Power supply rating?

[TarableCode]

I'm fiddling with a PCDuino3 Nano and the tech specs say it needs 2A to run, but my own small bit of testing shows it using much less.

Is the additional current rating for worst case scenarios or additional USB devices?

 

I made a tiny bash script to monitor /sys/power/axp_pmu/ac/amperage and /sys/power/axp_pmu/ac/voltage but I'm not seeing those figures.

 

Idle with nothing connected except power, 1Gbps ethernet, and a 16GB SD Card:

 

Same as above except with sysbench loading both cores:

 

I still need to get a USB wifi adapter for this, but I can't see that adding more than 100mA to the max which was around ~450mA.

Unfortunately I don't have the hardware on hand to measure the current through the 5v supply.

 

The reason I'm asking this is because I want to use this board as a kind of addon to my Apple iBook G3 and replace the Pi Zero W it's been using.

I've been looking for a suitable way to power this thing from the laptop and that's where I'm kinda stuck.

 

IIRC I can get 9v at 900mA from the firewire port and the USB port is the standard 5v 500mA that came on devices at the time, but I can't verify the max current rating and I also don't want to cook my iBook's USB ports finding out.

Would it be better to get a DC-DC converter and run off the firewire port or use a battery to handle peaks > 500mA and charge it using the USB port?

 

Either way it's been fun fiddling with this board so far and even if I don't use it as a companion to my iBook I'm sure I'll find a place for it.

 

e: Dumb script I made by googling and stuffing things together until it worked

http://paste.debian.net/981643/

[chwe]

6 minutes ago, TarableCode said:

Is the additional current rating for worst case scenarios or additional USB device

Probably not worst case, maybe enough that it runs smoothly for sure. If you suck the maximum current, the voltage drops normally below 5V (for sure if you use a cheap charger, never tested it on my notebook). 

 

this two threads could give you a general insight in power consumption:

Not about A20 in general but maybe it helps get a deeper insight in this topic. Normally these boards suck a lot of power during the boot process, something you might not see with your script. Optimizations can be done to avoid peaks during the boot process (If I'm right kaiser wrote something about it in these threads I mentioned).  If you search for 'USB power meter' on your prefered gimmick supplier, I'm sure you'll find one of these cheap USB thingies where you can see how much current it sucks. Or you go for an INA219 which should also do the job. But as zador mentioned these readings might be not correct cause voltage can pulsate due to overloading.

 

 

 

[zador.blood.stained]

58 minutes ago, TarableCode said:

I'm fiddling with a PCDuino3 Nano and the tech specs say it needs 2A to run, but my own small bit of testing shows it using much less.

Plug a (relatively old) SATA HDD and it alone may consume up to 500mA from the 5V line when it spins up. It also has 2 USB ports which can consume up to 500mA each or even more depending on the current limitation implementation (if it exists). USB OTG port, HDMI port and expansion ports also may provide 5V to connected devices.

 

So 2A is not a hard requirement but a rather conservative value at which you won't run into undervoltage issues straight away if you decide to plug a HDD and some USB devices into the board.

[tkaiser]

2 hours ago, TarableCode said:

Dumb script I made by googling and stuffing things together until it worked

http://paste.debian.net/981643/

 

Just for the record. In Armbian there's no need for such monitoring scripts since you get RPi-Monitor installed by doing a 'sudo armbianmonitor -r'. The appropriate sun7i_next.conf template for A20 boards running mainline kernel can be found here: https://github.com/armbian/build/tree/master/packages/bsp/armbianmonitor/templates

 

And another note wrt PCDuino3 Nano powering: There's a Micro USB jack for DC-IN and an Micro USB OTG port. You can provide power via both but unlike on almost all other A20 boards these two jacks are not separate power inputs (with different semantics eg. providing power through /sys/power/axp_pmu/vbus on almost all other AXP209 equipped boards you need to press a power button for the board to power up while using /sys/power/axp_pmu/ac the boards starts as soon as power is provided). Regardless which of both Micro USB jacks you use it will always be fed to DC-IN (/sys/power/axp_pmu/ac) but PCB traces are so tiny that you experience a severe voltage drop even with light loads when using the OTG port compared to the other Micro USB jack.

[TarableCode]

Would it be safe to assume that without attaching any USB peripherals (aside from WiFi) it should be able to run from the firewire port provided I use a DC-DC converter?

I would plan on hooking it directly into the 5v in header on the top and avoid the USB jack altogether.

 

The board's main function is to provide WiFi to the ethernet connection and run devkitPro's devkitARM toolchain.

In the end it'll stick onto the side of my iBook kind of like what the pi zero w does except through the ethernet jack instead of USB.

 

 

Edited August 16, 2017 by TarableCode
Forgot something

[chwe]

Yeah, years ago since I saw 'barbie's toilet seat' running.  Did you power your RPi zero via USB? It's long time ago but I have in mind that USB 1.1 had something like a handshake before the USB controller delivered more than 100mA to the powerlines?

 

[TarableCode]

Yep, it's powered through the OTG port which also provides an RNDIS ethernet connection.

It works fine it's just I want to go back to MacOS 9 which does not have an RNDIS driver so I'm looking to switch to ethernet.

 

Totally stable, I just like fiddling with things.

I brought it to Starbucks just so I could take a picture of living the hipster stereotype dream, totally worth it and the battery lasted a good 5 hours.

Got lots of work done on it too