Why two power ports on a budget laptop??

[wdt]

Well, I don't know if you have a DVM.......
Just out of curiosity, I found a bee-hive nail to fit in center
My auto switch did not like it at all,, 1V +-0.5
The dial meter said 5.35,,, FWIW
Obviously what really matters is reading under load,
but my pbp is a virgin, back has never been off

[moonwalkers]

Well, I don't know if you have a DVM.......
Just out of curiosity, I found a bee-hive nail to fit in center
My auto switch did not like it at all,, 1V +-0.5
The dial meter said 5.35,,, FWIW
Obviously what really matters is reading under load,
but my pbp is a virgin, back has never been off

I assume by DVM you mean digital volt-meter? If yes then yes, I do have a multi-meter. I'm not sure what you mean by bee-hive nail, and I'm not sure which readings you're talking about. I did remove the back on mine once to install 128GB eMMC instead of 64GB one.

[wdt]

Down the center of the plug from the brick, you could unfold a paper clip...
It is too small to fit probe into
Maybe your brick is weak??

[dsimic]

Did you try to run the CPU and GPU at full tilt and charge an empty battery while using a 90 W charger disguised as a 170 W unit?  I'm pretty sure that wouldn't work well.

I forgot the exact way the ThinkPad chargers identify themselves to ThinkPads, so this was a good reminder.  Here's detailed information, if anyone wishes to check it out.  However, sensing a resistor is also a way of communication between the charger and laptop; sure, not nearly as complicated as the communication required for establishing high-power USB charging, for example.

Haven't tried that specifically on W510 with 90W PSU - unless I load the GPU it just doesn't breach the PSU limit. But, as I already said, I have tried that on the much older R50 that AFAIK doesn't have that resistor and third pin in PSU for wattage sensing, and the result was that it works, but the battery discharges, though slower than without any PSU. I'm not an electrical engineer, but my understanding is if more current is drawn than PSU can supply without voltage dropping below certain limit then battery will be discharging instead of charging, providing the extra current needed for the machine to keep operating. If this is a misconception on my part I would love to see someone correct it. Until someone does - I'm pretty sure there is more to it than simply overestimating the power consumption. R50 on 30W PSU without battery - definitely unstable, unable to reach full performance. But plug the battery in and it chugs along just fine, with battery discharge rate that appears to be several times slower than running on battery alone.

The sad thing about PBP is that the internal charging circuitry isn't maxed out at around 3 A.  It can go up to 4 A, but it is simply limited to 3 A through the configuration of the BQ24171 charger IC.  By the way, guess what, the charger IC is configured using resistors.

Yes, yes, yes, I read the thread about it, no need to be so pedantic :-) Whatever are the details of the cause, the bottom line - overall power input to the machine is limited by the current battery can use to charge and the machine will not run without battery (the switch on the mobo aside). The whole power system in PBP is (artificially, though probably unintentionally) bottlenecked on the battery charge current, whereas on the same ThinkPads the bottleneck is the PSU itself - if it is not powerful enough the battery will not charge at full speed and may even discharge to supplement PSU, and if PSU has excess power the battery will not charge faster than is safe. But one more observation - it seems to me that unlike the already mentioned R50 on PBP the battery is either charging (however slowly), or just discharging with charger making no difference whether it is connected or not - while I couldn't get any meaningful data from powertop the time measurements on similar loads show little difference between run times with no AC connected and AC connected with red light next to charger port blinking. If this is also a misconception - again, by all means please correct me.

My apologies for the delayed response.

You're right on both of the points above.  If the battery charging current (or the max. current available from the charger) is greater than the current required by the computer, the battery may be charged.  Otherwise, the battery will be drained to compensate the difference.  The underrating of the charger outputs and overestimation of the system power consumption is what actually makes such "charger overclocking" physically possible.

As another example, while debugging some issues with the PinePhone, I've seen a 5 V, 1 A "plain" USB charger supplying 1.2 A happily for extended periods of time.  That's another example of the "charger overclcking".

The trouble with the PineBook Pro is that the battery charger IC is configured to think that the computer needs no power to operate, so the entire input power from the charger is capped at the max. battery charging power, while it could go higher.  All of that goes back to the above-described behavior.

