The round-off error obtained upon quantizating a signal is the quantization error. Using a smaller step size upon quantization, we can show that the mean squared error will be 1/12 of the square of the quantization step. The mean squared error mentioned above can also be referred to as the Quantization Noise Power.
[tex]Quantizationhspace {2mm}noisehspace {2mm}powerhspace {2mm}=hspace {2mm}sigma_{q}^{2}=frac{triangle^{2}}{12}[/tex]
[tex]Quantizationhspace {2mm}stephspace {2mm}=hspace {2mm}triangle=frac{A}{2^{n}}[/tex]
A is the full range of the analog signal to be digitized and n is number of bits per sample.
Now all you need is to do your math.