pins labeled on the physical mainboard. If I put together the component placement drawing (or the pinout on the wiki) with the schematics, I suppose the pinout goes like this.

Code:

PIN6 PIN2 PIN4
PIN5 PIN1 PIN3

or

Code:

INT    SDA  SCL
USB-5V DCIN GND

Now...

DCIN: DCIN from J1300 (40PIN) on the mainboard connects to VBUS on J101 (40PIN) on the USB-C small board, which goes to VBUS on J101 (USB_TYPE_C_RECEPTACLE). So DCIN would be the 5V input from a USB-C charger. It goes to ACIN1/ACIN2/VBUS on U600 (AXP803 PMIC). If ACIN/ACIN2/VBUS is available, the AXP803 uses that as the power source and charges the battery. DCIN also goes to pogo PIN1.

USB-5V: PS (IPSOUT? Power Supply?) comes from U600 (AXP803 PMIC) and feeds L606 (LP6226CB6F boost converter), to make USB-5V. USB-5V arrives at pogo PIN5.

DCIN and USB-5V loop: Don't know exactly how this works, but if DRVVBUS is enbled, USB-5V actually provides DCIN through U1302 (LPW5206H USB power loading switch). So we can provide a 5V output for a USB device on the USB-C small board J101 (USB_TYPE_C_RECEPTACLE).

VBAT: In schematics, VBAT comes from J600 (the battery connector). Pretty sure it never touches pogo PIN1/DCIN nor pogo PIN5/USB-5V. The wiki calls pogo PIN1 "VBAT".

Questions:

1. Is that a good read? Did I get it wrong?
   
2. Did I get the pinout right? Why are PIN1 and PIN2 in the middle?
   
3. Does USB-5V/PIN5/VBUS provide a 5V output up to 500 mA?
   
4. Is DCIN/PIN1/VBAT a 5V input for powering the device and charging the battery?
   
5. What happens if you connect a 5V input to DCIN/PIN1/VBAT and to the USB-C connector at the same time? Would that short together two power sources?
   
6. "VBAT, which connects to the battery voltage" would make me think that pogo PIN1 basically connects to the positive battery terminal VBAT. Not true though, right?
   

Reason for asking: Obviously I'd like to use the pogo pins. But we should also clarify the wiki.

[hiimtye]

the schematic on the breadboard you linked explains it fairly well. the top row is for communication, INT/I2C, and the bottom row is for power, 5V/BAT, and GND/N.

the bottom row can be connected to power your device, either from the 5V AC->DC source, or the DC source supplied by the battery, or both. the GND/N is your return path

[bokomaru]

the bottom row can be connected to power your device, either from the 5V AC->DC source, or the DC source supplied by the battery, or both.

So are you saying that PIN5/USB-5V and PIN1/DCIN are both only outputs from the PinePhone?

Doesn't DCIN go to the USB-C connector? Can you power the phone (charge the battery) via the pogo pins?

[scholbert]

the bottom row can be connected to power your device, either from the 5V AC->DC source, or the DC source supplied by the battery, or both.

So are you saying that PIN5/USB-5V and PIN1/DCIN are both only outputs from the PinePhone?

Doesn't DCIN go to the USB-C connector? Can you power the phone (charge the battery) via the pogo pins?

hiimtye explained it already...
USB-5V and DCIN should be used to power your extension board... so, yes they should be seen as outputs.

As DCIN is directly connected to the USB Type-C port it's on the same rail.
From my point of view it should not be used to power the phone or charge the battery, because it may conflict with an attached USB Type-C charger.
Powering DCIN from the pogo pins, conflicts with the logic inside the phone and there's no protective element in between.

Without a charger connected to USB Type-C DCIN is normally not present... as long as you use a USB-OTG cable to switch the role.

USB-5V is delivered from a DCDC-regulator and used in general to power attached USB devices with the Pinephone is in the role of the host.
In this case it is switched to USB Type-C connector.

