Good point Geaff, this is why we have to prove thermisters by geting the test data and proving it functions correctly.

I recently changed a Worcester 240 pcb for a customer because the relay was frequently making and breaking. Discussions with Worcester confirmed my choice of part. Now they've phoned me to say it's doing it again. Perhaps it was the thermister? I've bought one to take round. If I was wrong about the pcb I'll refund them in full and fit the correct part. I've also taken the precaution of purchasing for £2.50 a suitable relay from RS in case it's the relay. I hadn't learned from my own lesson. Relay clattering like mad, must be the relay, customer wants it fixed same day, can't get a relay in time round to hrpc for a pcb.

One bug bear, suppliers won't take pcb's back, so now that's another pcb in my stock.

Have you any take on this particular problem?