Direct Solar Charging

By direct solar charging, I mean charging a battery from a solar panel without going through an inverter to make AC and then a charger to again make DC.  Although this latter method is well-understood and straightforward, it's inefficient and expensive in terms of the equipment required.

The experimental setup shown below is charging an Electric Motion 5.7 battery (which requires 54.6 VDC to charge) from a Qcells 430W solar panel via a ZK-SJ30 ($32 Chinese buck/boost solar charging module).  Additionally, two inline power meters (not strictly required) are being used to monitor the experiment. 

Results

The EM 5.7 battery is rated ~1.2 kWh, and was at about a 50% SoC when the experiment began.  This battery requires 54.6 volts to achieve a full charge.  However, for the initial experiment I decided to use 53.3 volts instead.  This represents 4.1 volts per cell times 13 cells.  I wanted to leave a little margin for error in case the voltage overshot for some reason.  Although I believe the 5.7's BMS will open its charge relay to prevent cell damage, in never hurts to be cautious — especially when experimenting with unproven equipment.   

The solar panel produced 523 Wh during the charge interval.  Note that all three meters (coincidentally?) read the same power (209 watts) when the photo was taken.  Although I know the input-output conversion can't be perfect, the system appears to have excellent efficiency.  This is partly because the solar panel is emitting about 41 volts and the battery was at about 50 volts.  So the ZK-SJ30 boost converter did not have much work to do.  The current is also relatively low at 4 - 5 amps.

I know that the PZEM-51 meter (which monitors the output from ZK-SJ30) is fairly accurate (rated +/- 1%) but the display on the ZK-SJ30 itself is quite inaccurate.  I used a Fluke DMM to accurately set its output at 53.3 volts.

In prior testing, the inline solar power meter (which monitors the input to ZK-SJ30) seemed reasonably accurate.

At the end of the experiment, I used the 5.7's standard charger to “top up” the battery to 4.2 volts per cell.  It required about 100 Wh on the DC side to do so.

Qcells 430 Watt Solar Panel

A friend has a large (24 kW) solar power system.  One of his 430 watt Qcells panels was damaged during installation, but was repairable.  He offered it to me for experimentation.  This is a large commercial-quality panel measuring about 1 meter x 2 meters and weighing 25 kg.

I used heavy door hinges and a green-treated 2×4 to fabricate a temporary mount which was then affixed to a deck railing and adjusted to the proper solar angle via the table beneath. 

MPPT Definition

The power produced by a solar panel is related to the amount of sunlight falling on it.  The panel's voltage is maximum with zero load (but this also implies zero power transfer).  As the load is increased, the current increases, but the voltage simultaneously falls.  MPPT stands for Maximum Power-Point Tracking.  It is an algorithm that changes the load impedance (resistance) of the photovoltaic converter to maximize the power (volts x amps) delivered by the panel for a given amount of sunlight.

ZK-SJ30 

The ZK-SJ30 is a 700-watt Synchronous Rectification Buck/Boost Solar Charging Module with MPPT.  It sells for about $32 on AliExpress.  Specifications are shown below:

• Input voltage range: 6–80 VDC

• Output voltage range: 1.3–78 VDC

• Maximum output current: 30 A

• Maximum output power: 700W

• The input is not protected against reverse polarity

• The output has a diode to prevent “backflow” from the battery

The ZK-SJ30 also has a 12V @ 200 mA output for a DC fan, but I did not bother connecting one.  In my initial experiment the power was under 300 watts and the heatsink felt cool to the touch. 

The brain of the unit is Linear Technology's LT8795 (a synchronous buck/boost controller). 

My only complaint about the ZK-SJ30 that its three adjustment potentiometers (voltage, current limit, MPPT setpoint) are very poor quality and exhibited way too much hysteresis.

A 20-Amp version of this device is available slightly cheaper, but that seem like foolish economy as I expect the capabilities of both versions are somewhat exaggerated.  They also make a version that omits the display.

Setup

The devices were connected as follows:  

Solar Panel ==> Inline Solar Power Meter ==> ZK-SJ30 ==> PZEM-051 ==> 5.7 Battery

The input to the 5.7's battery was monitored by a PZEM-051.  This is described as an LCD DC Digital Ammeter Voltmeter Power Energy Monitor with a 6.5 - 100V operating range.  I use this device between charger and battery whenever I charge a Li-ion battery. 

The ZK-30SJ is configured via three onboard potentiometers: output voltage, output current limit, and MPPT tracking voltage.  In order to set the output current limit, I shorted the output and observed the current on the display while adjusting the pot.  

The MPPT voltage is recommended to be set at 3/4 of the solar panel's open-circuit voltage (49.38 V) so 37 volts.  Then I observed the in-line meter while tweaking the pot slightly to maximize power transfer.  After tweaking, the display showed 37.33 V for the MPPT voltage.