When using USB-5V as a power rail it should definitely be used as a power source coming from the Pinephone.
It could be controlled by GPIO PD8-VCC5V_EN

Hope this helps to understand and does not confuse more

Cheers,
scholbert

[bokomaru]

Thanks both for the replies. Forgive me if I seem stubborn. Still want to talk through this if you'll humor me :-)

I'm not asking about any single use case, but want to understand what PIN1 and PIN5 are and how they work. I'm sure that we can power an expansion board from pogo pins, we can set that aside as a given.

USB-5V is delivered from a DCDC-regulator and used in general to power attached USB devices with the Pinephone is in the role of the host.
In this case it is switched to USB Type-C connector.

When using USB-5V as a power rail it should definitely be used as a power source coming from the Pinephone.
It could be controlled by GPIO PD8-VCC5V_EN

Makes sense to me, I agree.

As DCIN is directly connected to the USB Type-C port it's on the same rail.
From my point of view it should not be used to power the phone or charge the battery, because it may conflict with an attached USB Type-C charger.
Powering DCIN from the pogo pins, conflicts with the logic inside the phone and there's no protective element in between.

I follow you that it's on the same rail. Pogo PIN1 goes directly to the PMIC's power inputs ACIN and VBUS. The USB C connector's VBUS pins also go directly there.

By "no protective element in between", you mean none between pogo PIN1 and the USB C connector? I agree. What doesn't make sense is why you couldn't power the phone from PIN1 (given USB C is not connected). I don't see any logic there to distinguish PIN1 from the USB C connector's VBUS. It's the same DCIN signal on the schematics.

Imagine a scenario: Powering a lone mainboard. There would be no USB-C smallboard attached via the 40-pin connector. Would it now be safe to connect a 5 V source to pogo PIN1?

Another scenario: this "wireless coil design used in the back case" from the december blog update. The may blog update even says "Wireless charging coming to the PinePhone via add-on / using pogo pins". My guess was the wireless charger provides 5 V to PIN1/DCIN.


Some investigation with a multimeter:

With a powered down PinePhone, I get 0 V from PIN5 and 4.0 V (battery voltage, unloaded) from PIN1.

Power up the PinePhone with no charger in the USB C port. I get 0 V from PIN5 and 3.9 V (battery voltage, loaded, modem is off) from PIN1.

PinePhone still powered on, now I plug in a USB C charger. I get 5.0 V from PIN5 and 4.8 V from PIN1.

As I understand it, 4.8 V is the voltage of my external AC to DC power brick in the wall under whatever load. I don't _think_ that's the voltage at which the battery gets charged. I think the max allowed for charging a Li battery is around 4.2 or 4.3 V, and anyways, since this battery is almost full, it's probably in the constant-voltage phase of charging. Maybe I'll hook up to a variable power supply to investigate more.

What I'm hearing is that if you're powering an expansion board from the PinePhone's pogo pins, you can pick between a 5 V source or the battery voltage. But what I'm measuring is that if you pick PIN5, you get 0 V or 5 V depending on whether or not USB-5V is enabled. If you pick PIN1, you get either the battery voltage or the USB C connector's voltage, depending on whether or not an external power source is plugged in.

USB-5V and DCIN should be used to power your extension board... so, yes they should be seen as outputs.

Without a charger connected to USB Type-C DCIN is normally not present

I noticed something else. Without the charger plugged in to the USB Type-C port. So if you use PIN1/DCIN as an output from the PinePhone. Power is going from the battery into the PMIC, and then out of the PMIC's ACIN and VBUS pins? The AXP803 data sheet says ACIN and VBUS are type "PI" (I think Power Input). Is that drawing power "backwards" through the PMIC?

[scholbert]

Some investigation with a multimeter:

With a powered down PinePhone, I get 0 V from PIN5 and 4.0 V (battery voltage, unloaded) from PIN1.

Power up the PinePhone with no charger in the USB C port. I get 0 V from PIN5 and 3.9 V (battery voltage, loaded, modem is off) from PIN1.

