PORT D Read errors.

[KMill]

I have an STM32CubeIDE project using HAL for a STM32G071RBT6

I have Port D bits 0 to 6 set up for GPIO Input, with pull-up enabled.

In hardware these are connected to an 8-WAY dip switch, which when each switch is "on" connects the respective pin to GND. (PB3[STOPBITS] is connected to the 8th switch)

 

The issue is that when I read the port using

id = (GPIOD->IDR & 0x7F); /* Lowest 7-bits are from PORT D0-D6 */

Bits 0 and 2 are always stuck at 0.

With the board powerd off I have tested the board connctions to the chip and the dip switch, all seems OK. There is no short to GND on bits 0 and 2.

I checked the resistance on all the bits of port D and bits 0 and 2 have about 500KΩ to ground, whereas the rest of port D has about 7 MΩ

I have checked through the rest of my code and it's not doing anything else to PORT D, and the alternate-function registers (PORTD->AFR) read all zero immediately before reading from the IDR.

I have five of these boards, all made by JLCPCB (including assembly) and they all do the same thing.

What's causing the bits 0 and 2 to be stuck at 0? Does PULL UP not work on those bits? I could try hand-soldering some stronger pullups on, but it works for the other bits in port d, and it's the same on all boards!

[Andrew Neil]

If you measure the voltage at the STM32 pins, are they going high?

[KMill]

Thanks @Andrew Neil 

The 2 stuck pins are reading 0.24v and the rest are reading 3.26v with the swicthes open.

[Andrew Neil]

So the port is actually reading correctly - those bits really are at zero.

The question becomes: why are they not being pulled-up.

As a test, can they be pulled-up with an external resistor?

 

PS:

https://www.st.com/resource/en/datasheet/stm32g071rb.pdf#page=48

[KMill]

With a 4K7 pull up they are reading as expected (1 when switchopen, 2 when closed) but the voltage on the pins is only getting pulled up to 1.65 V (VCC is 3.3 V)

 

So yip, I agree, the quesion is what has happened to the pull ups?

Let me go read the PUDR register...

[mfgkw]

Which board do you use?

In DS12232 section Pinouts... PD0 and PD2 are said to have an IO structure of "FT_c" where the _c means "USB Type-C PD capable" and they have an additional function of UCPD2_CC1 resp. UCPD2_CC2.

Does your board have an USB port connected to them?

[Andrew Neil]

@KMill did you see my PS?

 

 

[KMill]

AHA! A smoking gun!

Thank you - that sounds like exactly the sort of thing that would cause this. I'll try strobing that bit and see what happens.

[KMill]

@Andrew Neil Thank you!

The STOBE Bit was the issue.

I set the appropriate bit in SYSCFG1 and now it all works as expected.

You have saved me a lot of grief! Who would have thought the Pull-Downs would be enabled when the UCPD Device is disabled?

[Andrew Neil]

---

@KMill wrote:

The STOBE Bit was the issue.

---

Great - please mark the solution.

 

---

@KMill wrote:

Who would have thought the Pull-Downs would be enabled when the UCPD Device is disabled?

---

Always good to check those little footnotes on the the pinout table - there's a few gotchas!