Safe chemistry MPPT solar charger

[ssozonoff]

Hi All,

Excuse me for writing in English but its maybe the most compatible option

Personally I am not happy with sticking Lipo cells out in the nature, all it will take is one fire!! As a result have been developing a multi-chemistry MPPT solar charger with a personal preference for LTO mainly for its cold weather charging abilities. While there are a couple of solutions to be found on the internet to address this need, one seems to be no longer available, one is great but binds you to a RAK module and the aliexpress one is, I would say, incomplete. They are all based on the CN3795 as is this one.

I have produced my own PCB inspired by the great work of Vlastimil Slinták who makes the Solar mesh baseboard but my wish was for a “charger only” PCB.

So far I have made two prototype batches of 5 units in iterations. The first one did not include an onboard regulator, the second one adds an efficient onboard regulator with a solder jumper to select between 3.3V and 5V output and an option to disable the regulator completely. Note that the regulator will shutdown at around 1.2V but requires 1.8V to startup again which is mainly relevant to LTO. Output ripple was measure at 25mV.

Both units have been tested and work as designed. I am now preparing for a larger production run with a few additional changes to the previous two prototypes.

These would come with pre-defined resistor values for LTO, Sodium-ion and LifePO4 1S configurations, you just need to solder the jumper.

Solar panel Vmp voltages by default for 9V, 12V and around 18V panels, 7.5V input voltage is the minimum possible with this charger.

Max charge current is configurable in steps with a max current of 1.2A and other options at 120mAh, 483mAh and 847mAh

For the battery charge voltage and the Vmp voltage (MPPT) there is a place to solder your own through hole resistor to set a custom value for your needs.

I have been testing with a 9V panel of 147mm x 147mm and achieving around 260mAh in ideal conditions using an LTO cell. Example setup

Now I am reaching out to see if there is any interest from the Mesh community so I can size my first production run accordingly and additionally to understand if there are any other ideas I have not covered. I am not attempting to make any money with this but simply trying to cover my costs and a bit of time, the sales price would be roughly 30.00 CHF plus shipping.

Here is a picture of the pre production unit. The final one will not have two battery connectors. i.e. the “Load” connector will be removed but raw battery is still available at the 4 pin header J1.

PCB size is 52mm X 41mm ± 2mm

I do have 2-3 pre-production units immediately available. 10.00 CHF for the one without the regulator and 15.00 CHF with the regulator + shipping cost.

These test units have slightly different defaults for 1S LTO, 2S LTO and 1S Na-ion and 120mAh, 240mAh, 360mAh and 480mAh charging currents but if you tell me your needs I can add resistors to set the battery charge voltage and Vmp.

P.S. I also have a LTO 1S BMS coming in the form factor of an 18650 cell for easy integration with a 2-5P pack of cells. It will expose an I2C interface using a QWIIC connector for battery voltage and power consumption, compatible with an INA260 so it works in Meschore without any changes. Again it is based on the open source work of Vlastimil, its his engineering, I have simply taken a different approach to form factor and re-designed the PCB accordingly. I will have the first prototypes in a week or two.

Thanks,

Serge

# Spec. sheet below

Solar Charger Section

High-efficiency buck switching controller with hardware-configurable presets.

1. MPPT Configuration (Solar Panel Optimization)

*Sets the Maximum Power Point floor *

Jumper	Resistor	Target Voltage	Typical Use Case
JP8	90.9k	9.01V	Standard 9V/12V Panels
JP9	120k	11.35V	High-Voltage 12V Panels
JP10	200k	17.77V	Standard 18V Panels
JP11	DNP	Custom	Through-hole pads for custom panels

2. Charge Voltage Configuration (Chemistry Select)

Sets termination voltage

Jumper	Resistor	Target Voltage	Battery Chemistry
JP4	133k	3.65V	1S LiFePO4
JP5	205k	2.80V	1S Sodium-ion (SIB)
JP6	290k	2.33V	1S Lithium Titanate (LTO)
JP7	DNP	Custom	Through-hole pads for custom voltage

3. Charge Current Selection (1206 Shunt Bank)

Parallel resistor bank for thermal stability (Base 1.0R is always active).

Jumpers Closed	Total Resistance	Charge Current	Mode
None	1.0R	120mA	Safety / Trickle
JP1	0.248R	480mA	Standard Charge
JP2	0.142R	840mA	High-Power
JP1+JP2+JP3	0.099R	1.21A	Turbo Mode

System Output Regulator (TPS63802)

High-efficiency synchronous buck-boost converter providing a stabilized system rail.