PinePhone still powered on, now I plug in a USB C charger. I get 5.0 V from PIN5 and 4.8 V from PIN1.

As I understand it, 4.8 V is the voltage of my external AC to DC power brick in the wall under whatever load. I don't _think_ that's the voltage at which the battery gets charged. I think the max allowed for charging a Li battery is around 4.2 or 4.3 V, and anyways, since this battery is almost full, it's probably in the constant-voltage phase of charging. Maybe I'll hook up to a variable power supply to investigate more.

What I'm hearing is that if you're powering an expansion board from the PinePhone's pogo pins, you can pick between a 5 V source or the battery voltage. But what I'm measuring is that if you pick PIN5, you get 0 V or 5 V depending on whether or not USB-5V is enabled. If you pick PIN1, you get either the battery voltage or the USB C connector's voltage, depending on whether or not an external power source is plugged in.

USB-5V and DCIN should be used to power your extension board... so, yes they should be seen as outputs.

Without a charger connected to USB Type-C DCIN is normally not present

I noticed something else. Without the charger plugged in to the USB Type-C port. So if you use PIN1/DCIN as an output from the PinePhone. Power is going from the battery into the PMIC, and then out of the PMIC's ACIN and VBUS pins? The AXP803 data sheet says ACIN and VBUS are type "PI" (I think Power Input). Is that drawing power "backwards" through the PMIC?

So to not confuse people with long post and "Quotes" i decided to split up the relevant parts... as you already did

To make it short...
I'm in doubt that the schematic gives us the correct information... and the comments i gave in my last post were based on these.
I made some measurements myself... though i try prevent a complete disassembly right now.

Checked the following connections:
USB Type-C cable connection (DCIN) to Pogo-Pins 5/1 (wiki names VBUS/VBAT) -> no direct connection -> conclusion: the schematic is wrong at this point no DCIN at Pogo Pin1 (T902)

Battery Connector VBAT (J600, Pin4) to Pogo-Pins 5/1 (wiki names VBUS/VBAT) -> no direct connection -> conclusion: no direct connection to the battery at VBAT Pogo Pin1 (T902)

To finalize here:
Pogo Pin1 (wiki name VBAT) reflects the Battery Voltage (on my device: Battery Voltage - 0.1V, as i did a direct measurement on the battery and then at the pin it to compare). I realize a short ramp up to 4.8V but then drops back to 3.9V.
It is not clear to which exact net this pin is connected. It's definetely not DCIN... nor is it VBAT (to be clear here: net names taken from the schematic).

Pogo Pin5 (wiki name VBUS) reflects the Power Supply Voltage. If it's plugged, i measure 5V, if not connected -> 0V.
Not sure about the schematic net entry about USB-5V here... i'm in doubt again that this is correct.

Imagine a scenario: Powering a lone mainboard. There would be no USB-C smallboard attached via the 40-pin connector. Would it now be safe to connect a 5 V source to pogo PIN1?

Another scenario: this "wireless coil design used in the back case" from the december blog update. The may blog update even says "Wireless charging coming to the PinePhone via add-on / using pogo pins". My guess was the wireless charger provides 5 V to PIN1/DCIN.

Yes i understand your aim...
Powering a standalone board:
As long as we don't know the exact electric path of Pogo Pin1, i would say some caution should be taken here.
We need to measure some of the nodes on the mainboard in relation to Pin1 to make things clear (reverse engineering is required).

If i should power the single board, i would design a 4.0V Power supply and attach it to the battery pins!

Wireless Charger backcase:
O.K. agreed... they wrote it on their blog... they showed us an evaluation board of a Qi-Charger... i never seen a real implementation working.
Only written words so far, so maybe in the end they found out that the Pogo Pins could not be used for that.

If they connect the "real" Qi-Charger to Pin5 (wiki name VBUS), then this would be the pin to power the boards... obviously
Again, we need to measure some of the nodes in relation to Pin5 to get things clear (reverse engineering again).

