Pi Zero GPIO pin stuck at low even when pulled up

[yy502]

Background

I've been using a DHT22 sensor with my RP0 for some time already and it worked perfectly well.
I use a 10k Ohm resistor to pull up GPIO 5 (BCM 24) to 3.3V, and all the voltage tests fine using a multimeter. i.e. 3.3V across GPIO 5 and GND.

Issue

One day I found my DHT22 isn't reading data and it turned out GPIO 5 is stuck in low (reads 0 using "gpio readall" by WiringPi) even it's physically pull-up?! I tried to change it to out put mode and set it to high using WiringPi. The mode changes fine, but the output is always stuck at low.

Reboot, upgrade firmware didn't make a difference.

Please share your experiences in debugging similar issues with GPIO. Thanks.

[joan]

With nothing connected to the GPIO try wiringPi's pintest utility or try my gpiotest.

[yy502]

With nothing connected to the GPIO try wiringPi's pintest utility or try my gpiotest.

will do this evening and post an update. thanks for the advice!

[yy502]

With nothing connected to the GPIO try wiringPi's pintest utility or try my gpiotest.

got some faults

--> Pin 4 failure. Expected 1, got 0
--> Pin 5 failure. Expected 1, got 0

Is there a way to fix these?

[RaTTuS]

move to another pin, but post your circuit diagram first so we can fix it

[DougieLawson]

Is there a way to fix these?

It'll cost you a fiver or a tenner if you want on-board wireless.

[yy502]

move to another pin, but post your circuit diagram first so we can fix it

hmmm, it's a very simple circuit as below:

Code: Select all

+-------+
|       |
| DHT22 |
|       |
+-+-+-+-+
  | | |
  | | |
  | | +-----+ GND
  | |
  | +--+----+ GPIO
  |    |
  |  +-+-+
  |  |10k|
  |  +-+-+
  |    |
  +----+----+ 3.3V

I guess I can use another pin, but I'm curious why and how those GPIO pins are broken...

[yy502]

Is there a way to fix these?

It'll cost you a fiver or a tenner if you want on-board wireless.

yeh.... already ordered the new Zero W... but I just don't get it why the pins broke for no good reason?

[DougieLawson]

yeh.... already ordered the new Zero W... but I just don't get it why the pins broke for no good reason?

It's usually done by sending 5V (or higher) into a GPIO (that isn't true for the SDA/SCL pins which have 1K8 pullups).

[yy502]

yeh.... already ordered the new Zero W... but I just don't get it why the pins broke for no good reason?

It's usually done by sending 5V (or higher) into a GPIO (that isn't true for the SDA/SCL pins which have 1K8 pullups).

For the record, I moved the sensor data wire to GPIO 6 and it works again.

In my case, I soldered the wires to physical pins 17, 18 (now 22) and 20, so it couldn't be the 3.3v wire accidentally tapped onto the 5V neighbour pin.

And GPIO 4 was unused, which broke the same way as GPIO 5... stuck to low no matter what even it measures 3.3v.

[BMS Doug]

If the GPIO pin was ever set as an output and set high the lack of a current limiting resistor would burn out the pin. Add a 1k current limiting resistor to avoid the risk.

Code: Select all

+-------+
|       |
| DHT22 |
|       |
+-+-+-+-+
  | | |
  | | |
  | | +-----------+ GND
  | |       +--+
  | +----+--|1k|--+ GPIO
  |      |  +--+
  |    +-+-+
  |    |10k|
  |    +-+-+
  |      |
  +----------+----+ 3.3V

[yy502]

Thanks for the tip! I always thought RPi would have some kind of current protection built-in for the GPIOs.

I'm using 2 pins to control some relays. Do I need current limiting resistors for them too?

If the GPIO pin was ever set as an output and set high the lack of a current limiting resistor would burn out the pin. Add a 1k current limiting resistor to avoid the risk.

Code: Select all

+-------+
|       |
| DHT22 |
|       |
+-+-+-+-+
  | | |
  | | |
  | | +-----------+ GND
  | |       +--+
  | +----+--|1k|--+ GPIO
  |      |  +--+
  |    +-+-+
  |    |10k|
  |    +-+-+
  |      |
  +----------+----+ 3.3V

[BMS Doug]

Thanks for the tip! I always thought RPi would have some kind of current protection built-in for the GPIOs.

I'm using 2 pins to control some relays. Do I need current limiting resistors for them too?

I imagine that you are using some kind of relay driver board? If so then current limiting resistors shouldn't be necessary (as they will be built into the relay board). If you are trying to energise relays directly from the Pi then expect to lose another Pi, they cant handle the load directly.

[yy502]

I imagine that you are using some kind of relay driver board? If so then current limiting resistors shouldn't be necessary (as they will be built into the relay board). If you are trying to energise relays directly from the Pi then expect to lose another Pi, they cant handle the load directly.

Yes, it's one of the common blue relay modules.

Thanks again for your insights.

[BMS Doug]

I imagine that you are using some kind of relay driver board? If so then current limiting resistors shouldn't be necessary (as they will be built into the relay board). If you are trying to energise relays directly from the Pi then expect to lose another Pi, they cant handle the load directly.

Yes, it's one of the common blue relay modules.

Thanks again for your insights.

You are welcome, happy Pi'ing