• State
  • Locked Locked
  • Replies 6 replies
  • Subscribers 80 subscribers
  • Views 1929 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

UART failure on MSP430G2403, ESD?

Nick Hazzard
Prodigy 145 points
Other Parts Discussed in Thread: MSP430F5510

We recently had a failure of the UART RX (pin1.1).  There appears to be a low impedance path to ground which causes the voltage to drop to around 0.7V (3V expected) while connected direclty to the output of another micro (MSP430F5510).  When connected to a 3V source directly it sinks about 40mA into the pin, so it appears as though the impedance is less than 100Ohms.  However, when I probe with a DMM the resistance to ground is on the order of 45-50kOhms.

There are no current limiting ressitors or ESD clamping diodes connected to this pin and I suspect the failure is due to ESD damage during handling.  My question is if the failure mode is consistent with ESD damage, or if it could be something else.

  • 0
    Guru 12050 points

    When using two connected MCU try to use open-drain system to communicate with each other (10K pull-up's) the two states is: input and output-low
    As UART is old they did not think about that, so most of the time TX and RX is push-pull and short circuits can happen.

    Put a 100-300Ohm in series incase there is a software bug while setting up pins when you're developing the software.

    What measurement do you get if there is absolutely no software (full erase of main-memory) on the G2403.
    As this should guarantee to put the pin as input with no pull-up or pull down, and will have 1+mega ohm as long you input Vcc +0.3V max.

    45K Ohm sounds like the internal pull-down resistor, so it's looks like your software is doing something you don't have full control over.

  • 0 in reply to Tony Philipsson
    Prodigy 145 points
    Hi Tony, thanks for your suggestions. I probably didn't mention that when I probed with the DMM the circuit was not powered. If the MCU is powered on I get about 26Ohm to GND on RX. I did a full erase and still get 26Ohm on RX, while every other pin measures in the 20MOhm range.
  • 0
    Guru 227245 points
    There _are_ clamp diodes. Every digital I/O pin of an MSP has internal clamp diodes to VCC and GND. If a voltage >VCC or <VSS is applied to an MSP pin, it is shorted to VCC/GND by these diodes. However, a rated current >2mA will eventually destroy the diodes, which may lead to a short, rendering the I/O pin useless.
    If one MSP was unpowered while the other one was active, this might be the reason.
    A series resistor to protect the input side is a good idea. Of course it only works for plain input pins, as a series resistor on an pin that is used as output (even if open collector) would be counter-productive.
  • 0 in reply to Jens-Michael Gross
    Prodigy 145 points

    Thanks for the reply!  Do you think that if I measured 45-50k Ohms from the RX pin to ground with the MCU powered off that it still could be a shorted clamp diode?  I am only measuring about 26 Ohms into the pin when the MCU is powered on.  If it was a shorted clamp diode wouldn't I expect a low impedance path to ground regardless of the state of the MCU?  It appears as though the RX pin is clamped to ground as an output regardless of the state of the firmware.

  • 0 in reply to Nick Hazzard
    Guru 227245 points
    It doesn't look like. However, when it is about diodes, resistance is futile. :)
    Depending on the instrument you use for testing (fixed voltage with current sense, fixed current with voltage sense) a 'shorted' diode can show any resistance.
  • 0 in reply to Nick Hazzard
    Guru 27285 points

    Nick Hazzard said:
    If the MCU is powered on I get about 26Ohm to GND on RX. I did a full erase and still get 26Ohm on RX, while every other pin measures in the 20MOhm range.

     Try hold reset pin to gnd and measure again, if still persist low resistance then I fear pin is damaged by ESD or other problem arise from what pointed by others here.

     SO if device is bad no other solution than rework device fit a fresh one then add series resistor and ESD protection devices too, a good idea is to add also EMF network to fully protect lines come from external off board RF income from cell phone and or local radio transmitters, high power devices, brushed motors etc...

**Attention** This is a public forum