We are using the TPS2543 USB controller IC in one of our designs, and I’m having issues with it when powering/controlling a USB wifi module plugged into it. I’ve traced it down to a USB voltage issue that’s causing us to drop a lot of voltage across the USB cable between the two devices (with oscillation on the voltage, with drops above Ohms law drop of the cable).
I first thought it was an issue with the power supply we are using to power the IC, but I did a bunch of tests and can’t get the voltage drop/oscillation issue to clear up. I’ve completely by-passed the modules internal supply and provide my own 5V power to the IC. I’ve by-passed the PCB impact of the 5V output of the IC by flywiring around it (and adding extra output capacitance.
I’ve also taken similar voltage measurements using the same USB cable and same USB Wifi adapter but plugged into a USB port on my monitor and don’t see the same level of voltage drop and no oscillation on the voltage.
So I’m starting to wonder if we’ve got the IC configured wrong. I don’t know the full details of the software on the module that’s got the TPS2543 on it, but I probed the board to check the Control inputs and current limit inputs, and we are using
CTL1=Logic 0
CTL2=Logic 1
CTL3 = Logic 0
ILIM = Logic 1
I did probe on the fault status output, and don’t see that going low, so I don’t think it’s current limiting (and I changed the current limit config resistors to set the limit to ~2.3A). The Wifi adapter is a high-power USB device and pulls about 500mA.
If you’ve seen that type of behavior before, or have thoughts on what could cause that, any suggestions would be helpful. Below are the scope captures describing the behavior.
Here is a scope shot showing the behavior when the TP2543 is powering the Wifi module:
For comparison purpose, here is what the voltage traces look like when the Wifi device is powered with the same cable but using the USB port on my computer monitor to power it. I don’t see any of the voltage oscillations, and the voltage drop across the cable looks more like a normal IR drop (the cable has ~250Ohms of resistance on both the VUSB and GND lines).
