Page 2 of 2

Re: battery in mah needed for running rpi3 for an hour..?

Posted: Thu Dec 08, 2016 1:22 pm
by Heater
So, this all prompted me to do an experiment with my Pi 3 and 12000mAh (@3.7v) power bank.

Fully charged the power bank. Wrote a little script that imposed a 40% CPU load on the Pi and logged, load and uptime to a file every minute. Meanwhile the on board WIFI is in use. Not much traffic there just keeping in touch with dataplicity.com

Result: The thing ran for 23 hours! Perhaps a bit more as it died after I left for work today. I can check that log file when I get back.

Conclusion: Rounding up to 24 hours it looks like a power bank of 12000 / 24 = 500mAh would run a lightly loaded Pi 3 for one hour. Call it 1000mAh for safety margin.

All very unscientific of course. I should really get down to measuring these things properly. But given the variability is power banks and unknown loads on the Pi it's a good ball park figure.

Aside: An odd thing happened during that test. The ssh session over WIFI died after some hours. Could not ssh back in. Could not even ping the Pi. BUT I could still talk to it via the dataplicity service! Surprisingly ping and ssh resumed working after a few more hours.

Re: battery in mah needed for running rpi3 for an hour..?

Posted: Mon Dec 12, 2016 6:56 pm
by tpylkko
davidcoton wrote:
tpylkko wrote: Yes, and this was already stated in like the second post, but it still does not answer the question, nor frankly does it have much to do with it.
Well, that is a bit blinkered.

The original question was answered (as it turns out, inaccurately, due to caffeine deficiency -- but this was corrected).
However, to experienced Pi users it appears that the OP may be asking the wrong question. It is by no means clear that ananthu0000 is running a Pi with a total consumption of 2A. Therefore it is relevant and helpful to explore what the actual scenario is, and then to explain the misunderstandings that led to a wrong question. It is the nature of this forum (which supports an educational project) to explore wider issues that are of educational value or general interest -- often posters are taught how to answer their own question, rather than just given an answer without imparting any understanding of why it is the answer.

If you think an answer is off-topic, you should report it to the mods, not make comments which can inflame things. And you could always give your own concise and correct answer to the OP.
Yeah, well, the answer is: It depends on the load. I have run a pi a+ on a 2000 mah battery for 12 hours. On 2 A it would maybe last an hour, dpending on how quickly the voltage drops with the current. But, yeah sure, why not mention that you cannot expect to get a good estimation of how long the battery will last if you don't know how much current you are using. It didn't occur to me that somebody might think that their project is using the amount of current that their power source is rated at.

Re: battery in mah needed for running rpi3 for an hour..?

Posted: Mon Dec 12, 2016 9:18 pm
by Heater
tpylkko,
It didn't occur to me that somebody might think that their project is using the amount of current that their power source is rated at.
There is the thing.

One can never assume what somebody posting here knows or does not know.

They read the suggestions for good Pi power supplies, they read the label on their wall warts. Why would they not assume that is what is actually required?

Don't forget they stopped teaching simple physics and mathematics in schools 20 or 30 years ago :)

Having said that I have still to find a USB wall wart that will stop a Pi 3 from throttling down or producing wrong results under exteme load. Damn thing seems to need a tank battery :)

Re: battery in mah needed for running rpi3 for an hour..?

Posted: Mon Dec 12, 2016 9:54 pm
by lumsdot
Ive just put together an esp8266 with an oled screen, to make a voltmeter to read the voltage of my 18650s which power my portable pi.

The Pi3 takes less than 500ma,
Its the 10inch lcd and driver which hogs the current at over an amp

Re: battery in mah needed for running rpi3 for an hour..?

Posted: Tue Dec 13, 2016 12:48 am
by mosespi
Unless their field is specifically electronics, I find a lot of people (yes even techie people) will make the mistake of assuming the label tells them the average power draw. As some of you know.. unless it's a tea kettle or space heater, that number is generally the peak power draw.

I hear similar misconceptions about voltage, and why I prefer the term 'nominal voltage'.

They are both easy, but usually wrong, assumptions to make.

Regards,
-Moses

Re: battery in mah needed for running rpi3 for an hour..?

Posted: Tue Dec 13, 2016 9:43 am
by davidcoton
mosespi wrote:people (yes even techie people) will make the mistake of assuming the label tells them the average power draw.
As you say, on most appliances the label gives the peak power (or sometimes current) draw. In fact most PSUs have that info (wrt the input) too. But there is no way that the output of a PSU can be labelled with anything except the potential maximum power that can be delivered -- the actual consumption depends on what is connected.

And as you also say, the maximum power delivery on some PSUs will incur a voltage drop, especially if there is high current through not-too-thick wires. Bench supplies often have separate monitoring terminals so that the voltage can be regulated at the point of delivery, but I have not yet seen that in consumer supplies -- presumably the alternative of a higher but less regulated supply voltage and an on-board regulator is regarded overall as a "better" solution.

Re: battery in mAh needed for running rpi3 for an hour..?

Posted: Tue Dec 13, 2016 10:44 am
by tpylkko
Well, in my defense, I would still say that while It is true that people might think as is suggested in the above posts, it is just as true that most people (educated or not) will think that unless you measure something in reality you don't know what it is in reality. Furthermore, the internet is full of pages on how much the different models of RPi consume, so you could easily think that people absolutely have to realize that it alone cannot ever consume anything near 2 amperes. Also, most stores that sell Rpi's also sell PSU's with different ranges, so again you would think that anyone looking at these products would realize. Yet even furthermore, the OP was littered with technical units and technical stuff making it appear as if the poster was in the know and often here people get bent out of shape if you suggest something "simple" to them... but yeah.

Re: battery in mAh needed for running rpi3 for an hour..?

Posted: Tue Dec 13, 2016 7:36 pm
by davidcoton
tpylkko wrote: ... but yeah.
I think this thread is quite a good case study in how a question gets asked and answered. Your points all have some validity, but you will find there is more to a good answer than a simple and direct one. I hope you will take up the challenge of trying to answer questions. Expect to get it wrong sometimes (most of us do), but someone will correct you and both you and the questioner will learn. And of course, sometimes there is no correct answer, just opinion.

In this case there was a wrong answer (caffeine deficiency), followed by a technically correct answer and a suggestion that there was a flaw in the question. This led the OP to experiment to determine the truth. Not all posts are that interesting!

Re: battery in mAh needed for running rpi3 for an hour..?

Posted: Wed Dec 14, 2016 10:05 am
by tpylkko
davidcoton wrote: I think this thread is quite a good case study in how a question gets asked and answered. Your points all have some validity, but you will find there is more to a good answer than a simple and direct one. I hope you will take up the challenge of trying to answer questions. Expect to get it wrong sometimes (most of us do), but someone will correct you and both you and the questioner will learn. And of course, sometimes there is no correct answer, just opinion.

In this case there was a wrong answer (caffeine deficiency), followed by a technically correct answer and a suggestion that there was a flaw in the question. This led the OP to experiment to determine the truth. Not all posts are that interesting!
My opinion is that you are entirely wrong.