As you see i won't give a final advice here... as your mainboard may get damaged, if you try without knowing.
Right now i won't diassemble my Pinephone again, to help you with measurements... i'm sorry!

Hope this helps though!
Best regards,
scholbert

[bokomaru]

Here comes lots more quotes and replies!

I made some measurements myself... though i try prevent a complete disassembly right now.

I'm trying to avoid disassembling too. No worries :-)

Battery Connector VBAT (J600, Pin4) to Pogo-Pins 5/1 (wiki names VBUS/VBAT) -> no direct connection -> conclusion: no direct connection to the battery at VBAT Pogo Pin1 (T902)

Confirmed.

I noticed something else. Without the charger plugged in to the USB Type-C port. So if you use PIN1/DCIN as an output from the PinePhone. Power is going from the battery into the PMIC, and then out of the PMIC's ACIN and VBUS pins? The AXP803 data sheet says ACIN and VBUS are type "PI" (I think Power Input). Is that drawing power "backwards" through the PMIC?

I'm starting to think this is true. It explains why we see the battery voltage when the PMIC doesn't have an ACIN/VBUS input, and why I see a higher voltage at DCIN when ACIN/VBUS is present.

There's one thing it doesn't explain. See (3) below.

USB Type-C cable connection (DCIN) to Pogo-Pins 5/1 (wiki names VBUS/VBAT) -> no direct connection -> conclusion: the schematic is wrong at this point no DCIN at Pogo Pin1 (T902)

Measurement confirmed, but I'm not convinced about the conclusion. I hope the schematics are accurate!

Going by schematics, it's true that the same DCIN net doesn't actually go _all_ the way to the USB C connector. On the USB-C small board, there's one component U102, which is an AW338XX "load switch", which is I guess supposed to break the connection if there's too much current or voltage. Maybe the AW338XX doesn't behave like a zero resistance wire for whatever signal our multimeters use for checking connectivity?

PinePhone still powered on, now I plug in a USB C charger. I get 5.0 V from PIN5 and 4.8 V from PIN1.

As I understand it, 4.8 V is the voltage of my external AC to DC power brick in the wall under whatever load.

Confirmed.

I broke out a female/host USB A connector to play the part of a "power brick". Just the VCC pin and GND. Hooked it up to a variable power supply.

Made at least 20 voltage measurements.

Let "Vs" be the "supply voltage" output setting on the variable power supply. Let Vb be the "battery voltage" that you'd measure directly across the +/- terminals on the removed battery. Now, for Vb ~= 4.08 V, and Vs = 4.70 V, 4.80 V, 4.90 V, 5.00 V, 5.10 V, 5.20 V, or 5.30 V:

Phone off:

Code:

PIN5 = Vs - 0.20 V
PIN1 = Vs - 0.35 V

Phone on:

Code:

PIN5 = Vs - 0.20 V
PIN1 = Vb

Example:

Code:

When Vs = 5.00 V,
PIN5 = 5.00 V - 0.20 V = 4.80 V

It's just a rough formula to describe that pattern I saw. Actual deviation from this was just a few hundredths of a volt though. All the time, the variable supply indicated it was hitting its target voltage (not sagging or something). The phone pulled around 450 mA off, or if the phone was on and doing something, closer to the 1 A limit that I configured.

Next, I tried with three different USB "brick" AC to DC supplies, including one that's a fancy wall outlet with a built-in USB port:

Code:

PIN5    PIN1

old ipod brick, labeled 500 mA
4.3 V   4.1 V     (phone on)
4.4 V   4.3 V     (phone off)

fancy wall outlet, unknown characteristics
4.8 V   4.7 V     (phone on)
4.9 V   4.7 V     (phone off)

newer generic brick labeled 1 A
5.0 V   4.8 V     (phone on)
5.1 V   4.9 V     (phone off)

And we still know that without a power supply, we have:

Code:

PIN5 = 0.0 V
PIN1 = Vb

Some things to notice:

1. If a "charging brick" is connected to the USB C port, the voltages at both PIN5 and PIN1 seem to be directly determined by the voltage of the USB C port supply.
   
