A bunch of solar usb battery packs

[xevilstar1]

I happen to have a bunch of solar usb battery packs laying around....
I wander if is it possible to connect them all to a pi so that they all give the same 5v to the pi without trying to discharge one into the other....
how would you do that ?



Is it possible to have an onscreen battery indicator that will tell me how much energy I have real-time ?

[drgeoff]

I happen to have a bunch of solar usb battery packs laying around....
I wander if is it possible to connect them all to a pi so that they all give the same 5v to the pi without trying to discharge one into the other....
how would you do that ?

Not impossible but IMHO impractical.

[pidd]

Not really practical, they shouldn't discharge into each other, but the same one will always take the hammering first - the one with the highest output voltage. Then they will discharge in order one at a time.

In practice there will be some sharing of load but the principal remains that one will discharge first, possibly before some of the others start supplying any significant current.

Depending how close their voltages are, they could be levelled out with small resistors but that would be throwing energy away, the closer the voltages the more efficient that would be.

There is also a risk of oscillation where one starts to provide current but its output drops under load and then another takes over etc ad infinitum.

[drgeoff]

I'd think about connecting them in series and using a buck converter to convert the resulting higher voltage down to 5 volts. My concern would be if they aren't well matched in terms of capacity and fullness what (bad?) things happen when one or more start running out?

[xevilstar1]

what if they are all equal ?
how would I make the onscreen indicator ?

[davidcoton]

... what (bad?) things happen when one or more start running out?

Potentially (pun intended) reversed polarity. That's in Something Very Bad™ territory. But the discharge protection might kick in first, if it exists.

what if they are all equal ?

They won't be.

how would I make the onscreen indicator ?

With considerable difficulty.

[W. H. Heydt]

what if they are all equal ?
how would I make the onscreen indicator ?

This gets into the difference between Physics and Engineering. *In*theory*, they are all identical. In practice, they won't and can't be.

[xevilstar1]

is out there a pcb with screw terminals or usb ports I can use to switch seamlessly from one battery to the other ?

[Imperf3kt]

You're a step ahead.
Before trying to make something out of what you have, you first need to learn what you need for this to work.

Calculate your required wattage to run the Pi for your desired time, calculate how much energy each panel outputs and then see if you what you have will supply what you need.


Pwrsonally, I'd not even bother but if I was going to do it I'd remove the batteries and just use the solar panels, rewired to a different battery pack / setup instead of trying to balance all the different batteries.

[xevilstar1]

How can I know how much energy my pi rig draws in 24 hours ?

[W. H. Heydt]

How can I know how much energy my pi rig draws in 24 hours ?

5v time average current times the time interval. If you want to be (engineering) conservative, use the nominal max current spec of 3A. Doing that, you are drawing 15W or 15Wh/hr. 15Wh/hr * 24 hr = 360Wh.

[salvato]

How can I know how much energy my pi rig draws in 24 hours ?

5v time average current times the time interval. If you want to be (engineering) conservative, use the nominal max current spec of 3A. Doing that, you are drawing 15W or 15Wh/hr. 15Wh/hr * 24 hr = 360Wh.

360W to be pedantic...

[DougieLawson]

How can I know how much energy my pi rig draws in 24 hours ?

5v time average current times the time interval. If you want to be (engineering) conservative, use the nominal max current spec of 3A. Doing that, you are drawing 15W or 15Wh/hr. 15Wh/hr * 24 hr = 360Wh.

You'd have to be running flat out for that.

I was looking at my energy monitor the other night (while charging my EV). The grid was supplying 7.8KW. The car was charging at 7.5KW. Everything else running in the house at the time (that's freezers, fridge, raspberries, lights, TV, laptop & phone chargers) was using 300W.

[W. H. Heydt]

How can I know how much energy my pi rig draws in 24 hours ?

5v time average current times the time interval. If you want to be (engineering) conservative, use the nominal max current spec of 3A. Doing that, you are drawing 15W or 15Wh/hr. 15Wh/hr * 24 hr = 360Wh.

360W to be pedantic...

Nope. Do the dimensional analysis.

