On Mon, 26 Mar 2012 00:13:49 -0700, Jeff Liebermann wrote:

On Mon, 26 Mar 2012 00:53:26 -0500, The Daring Dufas
wrote:

Back about 4 decades when I tried my hand at TV repair, me and the guys
I worked with had to replace a lot of what were called
"glowbars/globars", they were actually PTC devices hooked to the
degaussing coil around the CRT of a TV set to power the degaussing coil
for a short period every time a TV set was turned on.

That type of thermistor works backwards from the flavor I'm proposing.
For the degaussing coil, the thermistor has a low resistance when
cold. When it gets warm, the resistance increases dramatically. When
installed in series with the degaussing coil, it allows an initial
blast of current through the coil, followed by effectively turning
itself off.

That's a PTC (positive temperature coefficient) thermistor. Its resistance
goes up with temperature.

In the case of protecting the thermal fuse, it's the other way around.
When cold, the thermistor has a fairly high resistance, thus limiting
the peak inrush current. As the current heats up the thermistor, the
resistance decreases, allowing the device to operate normally.

Nope. A PTC protection device (a.k.a. "polyfuse" or "polyswitch") is also a
PTC device. It has very low resistance when cold. When excessive current
flows through it, it heats and becomes open, more or less. It's only reset
when the voltage is taken off and it's allowed to cool. It wouldn't work if
it had a high resistance when it was cold.

Both devices are PTC thermistors, albeit with different characteristics. NTC
thermistors are normally used as temperature sensors.

The only problem would be
coming up with a thermistor that could handle the current of the heating
element when it's connected in series with it.

It's not current handling but rather energy handling capacity. If the
inrush current surge lasts too long, the thermistor will get rather
hot. If this were a design exercise, I would need the steady state
current, the peak inrush current, and the approximate time duration in
order to calculate the energy dissipated (in joules or watt-seconds)
and eventually the maximum thermistor resistance.

Any sane person would just buy the one designed for the application.