The common bicon balun (the traditional long-neck variety) may explain the differences encountered between V & H. If my feeble memory is correct, the balun is only marginally "balanced". Consequently, it could be sensitive to cable length and routing. (An easy test for imbalance would be to slide a ferrite along the coax and see if the measurement is affected.) I wouldn't be surprised to see such differences from V to H as reported by Don et al simply by turning the bicon around - while remaining at the same polarity. I'm not sure this problem exists on the small metal-box style of balun, or could it be worse somehow? Anyway, sometimes I wonder if shielded twinaxial cable, with a remotely placed balun, might perform more consistently.
Regards, Eric Lifsey National Instruments