Solar Power Manager Support

1

Hi Guys,

I recently purchased a solar power manager, specifically: https://core-electronics.com.au/sunflower-solar-power-manager-9v-12v-18v.html

I also purchased a 20,000mAH lipo battery to go along with it, specifically: https://core-electronics.com.au/3-pin-lipo-battery-for-pijuice-20000mah.html

I also purchased a 20W solar panel: Solar Panel 20W-12V Mono 440x350x25mm series 4a SPM040201200

I have a few questions I was hoping you could answer based on my application.

  1. When using the 20W solar panel above with the solar power manager, do I set the MPPT set ‘bit’ to 12v?
  2. The lipo battery I purchased is 3pin rather than 2 pin, so when cutting the wires I technically only have positive and negative terminal blocks to plug the battery into. From my understanding the yellow wire is designed for monitoring temperature, is there any way I can still use this functionality?
  3. If there is no way to use this functionality, is there anything inbuilt in the 20,000mAH LiPo that will detect overheating? From my understanding there is only under and overcharge detection.

My application is that I want the power to run 24/7 to allow my Pi Zero to run from the 5V 1.5A USB output or terminal block output. I also just always want it to charge the battery when there is sun.

With the heat sink installed do you believe I could get 1 year constant usage out of this?

Thanks heaps

1 Like
2

Hi Samuil,

I dont have any experience using the Sunflower board but from my reading the documentation the MPPT to 12V as that is the nominal panel voltage.
Certainly, the sunflower itself doesnt have an input but you could use a microcontroller to monitor the temperature and respond appropriately.
Unfortunately not, the PCB’s on the batteries only just stop it from over and undercharging at the very extreme. If you take a look at the PCM datasheet it is rated at 4.3 and 2.5V respectively which is very far off the desired voltage of a LiPo and will significantly/catastrophically decrease its life.
A whole BMS would be needed to ensure that it is charged 100% correctly taking into account thermal effects.

2 Likes
3

Hi Liam,

Thank you for the detailed response.

The solar panel states 12V however it states in the specification that the maximum PV is 18.5V. Will the solar charge control act as a DC-DC converter stepping down the voltage to 12?

Will it need to be a microcontroller to monitor the temperate or is there a temperature sensor that could do this instead? I am assuming a Pico or Aruino could do the job though?

I have tried researching BMS for LiPo however have not had much luck. Would you be able to point me in the right direction for this?

Once again I really appreciate your help.

Kind regards,
Sam

1 Like
4

Hi Sam,

I’ve had a look at the Sunflower board and the panel you have chosen to pair it with. There is more detailed information available here:

image
image968×649 31.9 KB

Your solar panel has an operating voltage range of 0 - 22.6 Volts and maximum power output at 18.5 Volts. I suspect it is listed with 12 Volts in the name for marketing reasons to identify it as portable and caravan friendly instead of a rooftop solar kind of product.

The sunflower module is basically just a buck converter with some clever add-ons and it can change its impedance to suit the input device, so the MPPT setting should be set to 18 Volts.

This sunflower unit does include a BMS, there is a similar DFRobot unit that does not so Liam may have mistaken it for that model.
The battery protection circuit integrated into the LiPo battery (which Liam referred to ) is usually enough to stop them from being drained dead flat to an un-rechargable state, but the under-voltage protection is set quite low so the longevity of your battery’s lifecycle is not well protected.

The sunflower board is not designed to work with a 3pin LiPo which includes an NTC (for temperature monitoring), so it will monitor the battery voltage, but cannot monitor the battery temperature.
The LiPo does include an NTC for monitoring temperature externally, so the integrated protection board does not have a temperature cutoff, for that you would need an external device like a microcontroller to monitor the temperature.

1 Like
5

I am going down a similar path to the OP, and this topic title is ideal for my observations, so …
I am not asking for specific help here, but trying to present an overview of my conclusion … and partly venting my frustration.

TL;DR … my conclusion is that the current range of “Solar Power Managers” (with the notable exception of PiJuice which works with a Raspberry Pi) are ok only for trivial applications like charging a powerbank, powering garden or christmas lights, or providing short-term mains power UPS function - where the power can suddenly stop without warning when the battery runs out.

The designers & marketing folks assume optimal solar power and battery in order to ignore the real world. I note that in the past month DFRobot have removed the example of powering an Arduino from their product pages and wiki, leaving only a simple solar battery charger, and a mains powered UPS examples.

My use case

I am a 64-year-old former computer programmer - not an electronics engineer - so looking for plug and play components, and something that will allow the computer to shut down before the power dies.

