Number of parameters in LeNet-5

In a week2 lecture2, Andrew mentions that LeNet-5 has 60,000 parameters. I am not able to understand how this number is arrived. There are three fully connected layers with 400, 120 and 84 nodes at the end of the network. Wouldn’t there be more W’s and b’s than 60000? Please clarify. Thank you!

The W matrix that links the first two layers has size 500 x 120. That’s 48,000 weights.
The W matrix that connects the 2nd and 3rd layers has size 120 x 84. That’s 10,080 weights.
Add them together and you get about 60,000.