1. Operational Limits

• Safe Input Voltage: 1.8V to 5.0V Maximum (Optimized for 1S packs)
• Continuous Output Current: 2.0A
• Peak Switching Current: 4.5A
• Function: Seamlessly transitions between buck and boost modes to maintain regulation even as battery voltage drops.

⠀2. Output Voltage Selection (JP12)

• Jumper Open: 3.3V (Optimized for ESP32, nRF, and low-power MCUs)
• Jumper Closed: 5.0V (Standard USB-level sensors and peripherals)

⠀3. Disable jumper (JP13)

• Cut JP13 to disable the built-in regulator, J1 provides solar input and raw battery output
  ⠀

[zh_fox]

Very interesting idea! I myself just got started with a simple DIY solar repeater with a 6V panel and 2x 18650s.

About charging of lithium-ion batteries in low temperatures, I have come across this very interesting article: Cold Weather Charging of Lithium-Ion Batteries — YYCMesh

It alleges that charging lithium-ion batteries in sub 0°C temperatures is actually not as big of a deal as we have been lead to believe. Due to this I have decided against a temperature switch or other measures, and we’ll see over time if this actually has an effect on my batteries. Because my node is on my balcony far from anything flammable, I’m not too worried.

[ssozonoff]

Hi,

I have seen that document and cannot validate or rebuke those claims since I have not tested it to this extent. The thing is you will be artificially limiting your charge current based on this assumption and usually during the season where you would like push as much energy to your batteries as possible when sun is available (high charge current and potentially short sunlight periods).

It also has nothing to do the the fact that Lipo’s are known to burst into flames for many different reasons.

18650’s are know to be more stable than pancake style Lipo cells so this helps a little, but honestly I dont feel good taking that risk outside my 4 walls.

And yes, I am more referring to repeaters installed out in nature.

[zh_fox]

I think for the most remote repeaters enduring extreme conditions, such as on mountains, batteries like that combined with more sophisticated charging solutions make sense.

For the more “normal” repeaters, I would argue that in winter during the day, the repeater as a whole will heat up quite a bit, thus raising the overall temperature before higher charging currents due to stronger sun occur. This is assuming you’re using some small panel, like 1-5W.

Also while it is true that lithium-polymer and lithium-ion batteries are dangerous and can burst into flames, they don’t do so spontaneously (assuming you have cells from a reputable manufacturer). They do degrade over time, depending on storage condition (temperature, voltage per cell), so it probably makes sense to replace your repeater’s cells after maybe a year. Most dangerous is over discharge and over charging, but I think both of these are very unlikely to happen if you use a BMS and MPTT solar charge board designed for 1s batteries.

Where do you see the use case for your PCB and batteries such as LTO? Do you think all repeaters should be using them? I personally haven’t tried LTO batteries or even had them in my hands. How do they compare to 18650s in terms of cost and availability in Switzerland?

[ssozonoff]

Hi,

a lot of if’s and assumptions

@zh_fox

Where do you see the use case for your PCB and batteries such as LTO? Do you think all repeaters should be using them? I personally haven’t tried LTO batteries or even had them in my hands. How do they compare to 18650s in terms of cost and availability in Switzerland?

Yes many repeaters are already out in the “wilderness”, tree’s, mountains, pastures etc … I know several of the guys around my area are using LTO already as am I.

LTO is only one option but even Sodium-ion or LifePO4 are more stable but they too suffer from below zero charging which is not the case with LTO. Power density is a little lower than with Lipo but its not really an issue for our low power repeaters.

You can also get LTO cells in an 18650 size factor and the largest capacity I have seen is 1500mAh per cell. The ones I am using now are 23680 (23 dia. by 680 long) at 2500mAh per cell.

I have not looked everywhere in Switzerland with regards to availability since I simply order directly from China. Price wise not a huge difference.

Dont get me wrong, I am not here tell people how to build their repeaters, I was just interested in solving a problem I had with the way I wanted to build mine.

Best,

Serge

[TNM]

Hi Serge thanks for taking the time to make such a detailed post.

TLDR: What is better an unknown Chinese LTO battery or a known good quality Li-Ion cell?

I feel also a little shadow in back in the depths of my mind where there could be a possibility of a repeater bursting in flames as all 3 repeaters are in the wilderness over (oftentimes) dry grass/leaves.

What I’m wondering is how solid the available cells are compared to the more established Li Ion or LiPo Cells. I recall a video where x-ray analysis showed huge differences in tolerance and quality between different 18650 Cells.