2. If a "charging brick" is not connected to the USB C port, the voltage at PIN1 is the battery voltage.
   
3. If the phone is powered on, my "hacked together" variable supply (fake "brick") doesn't affect PIN1's voltage. PIN1 remains at the battery's voltage.
   

On (3): That's pretty strange, no? There must be something else I need to do to convince the phone that I'm a charger. But this also makes me see it your way, doubting if PIN1/DCIN is connected to the USB C port like the schematics say.

In any case, it's pretty clear to me that the wiki's statement "the second power pin is VBAT, which connects to the battery voltage" is inaccurate, or at best is an incomplete description of PIN1.

If i should power the single board, i would design a 4.0V Power supply and attach it to the battery pins!

Unfortunately, an actual battery is actually necessary for the short bursts of high current draw from the modem, apparently on the order of Amps. So your solution to that problem would work to power part of the system, but not all. Properly powering an isolated mainboard requires a battery-like device, plus another power source for powering the phone + charging the battery. I don't see a way around it.

Only written words so far, so maybe in the end they found out that the Pogo Pins could not be used for that.

Maybe, but I hope not true! The blog says these have gone into production already. You'd think it works.

If they connect the "real" Qi-Charger to Pin5 (wiki name VBUS), then this would be the pin to power the boards... obviously

Ahaha, too bad that's the opposite conclusion of how I read the schematics. I feel like we still have no idea which pin would be the charging input or why.

We've definitely made some progress though :-)

[scholbert]

Some things to notice:

1. If a "charging brick" is connected to the USB C port, the voltages at both PIN5 and PIN1 seem to be directly determined by the voltage of the USB C port supply.
   
2. If a "charging brick" is not connected to the USB C port, the voltage at PIN1 is the battery voltage.
   
3. If the phone is powered on, my "hacked together" variable supply (fake "brick") doesn't affect PIN1's voltage. PIN1 remains at the battery's voltage.
   

On (3): That's pretty strange, no? There must be something else I need to do to convince the phone that I'm a charger. But this also makes me see it your way, doubting if PIN1/DCIN is connected to the USB C port like the schematics say.

In any case, it's pretty clear to me that the wiki's statement "the second power pin is VBAT, which connects to the battery voltage" is inaccurate, or at best is an incomplete description of PIN1.

If i should power the single board, i would design a 4.0V Power supply and attach it to the battery pins!

Unfortunately, an actual battery is actually necessary for the short bursts of high current draw from the modem, apparently on the order of Amps. So your solution to that problem would work to power part of the system, but not all. Properly powering an isolated mainboard requires a battery-like device, plus another power source for powering the phone + charging the battery. I don't see a way around it.

Only written words so far, so maybe in the end they found out that the Pogo Pins could not be used for that.

Maybe, but I hope not true! The blog says these have gone into production already. You'd think it works.

If they connect the "real" Qi-Charger to Pin5 (wiki name VBUS), then this would be the pin to power the boards... obviously

Ahaha, too bad that's the opposite conclusion of how I read the schematics. I feel like we still have no idea which pin would be the charging input or why.

We've definitely made some progress though :-)

O.K. try to keep it a little shorter this time.

The PMIC components on modern platforms are complex logic.
Intense reading of the datasheet may be helpful... or compare it with another reference design:
https://files.pine64.org/doc/Pine%20A64%...elease.pdf

Anyway, it would be a solution to pull to phone apart and use a logic tester to find out where the pins are connected to.

Another thing i realized by having a closer look at the schematic again (and a comparison with the Pine A64+ schematic as well).
It seems to me that there are two different DCIN nets with barely the same name... see my screenshots attached.
My next guess is there could be net names like "DCIN" and "DCIN ".
The red circles may point the same nets then... hope you know what i mean.

BTW, i missed U102 (load switch)... and you're completely right this component would block a direct measurement.

