In follow up to my previous post about the effects of inductive
ballasting a pole pig, which I don't totally believe myself, I've been
using spice to model this system. I'm getting results that I'd expect,
leading me to look for ground loops in my real measurement setup.
The result confirm some of my measured results - 53KV out instead of
23KV - not quite the 400KV that I think I measured;) ftp over to
ftp.bdt-dot-com/home/jim.fosse for the files: bal5_sch.jpg, bal5out.jpg,
bal5_sch.jpg shows my 9mH ballast in parallel with 5 ohms of
resistance feeding into the 110V tap of my LL-65 10KVA pig. A straight
220V into this should give 14.3kV RMS or 20.2kV peak.
bal5out.jpg The top graph is the pigs secondary output. Notice the
53kV peaks. This is with just loose wiring inductance in the HV
section - NO TC. The bottom graph shows the AC into the ballast
(Vplug) and the voltage into the pig's primary (Vpri) Note how the
spikes into the primary exceed the AC in the later 1/4 cycle.
bal5_i.jpg The top graph shows the pig's primary current. For
reference, the bottom graph is the same as in bal5out.jpg.