21 Feb
2011
21 Feb
'11
11:04 a.m.
Chuck writes:
I've been going over a small stack of failed T8
bulb solid-state
fluorescent lamp ballasts from about the mid 90's. They all have the
same failure and it suprised me.
There's a 2W 0.33 ohm carbon film resistor used as a current sensing
element. In all failure cases, the resistor has failed open, with no
signs of burning, but rather the outer paint flaking off.
To me, this is a puzzle. At 120v, the current through an 0.33 ohm
resistance in series with a 64W load is about half an amp. I2R gives
less than a tenth of a watt power dissipation across the resistor.
(There's also a 2A fuse in series with the whole circuit).
I replaced the failed resistors with 5W composition ones of the same
value, and they seem to work okay.
But the original failure has left me scratching my head. Does anyone
have an insight on this type of failure?
If the current is steady-state, your math works out correctly.
But... electronic ballasts are switching devices and the current
is not steady-state. If twice the current flows for half the time,
then the power dissipation doubles (Remember, I-squared-R,
you doubled I, so I squared goes up by a factor of four.)
All that said, it has little to do with your electronic ballasts :-)
I betcha the original resistor was a fusible resistor
to begin with. SMPS 101: when you see a blown up or burnt out
component, you can bet that some other component failed
and took it out.
Typically a fusible resistor will heat up to mildly-red-hot before
it blows. Do the same with a non-fusible metal film and it'll be
fine (if discolored!).
Tim.