It's a 500k logarithmic pot - not linear. Maybe it is faulty, but at half of its rotation, there will little change in resistance, (about 10% - 50K, then the other 90% - 450k would be in the other half of the rotation). That's quite unlike a 500K linear pot, which at the halfway point would read 250K. Before you change the pot, measure the resistance on a replacement one (470k log) at the halfway point and see what result you get.
You could do the same test on your existing pot - set the knob to the centre point of rotation, clip one Ohmmeter lead to the centre pin, then the other lead to each of the two pins on either side of the centre pin and note the two readings. If the result is roughly 50k and 450k, that suggests the pot is working correctly.
If you look at the graph here, you'll see how linear, log and reverse log pots behave.
http://www.geofex.com/Article_Folder...s/potscret.htm
As has been said several times, you won't make much headway by making assumptions and changing things on a whim hoping that you might get lucky.