So agreed with most of the things you mentioned...

Some last note concerning your comment on the power supply... you're right to point at the current pulses drawn by the modem.
All you would need here, are some additional decoupling caps with a proper capacity (see the EG25 hardware guides as a reference as well).
The caps on the pinephone mainboard (modem section) are barely sufficient in that case.

Anyway... we're not discussing to build up a DCDC supply for the phone... so good point!!

I enjoy this conversation, but to get clearness we need some reverse engineering... or definite information from the Pine64 team!

Regards,
scholbert

[bokomaru]

The PMIC components on modern platforms are complex logic.
Intense reading of the datasheet may be helpful... or compare it with another reference design:
https://files.pine64.org/doc/Pine%20A64%...elease.pdf

Anyway, it would be a solution to pull to phone apart and use a logic tester to find out where the pins are connected to.

I enjoy this conversation, but to get clearness we need some reverse engineering... or definite information from the Pine64 team!

For sure, I agree with all of it.

I did do a little bit of reading of the PMIC datasheet before. I found that ACIN and VBUS are allowed to be two distinct power sources, _or_ they are allowed to be connected together as one single power source. So we definitely could have the two separate DCIN nets, or one common connected one. Either way would make sense from the point of view of the PMIC.

Another thing i realized by having a closer look at the schematic again (and a comparison with the Pine A64+ schematic as well).
It seems to me that there are two different DCIN nets with barely the same name... see my screenshots attached.
My next guess is there could be net names like "DCIN" and "DCIN ".
The red circles may point the same nets then... hope you know what i mean.

Wow, good find!

I did wonder before why there was a separate drawing connecting DCIN to itself (screenshot attached). Now I see that one is "DCIN" and the other is "DCIN ".

Now I'm not sure if they're really supposed to be one net, or actually two isolated nets. But if they _are_ separated, we have an idea where or why. Neat!

[bokomaru]

Well, I took it apart.

After this, my plan is to reassemble the phone and try applying power to pogo pins to confirm. If somebody wants me to test something while I have the mainboard in pieces, speak up now :-)


There are 3 test points at J1300. A pair of them are the USB DP and DM pins. Convert the PinePhone's USB C connector to a USB A port. The USB A's middle two pins, the data lines, have connectivity with these two test points.

The remaining test point T3 at J1300 is DCIN. It doesn't have connectivity with the USB C port because of U102. This test point has connectivity with the bottom-left pogo pin.

Now, ya have to unplug all the connectors and take the mainboard out of the phone case. And you have to remove the big metal shield covering everything up on the back.

You need the "PinePhone mainboard bottom placement v1.1" drawing to identify two capacitors C646 and C645. Once you realize how large they are, and where they are in relation to the eMMC chip, they are easy to identify.

Capacitors C646 and C645 are connected between USB-5V and ground. And it's true, one side of each capacitor has connectivity with ground. The other side has connectivity with the bottom-middle pogo pin.

Unsurprisingly, ground has connectivity with the bottom-right pogo pin.

For extra credit, I checked the top row of pogo pins against the adjacent resistors. R907 has connectivity with the top-left pogo pin. R905 has connectivity with the top-middle pogo pin. R906 has connectivity with the top-right pogo pin.

I don't see the pogo pins labeled on the physical mainboard.

In fact, none of the test points are labeled on the "PinePhone mainboard top placement v1.1" drawing. And there isn't even a drawing for v1.2. Would be nice to have these things.

If I put together the component placement drawing (or the pinout on the wiki) with the schematics, I suppose the pinout goes like this.

Code:

PIN6 PIN2 PIN4
PIN5 PIN1 PIN3

or

Code:

INT    SDA  SCL
USB-5V DCIN GND

Wrong!

I made a bad assumption that the columns would line up. So I had PIN1 and PIN5 mixed up.

