Thanks for all the help, particularly the last two comments which rally make sense.
I think that what you're saying is that trying to send power from the decoder to the live electrofrog runs a high risk of shorting because of the timing of the switching vs the timing of the blade movement.
Do I read that the only solution is to do without the frog polarity wiring and really on switch blade contact?
If so, any hints on how to improve this contact.....some of my points seem very prone to pore contact.
Thanks,
Ian