240V SP Ind Fan Motor wiring help

[plugsandsparks]

Also i have looked at the relays and i have continuity across the relay that switches pin1 on the first PCB and on the other PCB i have continuity across the relay that switches pin 2.
Checking the relays R1 for pin 1 is std 5A capacity relay, R2 for Pin2 - they have used the high capacity version of the same relay type (10A) and R3 for Pin3 is a std 5A one again, all NO . So two poorly relays as well as burnt tracks,,,

 

[Lucien Nunes]

Relays welded shut are consistent with the evidence so far. As I mentioned in #5, if two relays close at once (or one closes while another is welded) part of the main winding is shorted and the circulating current can exceed the supply current by transformer action. The impedance of the non-shorted section might be high enough not to blow the fuse at first, giving things a chance to overheat.

 

[plugsandsparks]

Yep, it was your mention of multiple relay action that made me check that they were not welded.

 

[Simon47]

I've seen in at least one commercial design, and I'd do it that way myself, using changeover relays so that relay 1 applies power to one connection, only when relay 1 is off does it's NC contact pass power to relay 2, and similarly for power to relay 3. That way, barring an internal failure in a relay, it's impossible to cause the sort of "shorted section" fault described.
BTW, when running on fast (or to a lesser extent on medium) the the slow connection is typically at a significantly higher voltage than the mains supply - with the motor windings acting as an auto-transformer. I vaguely recall measuring over 400V to neutral on a a cooker hood fan motor.