Analog pin outputs 1.7v when set to input?

Hardware
41

Nice… let me take a look and see what might be the damaged part on the :spark: core :slight_smile:

42

kennethlimcp, this is the circuit:

It could be a bad or incorrectly oriented C15 or a flaky L2 (but not likely). The VDDA pin does have a voltage so L2 is not “open”. If C13 was “open”, the voltage would still be ok and “shorted” would ground VDDA which is not the case. If C15 is bad and/or has a high ESR, it could create the voltage drop being seen. :slight_smile:

1 Like
43

Yeah im hoping @TheNirEast can take a look and see what’s wrong before shipping back so we can learn more!

Thanks @peekay123. I’m out enjoying myself over the long weekend and haven’t see the circuitry ^^

1 Like
44

Has anyone figured this one out? I’m having the same problem on both of my spark cores:

  • Only using Tinker - no code written yet
  • 2000-2100 analogRead on all pins with nothing connected
  • ~2300 with TMP36 connected (using 3.3v* for reference)
  • The difference between the two reads approximates the correct temperature but not well enough to consider usable
  • Have tried the capacitor across the input and GND
45

Hi @jrl290

The analog inputs are not buffered on the Spark core, so 3.3V divided 2 or about 1.7 V is correct when the inputs are not connected. This is how the analog inputs work on the ST Micro ARM processor.

You need to drive the analog input with a medium-to-low impedance source (under say 20k ohm) to get a good value.

Have you tried adding the capacitor to the TMP36? That is known to help with that particular sensor.

Another thing to try if you think it is not working is a two resistor voltage divider with resistors 10k ohm or less tied to 3.3V* and GND.

46

Turns out my TMP36 was actually burning up on connection so the temperature reading might be accurate. I’ll have to figure that out first it seems.

My apologies for prematurely waking up this thread

1 Like