Many electronic projects benefit from being run from lithium-ion or lithium-polymer batteries. One of the complexities involved in this is the need to charge them. While I’ve seen some crazy people hook up “protected” Li-Poly cells directly to USB 5V (don’t do this), proper charging involves an initial trickle, a constant-current phase, a constant-voltage phase and a termination phase. As a result, it’s advisable to rely on a proper charger IC which can be rather inexpensive in module form.

One of the most common is the Top Power TP4056, a linear 1A single cell charger IC in an SOP-8 package. It’s simple to implement and low-cost, although the linear nature makes it somewhat inefficient. It’s so popular that there are plenty of different modules being sold – some of which have clones such as the XJS4056 from JXND Electronics Co. Ltd.

This module comes to me thanks to Platima (YouTube | GitHub) who also sells this in their shop for AU$0.35 at the time of writing.

The Item

The package arrives with the PCB inside an anti-static bag, but *do you spot a problem here? *Yep. The heat seal on this bag is below the easy-tear-open cut-outs meaning that you won’t have an easy time opening this package without some assistance. Thankfully, I do have a pair of scissors to break this nearly impenetrable forcefield (at least, with regards to ESD).

The board itself has no identification except for “090” below the IN+ marking. Input is via mini-USB which is quite old-fashioned. Seriously, *it’s getting quite hard to find those cables lying around anymore! *But thankfully, input is also available via two through-holes which you can solder your own wires to. Similarly the battery connection is via two through-holes as well. The implementation has a 0.4 ohm dropper resistor, two indicator LEDs with 1kohm series resistors and a 1.2kohm programming resistor for 1A charging current. There are ceramic bypass capacitors on input and output.

The rear of the board has a mostly solid copper plane, for a low-resistance ground path and better heat dissipation. There are no markings or components on this side, but there seems to be quite a lack of thermal stitching vias on this side, but also, a bit of an odd hole arrangement for the USB socket with a pair of them plated annular holes and two of them being unplated holes. A bit of a PCB error perhaps?

The chip itself is an XJS4056 from JXND Electronics Co. Ltd., a clone of the common TP4056.

Test Results

Testing was performed using two channels of my Rohde & Schwarz NGM202 power supply. By using the battery model of a 5000mAh 26650 Li-Ion cell, I could simulate a battery and observe the charger behaviour in a consistent and reproducible manner. The charger PCB was connected using 4-wire kelvin connections to avoid voltage drop errors. A 1200uF electrolytic capacitor was added to stabilise the battery emulation output to avoid oscillations.

The charger appears to perform as expected with initial trickle current of 98.94mA until the battery reached 2.8744V. Then, the charger attempted to charge at the full current, but was limited thermally, resulting in a sloped current until about 3.4V where the current levelled out. The peak current was only 847.93mA, short of the promised 1A, likely due to thermal limitations. Charge termination occurred at 110.18mA and quiescent charger current once completed was 2.501mA, much of which is likely the green indicator LED.

The self-consumption of the charger can be determined by subtracting the battery input current from the USB input current. This results in a consumption about 3.2mA during trickle, 3.8mA during bulk charging and 2.5mA on completion. Large oscillations were observed during the early thermally-limited charging and during the taper period due to (likely) individual pulses or oscillations in the battery emulation. The 255-sample average (10Hz sample rate) provides a smoothed representation of the average current.

Thermal imaging using the Thermal Master P3 shows that the board, when trickling a deeply-discharged cell manages to sit at a relatively comfortable 44 degrees C. You can even see the heat in the dropper resistor for the indicator LED. During bulk charge, the external 0.4 ohm dropper resistor and the XJS4056 both share dissipation duties, reaching 77 degrees C on its exterior. The inside of the chip is likely much hotter, given the datasheet 145 degree C Tj monitoring. The PCB doesn’t seem to have quite as much copper and vias underneath to carry heat away, it would seem, given the concentration of temperature in the component.

During charging, the red LED is lit. Once charging is completed, the green LED is lit – very simple and intuitive, but with no indication of any progress.

Conclusion

The XJS4056 is JXND’s clone of the popular TP4056, a linear 1A single-cell charger in an SOP8 package. This module is a very basic implementation of this charger IC pre-configured for 1A charging current and no thermistor connection, with a rather old-fashioned mini-USB input and no protection circuitry. It seems to work, although thermally, the module does not seem to be ideal with the charger never quite reaching 1A and showing thermal limitations in early charge where the dissipation is maximum in my testing using a battery emulator. It also seems to want to trickle cells that are below the minimum “safe” voltage (e.g. 2.0V) and doesn’t seem to have any charge safety timeout either. But that’s just par for the course – it’s a very simple, low-cost linear charger and a clone of a ubiquitous one. If you want faster or more efficient charging, better safety or more granular status, you’ll need to look elsewhere.

Thanks to Platima (YouTube | GitHub) who provided this module for review.

About lui_gough

I’m a bit of a nut for electronics, computing, photography, radio, satellite and other technical hobbies. Click for more about me!