Solar with RPi Zero v1.3

[Mickety]

So I've ordered a solar panel that is rated 6volts and 1watt for 1$ USD.
Wikipedia says Raspberry Pi Zero v1.2 and 1.3 has a power rating of 100 mA if you don't use monitor. keyboard and mouse.
Tha'ts good, because I only use a WiFi USB dongle thingy and boot in CLI .
Given the numbers the solar (cell I guess since it's small) unit should produce around 167 mA.
Well good. I could use 100 of those to power the board and the remaining 67 mA could be stored in a battery when the sun goes down.
I was thinking about using nickel metal hydride (NiMH) batteries because they don't need charging circuits and don't have a memory effect.

So does this mean I just put solar panel and a NiMH battery and RPiZero board in parallel with a voltage regulator (or a buck converter) somewhere in the circuit or is it a lot trickier than that?

Thanks!

[drgeoff]

I was thinking about using nickel metal hydride (NiMH) batteries because they don't need charging circuits

Where did you learn that bunkum?

Admittedly not as critical as rechargeable lithiums but you cannot just connect NiMH cells across a solar panel.

[rpdom]

Tha'ts good, because I only use a WiFi USB dongle thingy and boot in CLI .

WiFi uses a lot of power. By using that you can add at least another 100mA or more (probably a lot more) to your calculations.

[Mickety]

I was thinking about using nickel metal hydride (NiMH) batteries because they don't need charging circuits

Where did you learn that bunkum?

Admittedly not as critical as rechargeable lithiums but you cannot just connect NiMH cells across a solar panel.

I don't know, I thought only Lithium batteries needed special circuitry or they'll explode (LiPo) or lose capacitance (LiFePO4). How difficult is it to make one for NiMH? What about NiCd batteries? They too require circuitry to charge?

Tha'ts good, because I only use a WiFi USB dongle thingy and boot in CLI .

WiFi uses a lot of power. By using that you can add at least another 100mA or more (probably a lot more) to your calculations.

Well that's dissapointing. *sigh* I guess I could order an another solar cell in this case.

[tpylkko]

https://blog.adafruit.com/2014/06/20/ho ... pberry_pi/

[mfa298]

So I've ordered a solar panel that is rated 6volts and 1watt for 1$ USD.

Those are based on the absolute maximum the panels can manage.

Even in the glorious sunshine the UK has had over the last couple of weeks the couple of panels I've got (10w and 40w) probably only manage to provide about half their rated maximum, then when it's overcast you might only get a tenth of that maximum.

[pfletch101]

So I've ordered a solar panel that is rated 6volts and 1watt for 1$ USD.

Those are based on the absolute maximum the panels can manage.

Even in the glorious sunshine the UK has had over the last couple of weeks the couple of panels I've got (10w and 40w) probably only manage to provide about half their rated maximum, then when it's overcast you might only get a tenth of that maximum.

Basically correct for power output. The output voltage of a solar panel depends on the current being drawn from it, with its maximum value (for a given temperature and illumination level) being measured only when no current is being drawn. The maximum power point (MPP) of a solar cell typically corresponds with a voltage of 80-90% of its open circuit voltage (OCV). The OCV does not drop as far or as fast as the max power output as illumination is reduced - a cell with an OCV of 6V under standard illumination conditions (equivalent to reasonably clear sunshine) may still generate an OCV over 5V if the illumination drops to 20% of that, and the MPP under these conditions may still be over 4V, but the available power will be less than 20% of its value under standard conditions.

As a side note, you can only get anything approaching the maximum possible power out of a solar cell or array by manipulating the effective resistance across it on the fly as illumination changes to keep it close to its MPP. Inverters used in home (and commercial) PV systems include complex circuitry to do this, and to keep the output (AC) voltage constant despite the consequent variations of input (DC) voltage. The OP will need corresponding (if somewhat less sophisticated) circuitry to maintain a constant supply voltage to his Pi and to charge his battery.

[Mickety]

So I've ordered a solar panel that is rated 6volts and 1watt for 1$ USD.

Those are based on the absolute maximum the panels can manage.

Even in the glorious sunshine the UK has had over the last couple of weeks the couple of panels I've got (10w and 40w) probably only manage to provide about half their rated maximum, then when it's overcast you might only get a tenth of that maximum.

