Well, well, it was not the ballast at fault ! After further testing, the photocell was causing problems while testing earlier because I was doing the testing in my basement. Light reflecting off a white surface tripped the photocell, just as daylight would. Since it is a thermal relay, there is a delay of about 30-60 seconds for the bimetal strip to bend. As the lamp warmed up, intensity increased, and the PC did it's job. I intentionnally tripped the photocell with a spot light, waited a few moments and took the photocell off the fixture and in my hand. After the usual delay, I heard a click, but the PC was in my hand. The next test will be when the fixture is outdoor atop the pole. If problems persist, I will replace it.
On the plus side, I upgraded to a MagneTek ballast, which is superior than the Cooper ballast. Being MV only makes it lighter for a pole top fixture.
Many thanks to all who tried to help.
![Smiley :)](/Smileys/default/smiley.gif)