Page 1 of 1

DIY PiJuice?

Posted: Mon Mar 09, 2015 10:40 am
by bugsy_malone 666
Ok so I have been looking at ways to put a project I built last year back into operation with some enhanced performance. I built a 24/7 CCTV recorder using the raspicam project as my base, run from a large leisure battery with 17.5w of solar panel, you'd imagine it would run lovely. Not when your in the UK where theres a maximum of 8 hours a day light let alone sun in the winter. Once we had enough sun for my system to run for almost 2 weeks without the need for a battery change, but on average about 6 days was the most you could get. It would record 24 hours a day to the SD card, which made them last about 3 months with the fact the pi would just suddenly lose power.

Having seen alot of development in recent times for better power management (just moving from an A to an A+ would make a difference) I was looking to re-implement my project with a few changes, one of which being my power source. I have started harvesting lithium laptop cells, which will obviously give better performance than regular cells and was looking to build some sort of mega pack, with a phone these things will last a day maybe 2 or 3 on Standby, so I was thinking with several cells, a solar panel and some sort of time activated stuff I could maybe make the system 100% better.

I came across PiJuice while looking through search results on batteries and it seems to be the type thing I want, but I want something with slightly more capacity. What I havent really managed to find anywhere though is a DIY version? I have an RTC time clock, lithium chargers and stuff and a massive supply of cells I wanted to see if anyone had the DIY equivalent of a PiJuice to give me some base ideas.

The plan is to have the Pi and its battery pack about 10 feet apart as space is limited for the pi cam in its housing, but I can have a separate battery bank elsewhere.

Suggestions and ideas?

Re: DIY PiJuice?

Posted: Mon Mar 09, 2015 12:39 pm
by rahlquist
Well the whole point of DIY is DIY.. Not DSSED (Do Something Someone Else Did).. That said....

You have batteries and lithium chargers. You have a Pi, what you don't have it sounds like is enough Solar and possibly needed fast storage capacity to provide consistent output. Here is what I would do in your case.

Find the voltage of your solar panel, double (or triple) that value, and build a large capacitor bank at that voltage level. Build this with as much capacity as you can. The reason for this is your power vs light vs ability to use situation. You have a 17.5w panel. I am willing to bet though that your battery cannot charge at that rate, so the chargers lower it to whats not going to destroy your battery, so say the battery is a 3.7v lithium and can charge at up to 1700mA. You are looking at less than 6.3watts needed to charge, where does the rest of that 17.5w go? Out the window. Also cloudy day solar from what I have seen in the past ( and mind you I haven't played with solar in more than a decade), is very irregular, up and down. So on a sunny day while you may get 8 hours charge time the way you are setup now, on a cloudy day you may only get a fraction of that because your panel may only produce more than you need just for charging only a few times. I am rambling will try to tighten it up.

Cloudy day output;
2 hours @ 12w usable
2 hours @ 7w usable
4 hours @ 4w usable

For the sake of argument lets also assume your Pi is drawing 1000mA while running. That puts consumption at 10w with no headroom for conversion/charging losses. That means for only 2 hours a day in the above scenario are you able to charge because there is output left after consumption. All assuming your charging/PS can handle this. Lets add the assumption that your charging/PS is limited to 10w output total. For those 2 hours that means 2w is LOST though potentially available. With the addition of a large capacitor bank before your regulation/output/chargers those time when the panel is outputting more than you can consume, still gets stored, in those large capacitors where it can continue to be output at a later time without a lot of expensive charging like you need for battery tech. They can also absorb energy much faster. The key here would again be how much capacitor capacity can you provide. This capacity would never be a waste because worst case scenario their energy would continue to output their stored charge even after its gotten dark enough for the panel to no longer produce.

Some more info can be found at ... s-and.html
and ... /?ALLSTEPS

Re: DIY PiJuice?

Posted: Mon Mar 09, 2015 2:00 pm
by bugsy_malone 666
The capacitor thing is interesting, when I get time I'll look into that. I actually think power wise I'll have more batteries to charge than the capacity to charge them, as you suggest with excessive watts not doing anything you want to make use of them too by storing the energy for a latter point.

The reason i am interested to see what other people have made is for ideas, often i'll see an idea and instantly see a way I can improve on it or take ideas from the design to build something new. Sometimes its nice to try another persons idea and see how it might need to be altered to your particular application, been finding that with Arduino, where people have built weather stations but when I try to build what they have it doesnt work and I end up scratching my head!

So its nice to start with an idea to evolve. I certainly think the super capacitor idea as an application, even if the batteries charge in a day it can run the pi for a bit or something until that power runs out.

Re: DIY PiJuice?

Posted: Mon Mar 09, 2015 2:12 pm
by pluggy
To keep a Pi running all the time from solar in the UK, even in January, I reckon you'd realistically need 200W of solar panels and some bl**dy big batteries. Year round, you'll get approximately 10% of the plate capacity, reckon a quarter of that to get it through December - February, and less still to get it through a week of bad weather in January (when in all likelihood they'll produce nothing at all). If you had batteries that would run the Pi for a month without solar connected at all, it would probably manage. Of course you need some facility to stop the solar frying everything in July.....

Re: DIY PiJuice?

Posted: Tue Mar 10, 2015 1:56 pm
by bugsy_malone 666
Well when I say keep running, to develop the idea and make use of the new batteries is why I was looking at pijuice as its kinda designed to do a few things to keep the Pi in line when theres not enough power.

Running a CCTV 24/7 is a bit of a mission, but having timed and PIR activated stuff would greatly save on storage and also battery power, its the battery part I am looking into primarily.