Topics

OT: Powering field electronics (in winter?)

Gordon Haverland
 

Someone wrote a blurb on setting up a RPi for being powered by solar,
and they were using PiJuice and talking about a solar panel as big as
40W. The company in question I believe is in the UK, so it is about as
far north as we are, we just get colder.

For Dawson Creek, sunrise is 9:28 and sunset is 4:29 on the winter
solstice. On my farm 1 mile south of Dawson Creek, I think sunrise is
something like 10:30 and sunset is about 2:30 (in the NE corner).
Along the south border, it is much less than that. So, Dawson Creek
gets 7:01 of sunshine on the solstice, and I get about 4 hours.

Some of these RPis are 5V machines, and I think the better ones might
be drawing about 1A (max would be 2.5A). I think the article was
talking about 5V output cells, and a 12W solar cell in the best
sunshine should just about be able to support a RPi in bright sunlight
by itself. But this leaves no room for charging batteries to get
through the night.

If the RPi is only using 12W (2.4A), the night use would be 40.8 Ah.
To charge up, we have 7 hours, so we need 5.83A (or about 29W). So
really, a 40W solar panel only does the job is the summer solstice is
bright and sunny all that day (and really, it is still a little small).

An astronomer living in Arizona has an instructable on building a 60W
solar panel. He is building them from "seconds" that are 3x6 inches in
size, and produce about 0.5V each. So, wired in parallel (I've no idea
if this is best) that 60W panel gives 18V.

I think the biggest battery this PiJuice had, was 2500mAh, which isn't
even close. A quick look at amazon.ca, showed at that an 18650 could
be twice this (5000mAh) as Li-ion. I think LiFePO4 is probably better
suited, but that is a 3.2V cell and not 3.7.


And this note went out of control. So deleted much. Maybe the above
is somewhat useful.


If I never got cloudy days in winter, especially around the solstice, I
think I need slightly over 40Ah of capacity in a lithium type battery
(I am specifically thinking LiFePO4).

The specifications say that charging should be done with the battery
warmer than 0C, and that discharging can happen down to -25C.

My naive way of reading this, says I need to gently warm the batteries
before I charge them. Which probably means the battery container is
thermally insulated. Which is typically a problem.

An alternative reading, is that a person can charge them at lower
temperatures, but what you want to do is to really restrict charging
current until the batteries get as warm as 0C, and then you can move
towards a "normal" charging schedule.

Is that a better way to regard charging temperature? Are there battery
management systems that can do this, or is that extra circuitry that is
needed?

--

Gord

Frederick R. McDougall
 

Gordon
This is a topic I have pondered over the years.  Usually lead acid batteries are the best, except for weight, in our arctic climate.  Thermal containers which also house the electronics work very well because of the heat generated by the discharging battery and the electronics (also the charging battery).  But you possibly will need a cooling ventilator system for hot days.
A question I have wondered about but never researched is the heat generated by the solar panels.  How much is it and can it be captured?  If I ever build a house I will do some research on this (or scoff somebodies research).

Good luck on your project
Fred

On Tue., Dec. 3, 2019, 17:41 Gordon Haverland, <ghaverla@...> wrote:
Someone wrote a blurb on setting up a RPi for being powered by solar,
and they were using PiJuice and talking about a solar panel as big as
40W.  The company in question I believe is in the UK, so it is about as
far north as we are, we just get colder.

For Dawson Creek, sunrise is 9:28 and sunset is 4:29 on the winter
solstice.  On my farm 1 mile south of Dawson Creek, I think sunrise is
something like 10:30 and sunset is about 2:30 (in the NE corner).
Along the south border, it is much less than that.  So, Dawson Creek
gets 7:01 of sunshine on the solstice, and I get about 4 hours.

Some of these RPis are 5V machines, and I think the better ones might
be drawing about 1A (max would be 2.5A).  I think the article was
talking about 5V output cells, and a 12W solar cell in the best
sunshine should just about be able to support a RPi in bright sunlight
by itself.  But this leaves no room for charging batteries to get
through the night.

If the RPi is only using 12W (2.4A), the night use would be 40.8 Ah.
To charge up, we have 7 hours, so we need 5.83A (or about 29W).  So
really, a 40W solar panel only does the job is the summer solstice is
bright and sunny all that day (and really, it is still a little small).

An astronomer living in Arizona has an instructable on building a 60W
solar panel.  He is building them from "seconds" that are 3x6 inches in
size, and produce about 0.5V each.  So, wired in parallel (I've no idea
if this is best) that 60W panel gives 18V.