[W. H. Heydt]

How can I know how much energy my pi rig draws in 24 hours ?

5v time average current times the time interval. If you want to be (engineering) conservative, use the nominal max current spec of 3A. Doing that, you are drawing 15W or 15Wh/hr. 15Wh/hr * 24 hr = 360Wh.

You'd have to be running flat out for that.

I was looking at my energy monitor the other night (while charging my EV). The grid was supplying 7.8KW. The car was charging at 7.5KW. Everything else running in the house at the time (that's freezers, fridge, raspberries, lights, TV, laptop & phone chargers) was using 300W.

IT also ignores conversion losses and battery inefficiencies. So, taking a conservative approach, I'd still use the 360Wh value.

[salvato]

5v time average current times the time interval. If you want to be (engineering) conservative, use the nominal max current spec of 3A. Doing that, you are drawing 15W or 15Wh/hr. 15Wh/hr * 24 hr = 360Wh.

360W to be pedantic...

Nope. Do the dimensional analysis.

Sorry. You are right!

[xevilstar1]

is there a commecially available lipo that can suit my needs ?

[W. H. Heydt]

On the whole issue.... Is it necessary to use a Pi4B? There are a number of different Pis that could be used that would lower the power requirements to make a battery powered solution much easier.

Do you need 4 USB ports? If so, do you need the sort of computational power that comes with a Pi4B? If not, consider a Pi2B, which would cut your fully loaded power requirements by one third to 2A.

If you don't need a full array of USB ports, and if you don't need to computational power of a Pi4B, then the Pi3A+ or even A+ might be a better board to use. Even a Pi3A+ would cut power requirement significantly and an A+ should run on less than 1A, probably less than 500mA. In the extreme, you could look at the Pi0 or Pi0W and cut your power budget to around 200mA (1W).

[xevilstar1]

my portable rig has many things attached it is intended for on the go experimenting with the pi and I am thinking about a jacket or something around my arm haven't decided yet. got also two 7 inches touchscreens that give me a good nice desktop view.

by the way I found this https://nevonprojects.com/multi-power-s ... t-sources/
do you think something like this could be safely adapted to my needs ?

[Paul Hutch]

5v time average current times the time interval. If you want to be (engineering) conservative, use the nominal max current spec of 3A. Doing that, you are drawing 15W or 15Wh/hr. 15Wh/hr * 24 hr = 360Wh.

The units you used are a not quite correct causing some of the confusion, at least for me. Watt hours per hour is not appropriate here, that's a unit for change in power consumption over time.

To keep the units clear I would write it as:
15W or 15W/h, 15W * 24 hr = 360Wh

[dickon]

To keep the units clear I would write it as:
15W or 15W/h, 15W * 24 hr = 360Wh

Dimensionally, that's nonsense. 15W == 15J/s, 15W/h == (15/3600)J/(s^2). The two are not the same.

[pidd]

5v time average current times the time interval. If you want to be (engineering) conservative, use the nominal max current spec of 3A. Doing that, you are drawing 15W or 15Wh/hr. 15Wh/hr * 24 hr = 360Wh.

The units you used are a not quite correct causing some of the confusion, at least for me. Watt hours per hour is not appropriate here, that's a unit for change in power consumption over time.

To keep the units clear I would write it as:
15W or 15W/h, 15W * 24 hr = 360Wh

@Heydt's statment of "15W or 15Wh/hr. 15Wh/hr * 24 hr = 360Wh" is absolutely correct other than the mix of "h" and "hr" and it even makes sense if read out aloud. Yours is incorrect but I love trying to picture what a "W/h" is and whether it has any use.

[Paul Hutch]

Dimensionally, that's nonsense. 15W == 15J/s, 15W/h == (15/3600)J/(s^2). The two are not the same.

@Heydt's statment of "15W or 15Wh/hr. 15Wh/hr * 24 hr = 360Wh" is absolutely correct other than the mix of "h" and "hr" and it even makes sense if read out aloud. Yours is incorrect but I love trying to picture what a "W/h" is and whether it has any use.

Thanks guys and sorry W. H. Heydt.