However, I've switched from the factory-installed Debian to Manjaro ARM on my PineBook Pro, and it seems that one of the charging issues was caused by some weirdness inside the Debian's Rockchip-provided Linux kernel, which caused much higher power consumption.  The laptop is actually now able to stay away from discharging the battery while being subjected to pretty much the same load as it was the case with the factory-installed Debian, which used to cause the battery to become completely drained in a few hours, despite the laptop being connected to the 5 V, 3 A charger.

I'll provide more details about Manjaro in my another thread.

---

I think more obvious answers could be;
* They thought 5v @3amp would be enough
* Pinebook Pro is their first USB-C charging device, so they wanted a backup plan for power in case it did not work well, (or at all), see above.

Unfortunately, the PBP's Type-C charging option doesn't seem to have been directed toward a backup plan.  According to the PBP circuitry, the maximum current allowed to be taken from the Type-C port is even lower than the maximum current allowed to be taken from the barrel connector.  More precisely, it's 2.52 A vs. 2.98 A, both at 5 V of course.

Sorry, I meant that the barrel type jack was the backup plan to the USB-C charging.

As for the amperage limitation, I don't know. Could have been a mistake, (though unlikely).

One thing I have not yet read in this thread, (could have missed it), is that over-clocking the RK3399 SoC's CPUs, (and potentially the GPUs), is now "standard" for our Pinebook Pros. Remember, this SoC is not intended as a high performance device when compared to x64 laptop CPUs. But, a majority of people wanted their Pinebook Pros to be faster. (The Pinebook Pro IS already faster without over-clocking than the original Pinebook laptops.) So they accept the excess heat, power draw, reduced battery cycle, etc...

There are people requesting an even faster, (and potentially a more power hungry), SoC for either an upgrade. Or next version. (Some want a 15.5" or 17" screen too, or more internal peripherals like a 2.5" SATA disk bay... which would also take more power.)

I do agree that something can be improved with the power charging & delivery systems.
On the other hand, I think the Pinebook Pro is suitable as is for many tasks. Perhaps not as a daily laptop for people who need more power.

My apologies for the delayed response.

Please, don't get me wrong, I'm perfectly fine with the performance bracket that the PineBook Pro belongs to.  I do not expect wonders, and I would never try to overclock it.

However, I expect the laptop not to discharge its battery down to 0% and turn itself off while being connected to the factory-supplied charger.  Such behavior is simply not acceptable; however, that's exactly what I've experienced with some light usage, i.e., low CPU load, WiFi off, etc.

[CampGareth]

Hmm.  That's a new definition of simpler

All's I knows is that in five decades I can think of exactly two barrel jacks that have failed due to severe mechanical stresses.  I can think of barrel plug power supplies that failed because the cord got chewed by an animal, or the transformer played fuse because there wasn't one (this was pre made-in-China, so no fires).  But I have to replace a device every few years due to USB socket failure, and a power supply every few months due to USB plug failure.  Mind you I only have one USB-C jack and only two USB-C plugs, and I haven't even used them yet.  I think USB-C might be improved over all the other variants I have to use.  But history screams otherwise, and I fully expect that by the time I've had to give in and start really using USB-C, USB-E will be the new thing.

Yes, you can plug in a USB-C the wrong way.  I learnt this by reading the technical specifications.

I think the point is that while barrel jacks are immensely reliable they're not consumer friendly as they fail the principle of least surprise. Pick an arbitrary barrel jack. What voltage and current does it supply? Will it fit in your device? Will it be safe to plug in? No idea to all of the above.

USB-C on the other hand I know it will plug into my USB C device and I know it'll power it without causing any damage. It might not charge my laptop, I need a special charger for that but there are keywords to search for. The worst case is it doesn't charge, not 'it explodes'. It's also pleasing that most 20V barrel jack laptops can be charged off a PD supply with a simple trigger adapter (the other way around is a lot harder!). That said I tried a new smaller USB PD charger today that doesn't want to stick in 20V mode without re-negotiating every few seconds so there's room for improvement on the surprise front.

[moonwalkers]

You're right on both of the points above.  If the battery charging current (or the max. current available from the charger) is greater than the current required by the computer, the battery may be charged.  Otherwise, the battery will be drained to compensate the difference.  The underrating of the charger outputs and overestimation of the system power consumption is what actually makes such "charger overclocking" physically possible.

No offense, but frankly speaking you're not making much sense here - going back to my pretty tired by now example of R50 with a 30W PSU, it matters very little whether the PSU is capable of outputing 36W when it's rated for 30W if the machine expects 72W.