Well where I am the sky is mostly clear and sun is mostly generous with it's light so I though about maybe increasing the ammount of battery packs to be able to store more for the "rainy day" you know.
Now my idea is:

3 solar cells - 6volts 3 watts ~500 mA
If half is the best they can do it would mean 250 mA, 100 for the board and 150 to charge the battery
How many batteries in a pack (or Amp hours) should I buy to make it able to work, given that 8 out of 12 months have sunny days all day long so they work at half efficency (optimal case scenario) do you think? ( Like they charge batteries to full during 8 months and rely on batteries for the remaining time untill the next cycle )

[Mickety]

So I've ordered a solar panel that is rated 6volts and 1watt for 1$ USD.

Those are based on the absolute maximum the panels can manage.

Even in the glorious sunshine the UK has had over the last couple of weeks the couple of panels I've got (10w and 40w) probably only manage to provide about half their rated maximum, then when it's overcast you might only get a tenth of that maximum.

Basically correct for power output. The output voltage of a solar panel depends on the current being drawn from it, with its maximum value (for a given temperature and illumination level) being measured only when no current is being drawn. The maximum power point (MPP) of a solar cell typically corresponds with a voltage of 80-90% of its open circuit voltage (OCV). The OCV does not drop as far or as fast as the max power output as illumination is reduced - a cell with an OCV of 6V under standard illumination conditions (equivalent to reasonably clear sunshine) may still generate an OCV over 5V if the illumination drops to 20% of that, and the MPP under these conditions may still be over 4V, but the available power will be less than 20% of its value under standard conditions.

As a side note, you can only get anything approaching the maximum possible power out of a solar cell or array by manipulating the effective resistance across it on the fly as illumination changes to keep it close to its MPP. Inverters used in home (and commercial) PV systems include complex circuitry to do this, and to keep the output (AC) voltage constant despite the consequent variations of input (DC) voltage. The OP will need corresponding (if somewhat less sophisticated) circuitry to maintain a constant supply voltage to his Pi and to charge his battery.

This sounds kinda bleak for my skills. Have any links about MPP, OCV and drop rates of solar cells in various conditions that you could recommend? I have more of a "try first, think later" approach to building electronic stuff but in this case I'm afraid to even do anything unless I am absolutely positive about the numbers because I don't want to fry my board since a Zero board is difficult to obtain period.

[Mickety]

https://blog.adafruit.com/2014/06/20/ho ... pberry_pi/

Thanks for a link. I now have a starting point.

[Imperf3kt]

Is this for a 24 hour project?
To figure out what size battery you need is kind of complex, but not hard.


Here's an example.
Out of 24 hours, you get roughly 8 hours of useable light in perfect conditions.
That means you need at least 16 hours of battery backup.
If your setup uses 100mAh, then you need to store 1600mAh in the battery, minimum. Realistically, you'd want 2500mAh or more.

If you are charging batteries with 150mAh (which you won't be, due to inefficiencies in converting the voltages, etc) and this happens for 8 hours, then the maximum you can charge the battery would be 8x150 = 1200mAh. Not enough. Do remember that this is for 'perfect' weather, which I highly doubt you ever see.

This, of course, is just to run it for one day.


In my opinion, aim for at least a 20w solar setup for a Pi0

[mosespi]

Some thoughts..

Look here for a real world solar example.. http://www.allspectrum.com/mopower/sola ... solar.html
Convert your calculations in watt and watt hours, makes thing easier when dealing with different voltages, batteries, panels, DC converters, etc.
I generally take 3/4 of the rated solar output of the panel as my maximum starting calculation, MPPT chargers/converters.. means Maximum Power Point Tracking might do a bit better, look it up, but it's more money.
You can do -small- solar direct to NiMH (add a diode for prevent reverse flow), but generally below the trickle charge rate of about C (Capacity) / 20 (some list C/30 some C/40 as the safe rate). Any more and you should use a charge controller as NiMH batteries will not be able to safely absorb the over charge. NiCD does a bit better in this regard. You may be able to go with a -little- bit higher current if your sunlight is not 24/7 and you know you are going to be starting with a partially empty battery every day. Higher power charging really needs a charge controller of some sort. LM317s can be used as an effective current limiter too if you need one.

The simple tried and true method for solar would be a 12v (actually 18v for 12v nominal battery charging) solar panel + appropriate charge controller (small ones are cheap now) + a lead acid battery + a suitable 5v DC-DC converter for the Pi. What happens when power gets low is another matter, and depends how critical your project is and if it can stand an abrupt power cut.

Regards,
-Moses