Love-o-meter : Big Love....


I have implemented the project number 3 : the love-o-meter.

Once the first LED is lit, measured temperature is getting very high (more or less +10 °C for each LED which is lit).

As required in the book, each of the 3 resistors is 220 Ohm.

By switching to a 10 kOhm resistor, project works properly.

By measuring tension (Volt) around the TMP device, when temperature is getting very high, it goes from 5,07 V to 4,77 V : more or less 3 * 10 °C, because of 3 LEDs ? Maybe just coincidence.

Does anyone have an explanation ? Has someone faced the same situation ?


Hi, I'm facing the exact same situation. I think even the small load of 2 LEDs decreases the Arduino board's 5V pin voltage, affecting the TMP36's output.

I measured more or less the same voltage drops on the TMP36 as you did.

The approaches I'm planning are: - switching to an external power supply (no PS from the USB port, but using the jack) - using bigger resistors

Hi Guys,

I got exactly the same behavior. Each LED which is turned on increases the voltage on the measured analog input. Normally in bigger ECUs the ground for digital and analog input signal are separated because of the "dirty" ground a digital signal often brings with itself. But its hard to believe that only one LED has such an influence and that this has not been tested by the Arduino team. Because the result is that always all three LED are set to HIGH. If this would be a general problem more people should react on it. Maybe it is a hardware issue of some boards? What do you think? I don't have a solution here, but an answer or statement from the Arduino guys would be nice.

Greetings, fostgate

UPDATE: After I played around a little bit with the ground connection (e.g. moving the cable or moving the connector a little bit in the pin hole of the bread board) I found out that the ground connection was (or still is) not very good. I realized a big influence of this cable movements to the measured values in terminal (up to 300mV). After I plugged the ground line in and out the breadboard a few times the measurement became better and more stable. But it is still sensitive if I move the cable (even if much better now). So after all I think if the ground is connected properly, the measurement became more stable. The voltage toggles about 20mV now.


I encounter the exact same problem. The coding is right and all connection are properly plugged. The readings from the multimeter are as they should be. Still, when the temperature gets higher to light up the first LED, all of them light and the temperature readings get higher (more than +10 °C for each LED which is lit). The voltage of the sensor drops from 5.09 V to 4.80 V. I think the LEDs somehow draw from the voltage and messed up the readings. The voltage I measure from the sensor with the multimeter is 0.70 V which is the same as when the LED are not lit but the serial monitor shows much higher voltage and temperature.

Does anyone have a solution to this problem?

Best wishes, Future Projects

Hi all,

Same problem here, when the first LED is turned on, the voltage have a big jump. Output below:

Sensor Value: 151, Volts: 0.74, degrees C: 23.73
Sensor Value: 151, Volts: 0.74, degrees C: 23.73
Sensor Value: 152, Volts: 0.74, degrees C: 24.22
Sensor Value: 162, Volts: 0.79, degrees C: 29.10
Sensor Value: 171, Volts: 0.83, degrees C: 33.50
Sensor Value: 178, Volts: 0.87, degrees C: 36.91

The first LED turns on at 24 degrees C and immediately jumps to 30 degrees without never getting lower.

Did somebody find an explanation to this behavior? Thanks

UPDATE: I solved taking suggestions from this post. I changed wire and breadboard slot for the ground and now everything seems fine.

If anyone is searching for solution see this post