Ok I built a temperate sensor with a digital LED display for work. It works great. However when I designed the PCB I forgot one crucial part. This is also designed to connect to the monitoring equipment at my transmitter and that will give me an alarm when the AC fails and temperatures gets to high. This is a metal building in Florida with a 21kW transmitter in it. The place can get REALLY hot and fast.
So I fixed my PCB error after the fact by adding connection points inline to the temp sensor (LM35). These are just pads add to the trace that runs to the ADC. I also added a pad for ground to give my monitoring equipment. Ok insert black magic here ..... When I connect a wire just a single wire I didn't even connect it to the monitoring equipment yet. All I did was take a 2 conductor with shield cable and soldered one to the ADC input and one to the ground. They are not connected together at all on the other end. Now the registered temperature of the MCU is higher. This means the voltage into the ADC has increased. This is where it gets weird. I would expect the voltage to go down from a resistive load not up. So I shortened my cable thinking maybe some weird chance that it was picking up some stray frequency. No luck. It is not random either. Its the same every time. It increases the voltage enough to equal 5 degrees.
I'm using an Zilog Encore MCU. I took classes on it in college and am only now converting to AVRs. However the Encore run on 3.3 but the LM35 runs on 5 volts. So I have a voltage regulator in the circuit dropping the input voltage of 5 volts to 3.3. The voltage sensor is on the 5 volt side but is grounded to the same ground as the 3.3v side. The sensor will not output more then 3.3 volts unless the temperature is crazy high like 200 degrees F or something like that. So I just connected the temp sensor output to the ADC input on the MCU.
I'd post schematics but I only have them in visio documents as that's what the make us use at work. I'm trying to switch to eagle but the rest of my department only does diagram drawings instead of actual schematics so they want visio documents for reference and I don't want to redraw things twice. Anyway any help would be great. I am at whits end with this thing. I also noticed a 3.42 K Ohm resistance between the positive and negative power supply input (5V DC output wall wort)