I am wanting to setup an ESP32 with several sensors in my small greenhouse which will require powering from a solar panel & battery. I purchased a DFRobot Solar Power Manager 5V (DFR0559) to use with my 6600mAh 3.7V Li-Po battery, and added a 20W 5V Solar panel (which being Aliexpress, was actually a 4W model). Should be all I need to do the job, yes ? At least enough for a first pass and I will tweak as necessary … or so I thought.

Since I already want to monitor various greenhouse sensors, I would also like to monitor the battery, and solar panel output. Am I getting enough sunlight and power from the solar panel to keep the battery topped up ? Has the battery life run out and it’s time to replace it ? These need a history to see the trends. The Solar Power Manager board is the logical device to collect these data as a side-effect of managing the power.

I want my system to run remotely - ie without a human there watching it - and ideally for 3 rainy days before I need to top up the battery. Is it reasonable for the microcontroller to let me know, and do a controlled shut-down before the battery dies ? Apparently not. A couple of the Solar Power Managers include battery indicator LEDs (which requires a human to be standing there watching it), but NONE have any reporting capability built in.

Am I being unreasonable ? Am I really the only person who wants info on the power being managed ? I note that aspects of this topic are raised regularly in HA and Adafruit forums.

What about the commercial applications mentioned in the product descriptions ? Are they happy for their devices to just power off without warning ? Do they go for overkill on the battery and solar panels … but surely they don’t want to throw that sort of money around unnecessarily ?

To achieve remote solar operation

I can purchase additional products (eg INA219 or INA3221) to monitor voltages and report via i2c bus to the microcontroller, as per this diagram (which has since been removed from DFRobot’s DFR0580 page).

image
image900×740 68.1 KB

So with an automation on the microcontroller I can now be informed remotely if the battery voltage is low, hopefully giving me enough time to top up the main battery before it dies.

And what should the microcontroller do about these over- or under- voltage conditions ? Turn off the input or the load … so now I am adding relays ?

Also we keep hearing about fires caused by lithium batteries; and as the OP pointed out, many Li-Po batteries (currently all Core’s ones over 6000mAh) come with a 3-wire connector … which won’t fit the 2-wire connectors on any of Core’s current Solar Power Manager boards. We are told to cut the 3-wire connector off, connect the +ve and -ve to the Solar Power Manager, and either ignore the thermistor (to invalidate the whole purpose for having the thermistor), or extend the third wire to our microcontroller.

Even if you have a large battery (>=10,000mAh), and the largest 5V solar panel you can find (at least 20W) … sooner or later a run of rainy weather will run the battery flat.

And when the sun comes out and the battery starts recharging ? Your microcontroller will start to reboot and quickly drain the battery as quickly as it can charge. Apparently we also need something like a KA75330 to stop the microcontroller from attempting to reboot until the battery reaches 3.3V.

So, why buy a Solar Power Manager ?

I started my project thinking to use a Raspberry Pi and PiJuice (which does provide battery information to the Raspberry Pi; so is not the focus of my tirade here) but decided the Raspberry Pi would be overkill processing power, and the overhead would chew though the battery quicker. FWIW at that time, I got the impression that PiJuice was intended as a UPS due to the limited battery capacity; but I see now much higher capacity batteries available.

The BMS to control multiple battery cells for a house or EV needs to be much more sophisticated than for my maker project.

But surely the title “Solar Power Manager” implies that the boards do some managing;

Solutions ?

Curiously the ESPHome documentation for INA3221 component provides a link to switchdoc.com whose SunAir product did this back in 2016 (with a successful kickstarter), and at 2020 they had 5 products a user could select from. But with covid and parts shortages they exited the maker market and moved on to weather monitoring and hydroponics.

I note that Adafruit have several solar/LiPo charger boards, but unlike DFRobot and Waveshare they have avoided overselling them as power Managers. If we cannot find any existing power manager board maybe they would consider the market for one ? Certainly LiPo and solar have moved from fringe products in 2016 to almost mainstream now; and I expect there are plenty like me who would rather pay a little extra to get a true power manager so they can focus on their actual application.

1 Like
6

Hey @Donald23173,

It seems like you have found a gap in the available products from these suppliers. I would encourage you to reach out to them with this information if you haven’t already.

As you have broken down in your post, it seems like the additional components needed to adequately ‘manage’ a solar/battery power solution increase the complexity of the module by a significant margin. I would guess that the additional complexity and (I’m guessing here) the larger cost associated with producing these kind of boards has led most of these companies to offer dumber, lower-cost solutions.

Making your interest in this kind of product known will hopefully lead to something that meets your requirements being manufactured in the future.