Close

Charging output

A project log for urban-edc-flashlight

Multifunction flash light for everyday carry

Christoph TackChristoph Tack 12/07/2021 at 20:020 Comments

As a power bank, this device should be able to supply 5V/1.5A, because that's what each of my tablet and phone are drawing during charging.

Dedicated power bank ICs exist that incorporate a boost converter and battery charging and protection circuitry (Thinkplus MP5043/MP5045, Hotchip HT4936S & PB0059E, Injoinic IP5306).  These don't need many external components.  The main drawbacks are their limited output current capability and they turn off when the output current is too low.

Another option is to design our own PCB using DC/DC-converters available : FeelingTechnology FP6276, FP6293, FP6298 or XYSemi XR2981.  Designing our own power supply board would take more time.

Option 1: 5V/3A boost converter

AliExpress 5V 3A DC-DC Step Up Power Module : €1.03/pce

XR2981 module (cdn.hackaday.io/images/7300101638905789058.jpg))

Uses Xysemi XR2981 boost converter., can also be found as H&Msemi HM9228B (hmsemi.com)

5V output is actually 5.15V (cdn.hackaday.io/images/6733581639160347879.png)

The measurement stopped at an input current of 3.5A, which was the maximum for my lab power supply.

Option 2 : Qualcomm QC2.0/3.0 Fast charge module TPS61088A

AliExpress TPS61088 Boost power module : €2.21/pce

TPS61088A module (cdn.hackaday.io/images/4184451638906752621.jpeg)

This module is more expensive, but it has the potential of faster charging by boosting the USB-voltage to 12V:

The no-load current is 548µA with a supply voltage of 3.7V.  The EN-pin of the TP61088 is hard connected to VCC.  Even disabling the TPS61088 would still leave us the 200µA current draw of the FP6601Q.

UVLO falling is 2.4V after which 0.5A is drawn.  If the input voltage rises again above 2.6V, normal operation is resumed.

The TPS61088 is a synchronous boost converter, which minimizes losses that would otherwise be caused by the schottky diode.  There's also a larger inductor used here, which probably has a lower Rdc than the one in the 5V/3A-module.

The measured efficiency matches the TPS61088A datasheet very well.

Efficiency after correcting for resistive losses in cables and in the QC-selector module (cdn.hackaday.io/images/5690061639162162726.png)
Measurement setup (cdn.hackaday.io/images/3155791639167441265.jpg)

To be able to test it with an electronic load, I ordered a QC2.0/QC3.0 tester (AliExpress QC2.0 tester : €3.53/pce).

(cdn.hackaday.io/images/1161241638907139365.jpg)

Discussions