There is another thread going with some comments about resets. Perhaps it is timely to try to sort this one out.
Firstly, I don't think this is a picaxe problem as such. More a problem relating to the pics themselves.
Secondly, I've got about 50 picaxes doing all sorts of nifty things around my farm. Indeed, without picaxes I would have no water so it is all fairly important that it works. Also, I'm the crazy guy who bypassed the cooling fan controller on my car (after the standard mechanical solution corroded and cooked my engine) and replaced it with a picaxe. If that picaxe fails to turn on, I cook another engine, but it all has worked fine for three years, and I think I can say that I can build picaxe circuits that can survive in the engine bay of a car.
BUT - I still get resets and odd behaviour near 240V. The three culprits are 3.6kw motors, flouro lights and high current DC nearby. Of these, flouro lights turning on are the worst. The problem tends not to manifest itself on the workbench, but rather, when a circuit is assembled in situ and has wires more than 1 metre attached to the boards. These can be data wires, or they can be power supply wires.
I suspect that every wire behaves as an antenna, even wires that I might designate as "earth" or "ground".
I also suspect a solution involves something like a metal box, and every wire that goes into that box, especially supply and earth wires, goes through a ferrite of some sort. Maybe a few turns round a toroid won't hurt either (I've had some partial success with toroids).
In fairness to pic chips, I've also had problems with 4000 series CMOS and CMOS 555 timers. Robust chips include old school TTL, 324 op amps and 555 timers. But in one case, even those were beaten by the very close proximity of a large motor, and in the end the only solution was relay logic. Amazing what you can do with relays, but I digress...
One of the circuits in my pump controller is a DIN rail mounting timer. It has a led and a pot, and seriously, it does nothing more than a 555 timer connected to a relay. But, it cost over $150. We had a flood a few years back and everything had to be pulled to bits and dried out, and it was fascinating to see what the inside of this timer circuit contained, particularly as it was designed to sit right next to multi-kilowatt contactors on a DIN rail. It had metal shielding inside the box. There were filtering components, including some I didn't really recognise.
Bottom line is these things are really robust.
I wonder if it is possible to 'robustify' a picaxe? Metal box, ferrites, LC filtering on all wires going in and out. Local regulation inside the box?
And given all these components cost money, which ones can you leave out for which circuits?
