I'm currently working on a custom PCB where I have a microcontroller hooked up to 3 LiIon batteries which are meant to get charged by a CN3722 IC. However, when I connect a solar panel to the circuit the batteries are not getting charged. The datasheet of the CN3722 can be found here.

Some more information:

  • All three LiIon batteries have a voltage of 4V so they should be getting charged.
  • One has to specify what the battery regulation voltage is supposed to be. This is done with a resistor divider. The formula is Vreg = 2.416 * (1 + (R22 / R23)), in my case this equals to roughly 12.6V.
  • You also have to specify the MPPT voltage (since its not real MPPT). The formula for this is Vmppt = 1.04 * (1 + (R17 / R18)). I wanted to be able to adjust this in the finished PCB so R17 is an adjustable resistor in my case. At the moment I have set it to 150k Ohm in order to have the MPPT voltage set to roughly 16V.
  • A charging cycle should basically start, if...
    1. Vin rises above the battery voltage and
    2. Vin is no less than the MPPT voltage set by the resistor divider.

My problem is that it doesn't charge. I have measured the current draw from the solar panel and I only measure ~6mA under pretty much full sunlight. This is due to the "CHARGE" LED being lit. However it is not charging. I don't know why though. I designed everything like it is described in the datasheet and tried to take all PCB layout considerations into account.

This is my schematic: (typical application circuit can be found on page 2 in the datasheet)

The Vsolar net is connected to an INA219 IC, which basically just measures the charging current so I can monitor it.

This is the rather small INA219 circuit:

As you can see its simply measuring current/voltage. I don't think this might cause any problems, can it? The VCC net is directly connected to the positive of the battery.

Finally I will attach a few images of my PCB. Even though I don't think I made huge layout mistakes I could miss something. This is the complete PCB, which contains some circuits I haven't mentioned (like a 3.3V buck converter.), because they are not related to this problem.

This is the CN3722 circuit on the PCB:

I know this is a rather extensive question, but I would really appreciate any help.

  • \$\begingroup\$ What's the voltage at the TEMP pin? Switch node and gate drive waveforms would be useful. Also a link to the datasheet for the inductor might help. \$\endgroup\$
    – John D
    Oct 18, 2020 at 19:55
  • \$\begingroup\$ According to your schematics, the CN3722 input and output (Vin and Vsolar) are shorted through 0.1 ohm resistor, so it's useless. \$\endgroup\$ Oct 18, 2020 at 21:10
  • \$\begingroup\$ @JohnD I'm using this inductor: lcsc.com/product-detail/… The voltage at the TEMP pin is ~0.6V at room temperature. Sadly I don't have the equipment to get any waveforms. \$\endgroup\$
    – Gereon99
    Oct 19, 2020 at 15:55
  • 1
    \$\begingroup\$ Your schematic is confusing. Vsolar is where the battery is connected, and the solar panel is connected to the two-pin header? So R6 is between Vsolar and the battery? What are the specs of the solar panel? Also, the recharge threshhold according to the datasheet is 95.8% of Vreg, which, assuming Vreg = 4.2V/cell, is 4.02V/cell. This means odds are the charging doesn't start when you try batteries that sit at 4V. Not sure why the charge light goes on then, though. \$\endgroup\$
    – ocrdu
    Oct 19, 2020 at 16:01
  • 1
    \$\begingroup\$ Without a scope it would be hard. With a scope and a diff probe you could capture the current through the sense resistor to see what the peak current is. \$\endgroup\$
    – John D
    Oct 19, 2020 at 20:00


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.