Solar harvesting into Li-ion battery

This is a 0.8x0.6 inch board with AEM10941 solar harvesting IC charging a Lithium Ion battery

Similar projects worth following
Starting from
jaspersikken has 651 orders / 23reviews
Ships from Netherlands
The AEMLION is a 0.8x0.6 inch board for the AEM10941 Solar Harvesting IC from E-peas. It efficiently converts solar panel energy into Li-ion battery charge, it even works with indoor light. It features 3.3V and 1.8V regulated outputs that are enabled when the battery has useful charge, and a low battery warning that informs the user of impending shutdown when the battery runs low. It easily integrates in other projects because of the castelllated via's, and when soldered onto 0.1' pitch header it fits in a bread board. I am selling this board on Tindie. This project is related to and

I designed it because I took part in the HackadayPrize2018 Power Harvesting challenge with a 1x1 inch Tiny Solar Energy Module and found a lot of interest in selling it. However since most projects have different power demands and people want to choose their own solar panel, I decided to remove the onboard solar cells.

This board is special because it integrates maximum power tracking, Li-ion battery charging and two regulates outputs in a tiny and easy to integrate board. There's is no other board with so little passive components.

Ideal for indoor applications

The AEM10941 harvesting IC is very suitable for indoor applications because it has an ultra low power startup. The boost converter starts at a very low 380 mV input voltage and 3 uW input power. The IC gets most power out of the solar cells by doing MPPT maximum power point tracking every 5 seconds.


  • PCB 2 layers (0.8 x 0.6 inch)
  • solar input voltage 50mV to 5V
  • ultra low power startup 380mV/3uW
  • solar input current max 110mA
  • MPPT every 5 secs, MPPT set to 70% (adjustable)
  • Battery: connect your 3.7V Li-Ion battery
  • 3.3V/80mA and 1.8V/20mA regulated outputs. These are enabled when battery voltage is between 3.60V and 4.12V (max charge voltage)
  • battery reverse voltage protection
  • 3.3V status output pin that warns the host MCU if battery voltage drops below 3.60V

  • AEMLION R3 has adjustable output voltage

    Jasper Sikken11/16/2020 at 20:28 0 comments

    AEMLION R2 has the two LDOs configured to a fixed 3.3V and 1.8V. This works fine for most people. However some people wanted to further reduce power consumption of their application by lowering the output voltage to resp. 2.5V and 1.8V. Below is the table with Battery and LDO configuration pins of the AEM10941 chip. CFG0[1) can be changed from 1 to 0. With this setting the battery can be used down to 3.01V in stead of 3.60V in R2. 

    This change will greatly extend battery life of the application. People asked me how to change the pin. I had to tell them that the CFG[1] is set by a copper trace UNDER the chip, this makes it almost impossible to modify the AEMLION R2 board. 

    So no I have redesigned the PCB. In AEMLION R3 the LDO voltages can be set using a 0R jumper resistor.

    This board is not for sale yet because I still have large stock of R2, I think in few months I will order a new batch of 100 which has the new design.  

  • I compared AEM10941 to BQ25570

    Jasper Sikken04/22/2020 at 13:57 1 comment

    I compared AEM10941 to BQ25570 and it shows that AEM10941 is the best for solar. 

    For thermal I would choose BQ25570 

  • Video for the HackadayPrize final

    Jasper Sikken10/20/2018 at 14:21 0 comments

    For the HackadayPrize 2018 I designed a Tiny Solar Energy Module. It is cool because it can harvest enough energy from indoor light to power simple BLE or LoRa applications. I am selling two boards on Tinde for $24.90. One stores it's energy into a Li-ion battery, EAMLION, and the other stores it into a pair of supercapacitors, AEMSUCA. This is really cool because now you can make applications that runs forever from indoor light without a battery!

    Tiny Solar Energy Module project on

    AEMLION project on

    AEMSUCA project on

    AEMLION is sold on Tindie  for $24.90

    AEMSUCA is sold on Tndie for $24.90

  • 50 boards are assembled and for sale!

    Jasper Sikken10/16/2018 at 18:04 0 comments

    Yesterday I have received 50 assembled boards from Elecrow. I tested them with simple test rig and they all work well. I am very happy with the quality delivered by Elecrow. I started selling on Tindie.

  • 50 PCB assemblies finished

    Jasper Sikken10/14/2018 at 22:21 0 comments

    50 PCB assemblies were finished by Elecrow and shipped to the Netherlands. They arrive tomorrow, but I have no time to test and fulfill Tindie orders soon because few days ago my wife gave birth to a baby boy!

  • Test rig is ready

    Jasper Sikken10/09/2018 at 18:07 0 comments

    I have created a simple test rig to test the oncoming batch of 50 AEMLIONs. It applies 1.5V to the solar input and it has LEDs connected to the 3.3V and 1.8V. This basically tests the circuit. It does not have a bed of needles but it simply two 5 pins rows of vertical male headers that are under 15 degree angle. When you push the board between the pins they make contact with the castellated vias.

    The package with the 50 AEM10941 ICs shipped with PostNL on 6 Sept still has not arrived at assembler Elecrow. It is stuck since Sept 18th at customs. After a short Chinese holiday Elecrow is working with customs again and try to release the package. I expect It wont take long. 

