The problem with "diameters" of globular clusters is that there are different ways to measure this parameter, e.g. the angular radius or diameter containing half the total light, or the radius containing 90% of the total light.
(and the measured diameters for the same object can therefore be incompatible)
As far as how to find the actual
physical size of a specific globular, from its
angular diameter, the first thing we have to know in order to calculate the size is the actual
distance of the globular.
This link to a (year 2010) catalog by Willam E Harris, a noted researcher, gives
reliable distances and other properties for all of the Milky Way's globulars.
http://www.physics.mcmaster.ca/~harris/Databases
An interesting discussion, folks!
I might have a go at comparing the physical sizes of these globulars, and a few others.