It is also just possible that you have a N-E fault that is not apparent on the mains TN-C-S supply due to the very short neutral to the common point, but it becomes enough of a problem on the generator TN-S with a long neutral path back to the common N-E point.
Again, this ought to have been tested, a global IR measurement would show that up. Though if anything has been changed/disturbed after the changeover switch, etc, was added then it might not have been present.
Thanks again for the reply pc1966
My best pondering is done with a few pints in front of me and last night's session sat with dogs outside in the lovely evening sun gave me some good pondering time
What I believe is happening is there's a neutral to earth leakage in the garage after its consumer unit. The reason it is tripping a 100mA RCD in the house consumer unit only when the supply is from the generator and not from the mains is our mains voltage is rather low, under 220V a lot of the time.
The generator automatic voltage regulator is set at 245 to 250V (I need to check with a calibrated digital meter as the panel analogue gauge is probably none too accurate). I'm thinking Ohms Law is causing a higher current leakage at the higher generator voltage and it's then enough to trip the house consumer unit RCD.
What I really need is an AC low current clamp meter, but I don't have one and decent ones are expensive, so I'll have do it peasant's way and start disconnecting neutrals one by one in the garage consumer unit until it goes away. I'm suspecting an outside alarm PIR that played up then failed a couple of years ago. It
may have water ingress. But we'll find out! What fun.....
I suppose I could lower the AVR voltage on the generator to say 200V and see if it stops tripping, but an excessive neutral to earth leakage needs fixing anyway. I'll try and do it properly.
I can see me spending hours in the sweltering and cramped loft space above the garage again, and I'm not as nimble as I was!
EDIT:
It's a fluorescent light circuit in a workshop fed off the garage consumer unit. I used a Megger to test every separated neutral off the busbar until I found one with 400 or so Ohms to ground. Of course it's the lights in my workshop with a car lift in it, so they are hard to get to "up in the Gods", but at least I now only have eight fluorescent fittings to check. Running all on gennie power as I type
Thanks for the help. Anything I may have missed or misunderstood please let me know but we've been on generator power, house and workshops, running all and anything for two hours with no issues.