16-06-2015, 02:08 PM
(16-06-2015, 01:20 PM)Yorkie Wrote: That said, I can't really think of many, if any, situations in which 0.25 Watt resistors such as these are likely to have very high voltage across them. and presumably it's the potential difference across the resistor that determines is ability to survive. Hence, if a resistor is required say to drop voltage at a point in a circuit from 250V to 150V, it needs to tolerate 100V - not 250V, or am I missing something?
The voltage rating question is an interesting one.
We all know that Power = Voltage squared over R. Hence, Voltage = square-root(Power times R)
From this, we can say that a 100k 0.25W resistor shouldn't have more than 160V across it - if it does, the power rating will be exceeded. Simple stuff
Of course, the voltage rating of a resistor is distinctly separate to this. Most 0.25W resistors are rated for 250V. So, although the above suggests that a 1M resistor could have 500 volts across it before the power rating is exceeded, the voltage rating will cap this...
OK, so actually, that's also pretty simple. But there's a further issue...
High value film resistors with lots of volts across them will eventually fail high or O/C. By "high", my usual rule of thumb is 100k and 100V, but I've seen failures with half those values. I've written about this before, so I won't repeat myself here.
So, I wonder if Nick was alluding to that particular failure mode? It might not be a safety thing as such - although arcing is possible - but it might be that projects working from more than a few 10s of volts might not have the long-term reliability you'd ordinarily expect.
Regarding the question about 100V or 250V, there is no simple answer - every situation is different. General thoughts: can you be sure that there will never be more than the usual voltage under all conditions (normal and fault)? If not, what might it rise to? And for how long? And how much current could be supplied? What is the effect of failure? Ultimately, it's a judgement call. And it will be guided by knowledge and experience. So, for example, my "rule of 100s" would guide me away from using 0.25W film resistors from certain situations - choosing a metal oxide type instead.
For more insight into these sorts of issues, have a look at Bob Pease's book: Troubleshooting Analog Circuits. It's an old one, but it's still very useful. I bought a paper copy many years ago, and by coincidence was reading through it just last night. Chapter 3 talks about resistors.
Cheers,
Mark
