D12 input influenced by D9 output on Nano

[adam_studiorip]

I configured D12 as input with pullup, D9 as output. I would expect D12 to read a stable HIGH (being unconnected and having pullup), but instead it reads the value of D9. This happens on a naked Nano without any hardware attached whatsoever. As a sanity check, I tested it on two different Nanos, the result is the same on both.

Below a simple code to replicate the problem, the values read on D12 will follow the value of D9 instead of being stable 1.

void setup() {
  Serial.begin(9600);
  pinMode(12, INPUT_PULLUP);
  pinMode(9, OUTPUT); 
}

void loop() {
  digitalWrite(9, 1); 
  delay(100);
  Serial.println(digitalRead(12));
  delay(1000);
  digitalWrite(9, 0); 
  delay(100);
  Serial.println(digitalRead(12));
  delay(1000);
}

[PaulRB]

@adam_studiorip

I tested it on two different Nanos

Two Nanos purchased from the same vendor in the same order? Perhaps they could be from a bad batch.

I will try to test your code on one of my Nano later.

[adam_studiorip]

Yes, same batch and same $1-2 Chinese copy from Alibaba. But still, I guess the chip is original, and the pullups are in the chip.

[kolaha]

@adam_studiorip

I guess the chip is original

yeh, sure

[adam_studiorip]

It seems to be the model above. But even for a Chinese copy, this is quite weird. Just tried with an external pullup, the same happens.

[PaulRB]

I just tried your code on one of my Nano. Its not a genuine Nano. Results as expected:

21:34:06.958 -> 1
21:34:08.050 -> 1
21:34:09.143 -> 1
21:34:10.236 -> 1
21:34:11.362 -> 1
21:34:12.455 -> 1
21:34:13.548 -> 1
21:34:14.640 -> 1
21:34:15.765 -> 1
21:34:16.857 -> 1
21:34:17.949 -> 1

@adam_studiorip

the pullups are in the chip

Maybe the pullups are not the problem. Maybe there is a short between the tracks of those those 2 pins somewhere on the Nano board.

Use your multimeter to measure the resistance between those two pins, while the Nano is disconnected. I get about 16~17K between pin 12 and most other digital pins.

Have you tried other pairs of pins to see if you can get the same problem?

[van_der_decken]

My "I really don't think so" meter was firmly in the red, but I tried your code on a couple of different clone Nanos I had here and the value read from D9 was not influenced on either of them by the value of D12. And looking at the PCBs with a magnifier, the traces for the two pins don't go anywhere near each other on either of my boards.

I think this could be one of those "the bitterness of poor quality remains long after the sweetness of low price is forgotten" moments.

[JohnRob]

I would try reversing the pin assignments and see if the reverse happens.

Do you have a ohmmeter? If so check for shorts with power off.

[TomGeorge]

Hi,
My test code.

void setup()
{
  Serial.begin(9600);
  pinMode(12, INPUT_PULLUP);
  pinMode(9, OUTPUT);
}

void loop()
{
  digitalWrite(9, 1);
  delay(100);
  Serial.print("Pin 9 OUTPUT HIGH   Pin 12 INPUT ");
  Serial.println(digitalRead(12));
  delay(1000);
  digitalWrite(9, 0);
  Serial.print("Pin 9 OUTPUT LOW   Pin 12 INPUT ");
  delay(100);
  Serial.println(digitalRead(12));
  delay(1000);
}

Results;

Pin 9 OUTPUT HIGH Pin 12 INPUT 1
Pin 9 OUTPUT LOW Pin 12 INPUT 1
Pin 9 OUTPUT HIGH Pin 12 INPUT 1
Pin 9 OUTPUT LOW Pin 12 INPUT 1
Pin 9 OUTPUT HIGH Pin 12 INPUT 1
Pin 9 OUTPUT LOW Pin 12 INPUT 1
Pin 9 OUTPUT HIGH Pin 12 INPUT 1
Pin 9 OUTPUT LOW Pin 12 INPUT 1
Pin 9 OUTPUT HIGH Pin 12 INPUT 1
Pin 9 OUTPUT LOW Pin 12 INPUT 1
Pin 9 OUTPUT HIGH Pin 12 INPUT 1
Pin 9 OUTPUT LOW Pin 12 INPUT 1
Pin 9 OUTPUT HIGH Pin 12 INPUT 1
Pin 9 OUTPUT LOW Pin 12 INPUT 1

@van_der_decken

My "I really don't think so" meter was firmly in the red,

What do you mean, what meter in what range testing what?

Tom..

[adam_studiorip]

Thanks for everyone for doing the test!!!

In the meantime I realized that the problem may not be the poor quality cheap chip, but the fact that I forgot to tell an important piece of information (for which I apologize, I didn't realize how important it was!!!): it is not a simple Nano, but an RF-Nano (from a batch initially bought for another project).

And the RF-Nano seems to have an onboard nRF24L01+ chip connected to D9-D13. The chip still allows the use of these pins as output pins (well, with a potential unexpected behavior in radiofrequency...), but it seems D12-13 cannot be used as input pins.

So I believe the mystery is solved. Thanks again for your time, and I apologize for hiding this information!!!

[adam_studiorip]

Yep, D9 is the chip enable signal for the nRF24L01+ chip, that's where the interference came from. Cutting the D9-D10 wires on the RF-Nano PCB miraculously solved the problem, so this is a short term cure until the batch of regular Nanos arrives.

Sorry again for this, I didn't realize how important piece of information the RF part in the name was!!!