Rick's b.log - 2016/06/19
It is the 3rd of May 2024 You are 18.223.106.100, pleased to meet you!

Power

Some people rate this device as "poor" because it breaks chargers and loses its power within a week. Well, you will notice in the photo that only three LEDs are lit. I charged it in February and left it on my desk, so it can hold a charge. As for breaking chargers - yes, this is quite believable. Many USB chargers are rated for a certain output power. My phone's charger is rated 700mA. My iPad charger is rated 1A. This is important because, unlike a USB port on a computer that can limit the output current, many chargers (especially the cheaper ones) contain no current limiting circuitry within. The battery, while charging, will try to draw as much power as it can. When plugged into my iPad charger, it'll draw 1.21A which is the most the charger is capable of providing (plus it gets pretty hot). When plugged into my Pi power source, the consumption rises to 1.58A (and might have gone higher if I unplugged the Pi). If you are driving a power supply over its rated outputs, you really shouldn't be at all surprised if it overheats and subsequently fails. The power brick says on the side in tiny letters that the input is 5V/2A. Plugging it into anything less capable is asking for trouble.
At the moment, the brick is plugged into one of the USB ports on the computer. It is providing 0.41A and that's it. However with current limiting, this is a safe way to charge that won't blow anything up. It will just take an eternity to charge.

One should be aware, however, that the 10,000mAh is truthful, but it is also misleading. You see, the internal battery is rated 10,400mAh, but that is at 3.7V, the natural voltage of the battery pack. If, for the sake of calculation, we assume that the DC-DC converter inside (translates between the battery's 3.7V and USB's 5V) is flawless and 100% efficient, then a quick calculation will tell us that the power capacity at 5V USB is closer to 7,700mAh. This means it can power a load of one amp (1000mA) for a duration of 7.7 hours. Half an amp for 15.4h, two amps for 3.85h. You get the idea.

So the question that comes to mind is - could this run a Pi, and if so, for how long? Well obviously it can run a Pi. It cannot act as a "UPS" to kick in when the power fails, because power bricks like this tend to only contain the one DC-DC converter. This means it can either be charging, or it can be outputting current to connected devices. It isn't physically capable of doing both at the same time. Which is a shame, but then this device is not intended as a power backup in that sense. Maybe in time, with the growing number of USB powered ARM boards, we may see UPSs that run at 5V native (instead of 12V like security camera backup supplies, or 230VAC like traditional UPSs) become commercialised. For now, it is mostly a do-it-yourself job.

The next part of the question is for how long could it run a Pi? For this we need to know the power consumption of a Pi.
As it happens, I picked up this nifty little thing off Amazon (also on eBay) for about three euros. Took a while to come (from China), but it appears to work quite nicely. Here is the Pi:

So looking at this, the Pi has been on for 58h55m and has consumed 31731mAh. 58h55m is 3,535 minutes. If we divide the consumption by the time, this tells us that the Pi consumes approximately 9mA per minute. If the battery can provide 7,700mAh then this should give us about 855 minutes of use. That's a bit over fourteen hours when running from the battery.

Obviously this is looking at the Pi model B (second revision) when used as an embedded device. It has a mouse, keypad, OLED, and USB serial port attached (via a tiny hub) and is running RISC OS. Different, plus more or less, equipment will alter the consumption - obviously.

In my instance, I have two other USB related devices running alongside the Pi. The first is the Vonets VAP11G WiFi adaptor, which consumes 0.43A. Almost as much as the Pi, but it contains an entire computer within, just like the Pi. The only difference is that it is inaccessible (you can't mess with it).
The other device is the HDMI to VGA converter, which is not so much a computer as a dedicated frame buffer capable of on-the-fly translation of HDMI video (via a framebuffer) to VGA analogue video. Plus extracting HDMI audio to analogue, and translating the monitor's capabilities information into an EDID data block. This, strangely enough, also consumes 0.43A - which is why powering it from the HDMI port (the spec says that is only supposed to provide 0.05A!).

Given this, it can be interesting to plug in other devices to see what their consumption is...

• Freecom 500GB USB harddisc - up to 0.6A when starting, averages around 0.44A when copying files to it.
• USB multi card reader with a µSD card in use - a barely measurable 0.01A, rising to 0.04A when reading and 0.07A when writing.
• iPad Mini - 0.78A when charging, 0.24A when charged.
• Samsung S5 Mini - 0.90A when charging.
• Pentax 16mpx digital camera - 0.45A when charging.
• Samsung external DVD writer - about 0.35A when disc mounted, around 0.48A when playing a DVD. I did notice it spike at 0.9A and it probably uses a bit when writing, which is why it has a Y lead to claim power from two USB ports.
• Wi-Life N-300Mbps WiFi dongle - 0.11A.

Logging into Arcade the easy way

void main(void)
{
   // force terminal to be visible
   setterminal(ANSI);
   
   // connect (assumes current driver is telnet)
   kprints("arcade-bbs.net\n");
   
   // wait (up to 10s) for login screen
   if ( getprompt("(NEW for new user):", 1000) == 1)
   {
      // offer user number (replace XXX with your user number)
      kprints("XXX\r\n");
      pause(100);
      // offer password (replace YYYYYY with your password)
      kprints("YYYYYY\r\n");
      pause(100);
      // skip presentation
      kprints(" ");
      pause(100);
      // skip information
      kprints(" ");
      pause(100);
      // skip today in history
      kprints(" ");
      pause(100);
      // skip lifeline
      kprints(" ");
      pause(100);
      // exit out of mailbox (CR to exit)
      kprints("\r");
      pause(100);
      // main menu -> public messaging menu
      kprints("P");
      pause(100);
      // -> zip through new messages
      kprints("Z");
   }
}

When you want to log into Arcade and check for new public messages, just double-click on the script and let it take you there...

Your comments:

Patrick M, 21st June 2016, 00:25

Thanks for reminding me that I have an external phone battery sitting in a drawer that needs attention, hopefully it isn't dead from being discharged for too long. I don't think I've touched it since mid/late last year.