A new attack against the Kirchhoff-Law-Johnson-Noise (KLJN) key distribution system is introduced. The attack is based on 1) Utilizing the dc-voltage-source - which we put at Alice’s end in our mathematical modeling of the attack-that could exist due to errors, imbalances, Electromagnetic interference and etc.2) On studying the number of samples per bit in the security key that the measured Alice/Bob voltages exceeds or falls below a threshold voltage, respectively. The threshold voltage is the average between dc voltages across low and high resistors- generated by a dc-voltage source. We count the number of samples the voltage at Bob’s end (containing both the noise and dc components) exceeds the threshold voltage and how many times it falls below the threshold to judge whether the resistor can be guessed as low or high for every cycle. Also for a pre-specified key-length we count the number of high resistance estimations per bit –at Bob’s end-according to the previous criterion and the non-successful estimations per bit to judge the final guessed resistor value at Alice’s end and Bob’s end, where if we have more bits with most of its measured samples are above the threshold voltage then we will predict Bob’s resistance as high resistance, otherwise we predict Bob’s resistance to be low resistance. The Simulation was conducted and the attack proved that it is successful unless the temperature increased dramatically to ranges more than a threshold temperature ~ Kelvin that increases when the number of samples per bit increases.