If you think about it, that makes sense given the several voltage measurements I made earlier. The bottom-left pin was always 0.2 V below my power supply, no matter what. That's because this pin connects right up to the power supply. The 0.2 V discrepancy is going to be due to resistance, or a voltage drop across U102, or even differences between the meter on my variable supply vs my multimeter.

I'm also renaming from SCL to SCK to match the name in the schematics.

The confirmed pinout is this:

Code:

PIN6 PIN2 PIN4
PIN1 PIN5 PIN3

Code:

INT  SDA    SCK
DCIN USB-5V GND

Why are the pins numbered in this order? (PIN6, PIN2, PIN4, PIN1, PIN5, PIN3) Good question. Probably, the logical names for schematics were picked before the physical layout was done. So they got rearranged into the order that made physical sense.

The wiki seems unclear to me

Confirmed. You could even argue the wiki is _inaccurate_, depending on how you look at it or how picky you want to be.

Wiki:

Code:

INT  SDA  SCL
VBUS VBAT GND

The names "SDA" and "SCL" would be enough to connect up an I2C device. You could guess that the PinePhone is the I2C master and your device is the slave. Whatever.

The names "VBUS" and "VBAT" are far from enough to understand what these pins are for. Ok, VBUS is probably 5 V for USB. But is it an input or an output? Ok, VBAT must go to the positive battery terminal. But oops, nope, it doesn't! So what voltage should you expect to see there?

Wiki:

The VBUS pin is powered by USB and is 5V at up to 500mA.

This description is ambiguous. Which USB? Does 5 V at up to 500 mA come from inside the phone, or are you supposed to supply that to this pin from outside?

This pin is called DCIN in the schematics!

I now have evidence that it would be safe to provide power _to_ this pin as an _input_ to the PinePhone, under one condition: You don't provide power to the USB C connector simultaneously.

It seems like it would also be OK to use power _from_ this pin as an _output_ from the PinePhone, if one of two conditions are true. 1) A 5 V power supply is connected to DCIN at the USB C connector, which means you're pulling power directly from your external USB supply, or 2) the PinePhone is driving DCIN from USB-5V via U1302, which means you're ultimately pulling power from the battery, possibly boosted up to 5 V. I don't know how much current you could draw safely.

Wiki:

The second power pin is VBAT, which connects to the battery voltage.

This pin does _not_ connect directly to the battery. It's actually called USB-5V in the schematics!

This pin is at battery voltage _sometimes_, depending on what power sources are connected, and the operation of the PMIC and external regulators. Sometimes, this pin is at 5 V.

It seems like it would be OK to use power _from_ this pin as an _output_ from the PinePhone. I don't know how much current you could draw safely.

Wiki:

The pogo pins provide access to an interrupt line, power input to charge the battery, power source and an I2C interface.

Ok, but which pin is the power input to charge the battery, and which is the power source?

My findings are that you could use _either_ of PIN1 (bottom-left) or PIN5 (bottom-middle) as a power output from the PinePhone, and you could use _only_ PIN1 (bottom-left) as a power input to the PinePhone. But only if certain conditions are met, which I already listed. And how much current you could draw or provide is unclear.

I found that ACIN and VBUS are allowed to be two distinct power sources, _or_ they are allowed to be connected together as one single power source.

ACIN and VBUS are tied together.

Now I'm not sure if they're really supposed to be one net, or actually two isolated nets.

There is only one, connected DCIN net.

I noticed something else. Without the charger plugged in to the USB Type-C port. So if you use PIN1/DCIN as an output from the PinePhone. Power is going from the battery into the PMIC, and then out of the PMIC's ACIN and VBUS pins? The AXP803 data sheet says ACIN and VBUS are type "PI" (I think Power Input). Is that drawing power "backwards" through the PMIC?

Not true, not a problem. I reached this conclusion when it seemed that battery voltage was appearing on PIN1/DCIN.

Now we know that battery voltage appears on PIN5/USB-5V, _not_ PIN1/DCIN.

The power from the battery ends up at PS, which goes through L606, through D600, to PIN5/USB-5V. That's fine.