I've put a circuit together for a 3S Li-Ion battery pack with an intended float voltage of 12.15V (max charge current set to 0.4A) based around the LT3652, (LT3652 Datasheet). When charging the pack from a bench top power supply (usually set to 18V and limited to 0.4A) the pack charges properly to the full voltage and towards the end it ramps down the current as expected before terminating the charge.

When charged from a solar panel (the original datasheet specified 5W, 0.3A Imp, 16.8Vmp, 21 Voc, 0.39A Isc, I've since contacted the supplier who have released an updated one: 5W, 0.29A Imp, 17.5 Vmp, 22 Voc, 0.32A Isc) it will only charge to around 11.9V before stopping. Measuring the voltage drop across the sense resistor I appear to be getting at least 0.31A of charging current after conversion. When handling the board I can't say I've noticed any thermal issues in the vicinity of the chip.

Additionally the input capacitor produces some acoustic noise when charging from the solar panel in some situations, I've been planning to swap to two lower value capacitors to prevent this. This may also be a symptom of a voltage oscillation on the input side which would explain the premature charge termination? I'm also a bit suspicious that my inductor value might be a bit on the high side.

I've attached the board schematic below (please note that the Vin blocking diode is integral to the solar panel so not on the board).

LT3652 Solar Charger Schematic

Any ideas why the charger isn't charging to the full float voltage when solar powered?

  • \$\begingroup\$ Measure panel voltage under charge conditions. It sounds as though the PV panel is charging the battery directly without the buck converter working as it should. || I have not looked at your IC specs ( please linkntoba good data sheet) but it sounds like it expects current than the panel will provide. There can be a number of of reasons for this that depend on IC and setup. \$\endgroup\$
    – Russell McMahon
    Commented Sep 12, 2021 at 12:50
  • \$\begingroup\$ @RussellMcMahon So the PV panel voltage (measured at the input to the board) hangs around 16.5V during charging which is what I would expect whereas the output voltage (measured at the input to the battery pack) follows the battery pack voltage (10.5-11.9V). In addition when the circuit is used with the benchtop supply is works just fine, no attempts at over current or over voltage charging. Thanks for your suggestion about the data sheet, I've added a link in the main description. \$\endgroup\$
    – user167210
    Commented Sep 12, 2021 at 14:16
  • \$\begingroup\$ Why a zener is needed for the BOOST connection? \$\endgroup\$ Commented Sep 12, 2021 at 18:51
  • \$\begingroup\$ @MarkoBuršič Page 12 of the datasheet: "To refresh the decoupling capacitor with a rectifying diode from the battery with battery float voltages higher than 8.4V, a >100mA Zener diode can be put in series with the rectifying diode to prevent exceeding the BOOST pin operating voltage range." \$\endgroup\$
    – user167210
    Commented Sep 12, 2021 at 20:10
  • \$\begingroup\$ Mystifying. You know that :-). I'd probably simplify the circui as much as possible and check all points with a meter and oscilloscope. || An unlikely just-maybe is to put a large capacitor across the panel - the MPPT may be being fooled by the panel response. (Less likely than not, but easily tried.). || Remove the LM234 temporarily. It shouldn't matter BUT it will work without it. || D2 value unstated - ensure Vboost is within spec. || Oscillation looking sensible ? ... || ??? ... :-) \$\endgroup\$
    – Russell McMahon
    Commented Sep 13, 2021 at 6:04

1 Answer 1


enter image description here \$V_{BAT_{FLT}}=12.15V \$

\$R_{FB1}\approx 920.5k\Omega \$

\$R_{FB2}\approx 343.2k\Omega \$

Lets calculate the voltage divider: $$\dfrac{V_{FB}}{V{out}}=\dfrac{343.2}{1263.7}=0.27$$

Yours: $$\dfrac{V_{FB}}{V{out}}=\dfrac{342}{1262}=0.27$$

I thought the divider was wrong, but it's OK.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.