On 20 Feb 2011 at 20:16, Tony Duell wrote:
As I am sure you know, the '2W' rating applies
if the resistor is in
air at 20 degC or something like that. It derates as the surrounding
air gets hotter. This is why, for example, a 1M, 1W resistor with a
amximum voltage rating of 300V is not as silly as it first appears...
But it would have to be in awfully hot surroundings for that to
matter.
And one would expect to see at least a little browning of the PCB
underneath, but no.
Is it possible the resisotrs were defecting (e.g. the
carbon film was
contaminated in some way)?
I can't think why it would matter, but is there a
significant HF
component to the current through this resistor?
Well, there's 40KHz at full line voltage flowing through it.
I'm going to strip the paint off the body of the resistor and see if
I can find what the failure mode is. It's very odd--I'm not used to
seeing resistors that overrated failing.
It could in fact be a manufacturing defect.
The circuit follows a couple of reference designs with no apparent
corners being cut--the PCB solder side has been sealed with a coat of
clear lacquer of some sort. These, unlike some other T8 ballasts,
are true 4-wire ones with a preheat circuit for the filaments--after
they're powered on, about 5 seconds elapses before the HV is applied.
A lot of cheap ones just apply HF to both ends of the lamp and not
even bother about the filaments.
I'll keep you posted on what I discover. It might be that there was
simply a bad batch of resistors.
--Chuck
Any chance ofa very high peak voltage accross this resistor from time
to time whoch could lead to flash-over between the 'turns' of the
carbon film?
-tony