I think the biggest battery this PiJuice had, was 2500mAh, which isn't
even close.  A quick look at amazon.ca, showed at that an 18650 could
be twice this (5000mAh) as Li-ion.  I think LiFePO4 is probably better
suited, but that is a 3.2V cell and not 3.7.


And this note went out of control.  So deleted much.  Maybe the above
is somewhat useful.


If I never got cloudy days in winter, especially around the solstice, I
think I need slightly over 40Ah of capacity in a lithium type battery
(I am specifically thinking LiFePO4).

The specifications say that charging should be done with the battery
warmer than 0C, and that discharging can happen down to -25C.

My naive way of reading this, says I need to gently warm the batteries
before I charge them.  Which probably means the battery container is
thermally insulated.  Which is typically a problem.

An alternative reading, is that a person can charge them at lower
temperatures, but what you want to do is to really restrict charging
current until the batteries get as warm as 0C, and then you can move
towards a "normal" charging schedule.

Is that a better way to regard charging temperature?  Are there battery
management systems that can do this, or is that extra circuitry that is
needed?

--

Gord




Gordon Haverland
 

On Tue, 3 Dec 2019 21:45:00 -0700
"Frederick R. McDougall" <frmcdo@...> wrote:

A question I have wondered about but never researched is the heat
generated by the solar panels.
To generate the electron/hole pair takes a defined amount of energy.,
If the incoming quanta which produced the electron/hole pair has a
higher energy than that, the excess energy can be imparted to speed of
the "two particles" generated, or I suspect some of it may try to move
the lattice in general. The speed of the two particles can easily be
considered to eventually degrade into heat. Lattice recoil not so much.

--

Gord

Gordon Haverland
 

BatteryUniversity and StackExchange had the generic answers. Don't
charge lithium batteries below 0C. Doesn't matter if they are Li-ion
or Li-polymer, the damage that happens can make the batteries more
prone to serious problems later on.

I guess in 2017 or 2018, someone found a way to change the anode, so
that lithium doesn't "plate" out below 0C, and apparently can be
charged at -50C. But, I suspect those are much higher priced
batteries. I suspect the LiFePO4 batteries also fit into the (no
charging below 0C).

Which means, that for winter service, you need to include insulation
and heating for lithium battery solar.

--

Gord

Gordon Haverland
 

Greetings.

People have considered "arctic" conditions, but I am not seeing plans.

I did some back of the envelope stuff, and came up with a battery need
of just over 40 Ah for the winter solstice (if it is sunny). And it
kind of looks like you really want to be running batteries that are
about 20Ah in size. So, the smallest multiple of 20Ah that is more
than 40, is 60. I don't know if anyone makes a 3S BMS, in which case
you might want to go to a 4S which means having 80Ah of battery.

The battery container has some insulation and some thermal mass, and
maybe other stuff. We start our "day" at sundown. For my farm, it is
20 hours until I get direct sun (maybe). While the system is drawing
1A or so, it is generating some heat, so the idea is to have the
thermal mass and insulation help hold the battery at warmer
temperatures as long as it can. At some point, someone needs to do a
calculation as to how much heat for how long is needed to warm the
batteries up, so that they can start charging. I will guess 1 hour.
So at 19 hours after sundown, the heater kicks in to start warming the
batteries up. During our 4 hours of sunshine, we need to store however
many Ah the computer used in 20 hours plus however many Ah were used to
warm the batteries. And we need to produce at least 1A more than that
in order to keep running the computer during the day.

This may be something where tracking is useful, as I think for Dawson
Creek and those 4 hours, I only have to vary the angle by about 40
degrees horizontally, and less than that vertically.

--

Gord

Gordon Haverland
 

On Tue, 3 Dec 2019 21:45:00 -0700
"Frederick R. McDougall" <frmcdo@...> wrote:

A question I have wondered about but never researched is the heat
generated by the solar panels. How much is it and can it be
captured? If I ever build a house I will do some research on this
(or scoff somebodies research).
SolaCity (Kempville, Ontario - near Ottawa) seems to have some useful
documentation (Learning). They say that a solar panel can reach 60C.

I don't think trying to capture this into electricity is going to be
easy. Thermoelectric isn't that great a system. But, if a person also
has solar thermal (say for hot water), it may be possible to have that
water circuit first go past the solar cells, to pick up a little heat
before they go to their main heating thing?

--

Gord