What does matter, however, is that even the simplest PSUs like the one I have in my Soviet EleKtronika MK-52 programmable calculator from the early 80s usually have voltage regulating and current limiting circuits that protect both the device and the PSU itself by providing stable voltage up until certain load, at which point the voltage starts to drop. And when voltage drops the device can no longer draw any more power from the PSU and will have to supplement it from a battery, if one is available, regardless of the excess tolerance on the PSU.

The trouble with the PineBook Pro is that the battery charger IC is configured to think that the computer needs no power to operate, so the entire input power from the charger is capped at the max. battery charging power, while it could go higher.  All of that goes back to the above-described behavior.

That's one way to put it.

However, I've switched from the factory-installed Debian to Manjaro ARM on my PineBook Pro, and it seems that one of the charging issues was caused by some weirdness inside the Debian's Rockchip-provided Linux kernel, which caused much higher power consumption.  The laptop is actually now able to stay away from discharging the battery while being subjected to pretty much the same load as it was the case with the factory-installed Debian, which used to cause the battery to become completely drained in a few hours, despite the laptop being connected to the 5 V, 3 A charger.

I used the Rockchip-provided kernel only for as long as it took me to prepare a microSD card with (nearly) stock Debian using the unofficial "installer" script by danielt@, which is barely a couple hours after I opened the DHL package. And even then, soon afterwards and all through now I've been using the Manjaro kernel with my Debian installation. So whatever charging weirdness I've been experiencing is definitely _not_ due to Rockchip kernel. I've been keeping all my firmware blobs up-to-date, BTW.

However, I expect the laptop not to discharge its battery down to 0% and turn itself off while being connected to the factory-supplied charger.  Such behavior is simply not acceptable; however, that's exactly what I've experienced with some light usage, i.e., low CPU load, WiFi off, etc.

100% with you - even if the stock PSU doesn't support full load without discharging, the design should at least allow for such option to exist as an aftermarket purchase.

---

I think the point is that while barrel jacks are immensely reliable they're not consumer friendly as they fail the principle of least surprise. Pick an arbitrary barrel jack. What voltage and current does it supply? Will it fit in your device? Will it be safe to plug in? No idea to all of the above.

USB-C on the other hand I know it will plug into my USB C device and I know it'll power it without causing any damage. It might not charge my laptop, I need a special charger for that but there are keywords to search for. The worst case is it doesn't charge, not 'it explodes'. It's also pleasing that most 20V barrel jack laptops can be charged off a PD supply with a simple trigger adapter (the other way around is a lot harder!). That said I tried a new smaller USB PD charger today that doesn't want to stick in 20V mode without re-negotiating every few seconds so there's room for improvement on the surprise front.

Not to rain on anyone's parade--I gladly take advantage of the USB charging PBP allows--but only to explain why USB charging is neither the panacea nor the silver bullet of charging it's been treated as....

There is a difference between simplicity and POLA (principle of least astonishment), and traditional non-USB power supplies are definitely simpler.

And even when it comes to POLA... When I grab a barrel plug PSU it has markings for the voltage, current, and polarity it provides. I can't remember seeing even one that doesn't. And virtually all devices that use barrel plug PSUs have markings next to the socket that tell you the voltage and polarity they need, except for maybe the hobbyist devices and the likes that don't go through proper mass-market certifications. Make sure the voltage and polarity match and you can actually plug the cord without it being loose - and the worst case scenario the PSU doesn't provide enough current, the same exact situation as with your described USB C deal.

If you know how much current your device needs you can tell right away whether any random PSU will be able to provide that - it's written on the PSU itself. That said, while barrel plug PSUs come with the cable integrated, when you pick up any random USB C cable you cannot be sure whether it will let you use the full power the available PSU can provide, and as far as I'm concerned that goes right against the very POLA you brought up, leaving you wondering why your clever 100W USB PD PSU doesn't charge your shiny new USB PD laptop--a situation I've ran into myself a couple of times with the employer-provided machines.

[KC9UDX]

Hmm.  That's a new definition of simpler

All's I knows is that in five decades I can think of exactly two barrel jacks that have failed due to severe mechanical stresses.  I can think of barrel plug power supplies that failed because the cord got chewed by an animal, or the transformer played fuse because there wasn't one (this was pre made-in-China, so no fires).  But I have to replace a device every few years due to USB socket failure, and a power supply every few months due to USB plug failure.  Mind you I only have one USB-C jack and only two USB-C plugs, and I haven't even used them yet.  I think USB-C might be improved over all the other variants I have to use.  But history screams otherwise, and I fully expect that by the time I've had to give in and start really using USB-C, USB-E will be the new thing.

