Andy,? I may have missed a couple of other observations but I would like to mention:
1) In situations with a lot of old fashioned 1/2 wave rectifiers or SCR voltage controls (arc furnaces, electroplating controls, certain PCs and the like) or in particular, simple light dimmers, there is a DC unbalance reflected back to the line transformer source.?? Line power transformers are likely to be running near saturation (it's an economic design choice.)?? The transformer is now non-linear and harmonic mixing will take place.? This effect even shows up in big substation links from ground currents such as Schumann resonances generating modulation side bands.
Another example.? I went to a customer site for a problem with our equipment.? They had racks of PCs being turned and off for the start and finish of test sequences.? The surge of too many adjacent test PCs simultaneously coming on coupled with too small wiring, led to PC power supply glitches affecting the test as each test start required the charging of the electrolytic capacitors in the step down power supplies.? A couple of cycles, but more than enough to cause trouble. The turning-on PCs ignored the glitch as they have have delay circuits built in and ignore their own disruptive effect, but already ON PCs were not happy.
Thinking back about the past, mercury vapor rectifiers were often in 1/2 wave configurations -- too expensive to wire up full wave rectifiers --
a 4:1 cost.
Your linear accelerator may have had a similar problem, needs for a lot of power on a pulse basis.? It may? even have been synchronized to the line as I believe some are to minimize line noise from data capture,
Regards,
Charles Patton