  • 50 PCB assemblies are still not assembled

    Jasper Sikken09/30/2018 at 19:13 0 comments

    I have shipped AEM01941 ICs using PostNL to the assembler Elecrow in China for assembly of 50x AEMLION. I have prepared the Tindie webshop were I will be selling them. However the parcel has not arrived at Elecrow. PostNL track and trace shows on Sept 18th the package was released by customs in China, after that no more updates :-(. Elecrow now is working with customs to get the package. PostNL will not try resolve the problem before Oct 2nd, so that sucks. The package is just sitting there until then... To try out my own webshop I have sold 1 prototype to someone in California and received a lot of interest, at this time there is a waitlist for 14 pieces. 

  • Three designs with the AEM10941

    Jasper Sikken09/23/2018 at 11:34 0 comments

    In May/June I have designed the 1x1 inch Tiny Solar Energy Module to take part in the HackadayPrize2018 Power Harvesting challenge and won! And then I received a lot of interested in selling it. However since most projects have different power demands and people want to choose their own solar panel, I decided to remove the onboard solar cells and make two even tinier designs derived from the TSEM. They are both 0.8x0.6 inch small. The first is the AEMLION which is charging a Li-ion battery and the second is the AEMSUCA which charges a pair of supercapacitors. 

    I will be selling the AEMLION and AEMSUCA on Tindie for $24.90. At the moment ELecrow is working on 50 assemblies of each.

  • How large is the battery charge current?

    Jasper Sikken09/21/2018 at 14:45 0 comments

    I have measured AEMLION battery charge current using three different solar panels in indoor light (500 lux), outdoors in the shadow, and in full sun (~500W/m^2). 

    Solar panelindoor 500 luxoutdoor shadowoutdoor full sun 500W/m^2
    1V/100mA 30x25mm50uA700uA5mA
    2V/100mA 79x28mm100uA1.8mA10mA
    4V/100mA 70x70mm280uA4.8mA20mA

    Indoors (~500 lux) and with the smallest solar cell the battery is charged at 50uA for 10 hours. Then the application must have an average current of (50uA*10hrs/24hrs) 20uA or less. That's enough for a simple Bluetooth Low Energy beacon or a very simple LoRa application. If that is not enough you need to select a larger solar panel. 

    Outdoors in shadow current is 10-20 times larger. And in sun 100 times larger. 

    As you can see the AEM10941 can charge a battery from indoor light but indoor light is really only suitable for very low power applications.

  • I have build a simple test rig

    Jasper Sikken09/11/2018 at 20:40 0 comments

    Recently I ordered 50x PCB assemblies from Elecrow that i want to sell on tindie,  but I need a testrig for it. I quickly made one with prototype board, 0.1" headers, a battery, a capacitor, some leds and resistors.

    I just press the board-under-test between the two rows of angled male headers, and then the solar input gets 1.5V from the battery and the LEDs indicate the 3.3 and 1.8V outputs work. Basic functionality checked!

View all 14 project logs

Enjoy this project?



Mark Jeronimus wrote 05/29/2021 at 14:59 point

Between all the solar harvesters we sampled for work (most of them from China though), this one worked by far the best!

  Are you sure? yes | no

brianlam3288 wrote 04/30/2021 at 05:08 point

HI, can the Vlv be configured to 1.4V? thx alot

  Are you sure? yes | no

Jacques Caron wrote 02/22/2019 at 00:50 point

Hi all,

Very interesting project, but I see the 3.3V output is limited to 80 mA. This may be a bit too small for some chips while transmitting, as they can have higher peak currents.

Is the 80 mA coming from the chip itself, or something else on the board? Any way to get more?


  Are you sure? yes | no

((( TTN_Berlin ))) wrote 07/14/2019 at 13:06 point

interesting remark. 80 mA will not fit for LoRa technologies, when TX... so my question would be... can we expect a more powerful device?



  Are you sure? yes | no

arthur_jordan05 wrote 10/04/2018 at 13:36 point

Nice project! Thanks for making it open source. I ordered some chips from Fujitsu (which was not that easy btw), built one of these and it works quite well! I can think of many applications for the module.

Gonna try the AEMSUCA next. Like the fact that it makes use of the chip's balancing function for supercaps...but one has to find caps that don't have a self-discharge current that's higher than the harvested charge current. I think a bigger solar cell is definitely necessary here.

Have you thought about making a LiFePO4 version? LiFePO4 batteries have a stable chemistry, are less dangerous and have about 2000 charge/discharge cycles ...

  Are you sure? yes | no

Robert Poser wrote 10/15/2018 at 06:35 point

I am using aem10941 in combination with organic supercaps from cap-xx. they have low self discharge rates of only 1µA/cell after some days of stabilisation. If try to use single cell supercaps rather than double cell supercaps since the balancing consumes additional 1..2µA too.

  Are you sure? yes | no

Kurt Van Kelst wrote 08/21/2018 at 20:04 point

where do u buy the chips? ....cant find em online

  Are you sure? yes | no

Boris Bershadsky wrote 09/08/2018 at 12:26 point

Kurt, he hasn't made the tindie product page yet...

Jasper, please expedite! I have a bunch of 5V solar cells begging for these chips :)

  Are you sure? yes | no

Jasper Sikken wrote 09/08/2018 at 12:39 point

Hi Boris, thank for liking and following my project. The chips are not available at the large distributors. Here you can buy them from a few distributors. I bought them directly from E-pease.

The chip can handle up to 5V, please make sure your solar panel open circuit voltage is below that

  Are you sure? yes | no

Jasper Sikken wrote 09/08/2018 at 12:34 point

These chips are not available at the large distributors. I bought them directly from E-peas, but you can also buy them at distributors listed here.

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates