0

I'm trying to create a charger that knows how to switch between OTG and charging mode. The OTG side is working, and my Samsung S4 goes into charging mode. However, charging is really slow. It even loses charge, even with the charging indicator on.

There are two things that do concern me about the schematic I came up with:

  • A diode is used on the Vbus line, causing my 5V supply to become a 4.3V supply. However, current is coming back as it should, at 760mA
  • I did not follow the S4 charging specification, shorting the data lines, adding a 10k resistor, and grounding the connection

I placed my order of 15 PCBs, and cannot exactly place a new order with new fixes. What's a good workaround for my current situation?

Some things I looked up due to the current situation:


Edit 2/5/15:

I modified my circuit a little bit, so that my attachment does not need to worry with the charging specifications of different devices.

Would it work if I:

  • Used a switching regulator to raise the voltage up from 4.3v to 5.0v rated for 1.5mA, and
  • Used a USB multiplexer so that my circuit can support different charging specifications, while still fully functional for USB-OTG devices? I have attached a schematic for reference: enter image description here

The chips I used were:

WayneDinh
  • 61
  • 7

1 Answers1

1
  1. You may want to try to substitute a diode with a Shottky. This will give you back ~300mV and your phone may start charging again.
  2. Data lines can be shorted at USB connector during assembly.

Both workarounds can be done without PCB modification.

Oleg Mazurov
  • 3,231
  • 1
  • 14
  • 17
  • 1
    If he shorts the data lines in assembly, the OTG option would stop working. – Passerby Jan 12 '15 at 18:24
  • 1
    In this case, I would opt for P channel MOSFET reverse polarity protection over any type of diode. http://electronics.stackexchange.com/questions/61692/reverse-polarity-protection/ – Matt Young Jan 12 '15 at 18:56
  • Thank you guys! Will be looking through our parts to see if we have one laying around later in the afternoon. Shorting the data lines at this point is out of the question though. That modification would have to be made with our next revision. As for the MOSFET, we don't have a SMD solution for that right away...but I will keep that in mind. – WayneDinh Jan 12 '15 at 19:53