Yes, you can plug in a USB-C the wrong way.  I learnt this by reading the technical specifications.

I think the point is that while barrel jacks are immensely reliable they're not consumer friendly as they fail the principle of least surprise. Pick an arbitrary barrel jack. What voltage and current does it supply? Will it fit in your device? Will it be safe to plug in? No idea to all of the above.

USB-C on the other hand I know it will plug into my USB C device and I know it'll power it without causing any damage. It might not charge my laptop, I need a special charger for that but there are keywords to search for. The worst case is it doesn't charge, not 'it explodes'. It's also pleasing that most 20V barrel jack laptops can be charged off a PD supply with a simple trigger adapter (the other way around is a lot harder!). That said I tried a new smaller USB PD charger today that doesn't want to stick in 20V mode without re-negotiating every few seconds so there's room for improvement on the surprise front.

I think that it boils down to carelessness and complacency over time.

Forty years ago just about every consumer device used coaxial power connectors that we now call barrel plugs/jacks.  Everyone knew that the voltage, current, and polarity of ever device was different, so everyone made sure to use the right one.

I do have some devices from the made-in-China era that have barrel type wall warts and neither the device nor the wall wart give any indication of electrical parameters. I find this troubling.

[xyzzy]

I have seen my PBP go from 60% charged using the Pine supplied charger to 100% charged
While watching YouTube videos with the volume turned up to the Maximum.

I see my battery drop by about 9% to compile the manjaro kernel package, with screen at 75%, using the pine supplied charger.  No NVMe drive, which would be a considerable extra power draw.

If you are able to use hardware accel, then I don't think youtube video is really that much of a power draw.

PSU in the PBP seems under powered to me.  They could easily get 36W with USB-PD if the regulator could go up to 12V and the rest of circuitry could handle the extra power.

[dsimic]

You're right on both of the points above.  If the battery charging current (or the max. current available from the charger) is greater than the current required by the computer, the battery may be charged.  Otherwise, the battery will be drained to compensate the difference.  The underrating of the charger outputs and overestimation of the system power consumption is what actually makes such "charger overclocking" physically possible.

No offense, but frankly speaking you're not making much sense here - going back to my pretty tired by now example of R50 with a 30W PSU, it matters very little whether the PSU is capable of outputing 36W when it's rated for 30W if the machine expects 72W.

I'd say that we're basically conveying the same thing all the time, but for some strange reason we're unable to acknowledge that.

A laptop that is declared to consume 72 W will actually use 72 W only under 100% system load, plus 100% screen brightness, etc.  Otherwise, it would consume much less power, which allows it to be powered most of the time by a 30 or 36 W power supply.

However, I've switched from the factory-installed Debian to Manjaro ARM on my PineBook Pro, and it seems that one of the charging issues was caused by some weirdness inside the Debian's Rockchip-provided Linux kernel, which caused much higher power consumption.  The laptop is actually now able to stay away from discharging the battery while being subjected to pretty much the same load as it was the case with the factory-installed Debian, which used to cause the battery to become completely drained in a few hours, despite the laptop being connected to the 5 V, 3 A charger.

I used the Rockchip-provided kernel only for as long as it took me to prepare a microSD card with (nearly) stock Debian using the unofficial "installer" script by danielt@, which is barely a couple hours after I opened the DHL package. And even then, soon afterwards and all through now I've been using the Manjaro kernel with my Debian installation. So whatever charging weirdness I've been experiencing is definitely _not_ due to Rockchip kernel. I've been keeping all my firmware blobs up-to-date, BTW.

It would be very interesting to perform a thorough investigation, because my own experience is rather different.  I can see significant improvements to the power consumption from running the Linux kernel shipped with Manjaro.

---

I see my battery drop by about 9% to compile the manjaro kernel package, with screen at 75%, using the pine supplied charger.  No NVMe drive, which would be a considerable extra power draw.

I saw my PineBook Pro battery going from about 95% down to 60% as a result of running "make -j6" inside a Linux kernel tree for just about half a minute.  That was truly disappointing.

PSU in the PBP seems under powered to me.  They could easily get 36W with USB-PD if the regulator could go up to 12V and the rest of circuitry could handle the extra power.

Let's hope that the hardware modifications required to use a higher voltage over USB PD will be actually possible.