Taking inspiration from the Canadian topic
There are two ways I could do this. Firstly I could have one seat for approximately 70,000 voters. This of course has happened by chance; the number of seats in the UK has stayed broadly the same even as the electorate has increased. Instead I have divided the total adult population (18+) of the USA by 650 to give an electorate. This is making an assumption that voter registration in the US is higher than it is in reality.
So 234,646,600 divided by 650 =
360,995The Boundary Comissions are usually happy with between a 5% and 10% variance
So 5% variance 342,946 - 379,045
10% variance 324,896 - 397,095
For the purposes of this exercise, state boundaries will not be crossed. For those familiar with the Boundary Comissions work, in this exercise States will be treated similar to Counties and US Counties similar to Districts
So here we go. Staring with the Eastern Seaboard
Maine - 2.92 = 3
New Hampshire - 2.85 = 3
Vermont - 1.37 = 1 or 2
(Note: If state lines could be crossed a case could be made for a Vermont and New Hampshire to be allocated 4 seats between them)
Massachussets - 14.2 = 14
Rhode Island - 2.29 = 2
Connecticut - 7.63 = 8
New York - 41.7 = 42
Pennsylvania - 27.45 = 27
New Jersey - 18.63 = 19
Delaware - 1.92 = 2
Maryland - 12.25 = 12
District of Columbia - 1.39 = 1 or 2
West Virginia - 4.07 = 4
Virginia - 17.02 = 17
North Carolina - 20.10 = 20
South Carolina - 9.81 = 10
Georgia - 19.94 = 20
And for the record, California would be entitled to 77 seats