And I’m just wondering if you know how mature the whole ecosystem of LTO is - I have had 0 exposure to it so far.

[ssozonoff]

Hi,

At the end of the day most of our cells are made in China one way or another. With Lipo cells you typically have different “Grade” cells, paying more for the A grade and going down to grade C which are basically failed cells with significant performance issues but still sold on the market. You often dont know what you are buying other than maybe making an assumption based on the brand or price on what you are getting. Typically Grade A go to reputable cell phone manufacturers who pay the premium.

In any case from a purely chemical/design point of view an unknown Chinese LTO cell will always be more stable then a well known Li-ion cell. Worst case your LTO cell has been sold as a higher capacity cell than it really is.

There are many differences between Lipo/Li-ion and LTO like for example the anode material but possibly the most significant one, relevant to our current discussion, is thermal runaway which is a significant risk with Lipo chemistry.

Tests have shown that LTO cells can be punctured, crushed, or overcharged without catching fire or producing smoke.

One other benefit, in my opinion, is the number of charge cycles. Good LTO batteries can have > 20’000 cycles (in temperatures from -40 to +50C) versus Lipo at maybe 500-1’500 cycles and a lower charge limit of 0C.

Today LTO batteries are mainly used in high intensity commercial and industrial applications, hence the lack of variety in smaller cells suitable for our needs and the lack of chargers and BMS solutions for low power IoT based applications.

Best,

Serge

[ssozonoff]

I will further add that I have also produced very simple pcb’s for assembling a 2p or 3p pack based on the following LTO cells.

[ursmii]

Replacing the LiPo with saver batteries makes only sense if we can reuse the other parts.
So coming into business your charger board should have the same size as them in the most used repeaters e.g. D5 & SenseCap Solar Node.
Building a repeater from scratch is for many of us to complicated and insecure in case of waterproof and stability.

[zh_fox]

These are the exact LTOs you showed in your post before: HAKADI LTO 2.4V 2500mAh 23680 Lithium Titanate Cell 15C Power Recharge – hakadibattery

I just went over the specs and noticed the following:

Operating Temperature:

Charge: 0~45℃

Are you aware of that?

[ssozonoff]

Hi.

Thanks for pointing this out but it is surely a mistake on their side. I have reached out to them directly for clarification.

The characteristic of charging in under sub zero temperatures is a feature of this chemistry and not vendor specific.

[ssozonoff]

@ursmii

Replacing the LiPo with saver batteries makes only sense if we can reuse the other parts.

I understand however as mentioned I am building this in my spare time out of interest in IoT projects. I will not be able to adapt the PCB for multiple cases.

I dont have a D5 but if someone wants to measure the direct dimensions of the charge PCB I can at least give it some thought. I would also need an estimate of the height available under the lid.

[ursmii]

D5: The dimensions of the board are 31 x 74mm
Inner height is about 25mm
here also two pictures

[ssozonoff]

Thank you for this and although I could adapt the PCB to fit I just noticed that the output voltage of the solar panel on the D5 is 6V which sadly will not work with the charger which requires 7.5V or more so its a no go.

[chix]

I have been looking into different battery technologies recently and I have also ordered some Sodium-Ion 18650 cells.

I am now looking for solutions on how to handle the charging, but there are not a lot of good cheap options yet. The best option I have found and ordered (but not yet tested) is this one:

BQ24650 2A 4A 8A 10A Lithium-Batterieladegerät 1S 2S 3S 4S 5S 6S Solar...

I’d love to buy some of your PCBs, but since every single solar node I own is running on a 6 Volt solar cell it is, just like you also mentioned, not compatible.

[ssozonoff]

@chix

I’d love to buy some of your PCBs, but since every single solar node I own is running on a 6 Volt solar cell it is, just like you also mentioned, not compatible.

I realise this (I had the same problem) but unfortunately this is a requirement of the charging IC itself and there are not a lot of options to choose from…

[chix]

I know, this becomes obvious when searching for such PCBs on Aliexpress, there are only few supporting such low voltage panels.

Did you ever check out the BQ24650 chip which the linked board above is using?

[ssozonoff]

Hi,

Yes and it would work for a 2S LTO setup but sadly not for a 1S LTO setup. With some fine tuning, it may work with Sodium-ion…

[jedie]

take a look on Solar Mesh Baseboard by μArt.cz

[ssozonoff]

Yes if you want an “all in one” solution with the RAK module this is the board to get.