NTC reduced lifetime at -25degc ambient

[Flyback]

Hello,
We have an 240VAC offline 60W Flyback SMPS LED driver.
It will be ON for 8 hours per night, and only switches on once per 24 hrs.
At the input we have a 3W , 2R7 power resistor (ROX3JS2R7) for inrush limitation.

We wish to consider exchanging this for a 3 Ohm NTC instead (link below)
(B57213P0309M3)

The thing that we fear about the NTC, is that at minus 25degC, its resistance is 20 Ohms…as such, when powered up at -25degC, the NTC will obviously be suddenly dissipating a lot of power (1.8W)…..As such, its internal temperature will suddenly rise and surely the associated sudden expansion of the NTC’s internals will result in it being weakened, and then failing early in its life?

B57213P0309M3 datasheet
https://www.farnell.com/datasheets/...MI4LOA0c_P3wIVWflRCh2r2A9_EAAYASAAEgJgR_D_BwE

 

[dknguyen]

So? That's no different from operation at any other temperature. Ambient rise is still going to be the same. It's rated for 3.8W anyways.

Maybe I'm missing something but would it not dissipate less power on startup when R is higher? Inrush current are not constant current loads (I think) and you're implying that if you used a 20 ohm resistor that it would heat up more than the 2.7 ohm which would require it to be a constant current load.

 

[Flyback]

The thing is,, at the lower temps, eg -25c, the resistance is very high, and as the temp gets higher, the resistance is so low above say 30degs that the dissipation in the ntc is quite low.

Also, for some reason it doesnt say anything about the resistance below -25degc, which seems ominous.? why is that?

 

[gophert]

The thing is,, at the lower temps, eg -25c, the resistance is very high, and as the temp gets higher, the resistance is so low above say 30degs that the dissipation in the ntc is quite low.

Also, for some reason it doesnt say anything about the resistance below -25degc, which seems ominous.? why is that?

Can you make a sketch of the circuit, snap a photo and post it?

 

[dknguyen]

Also, for some reason it doesnt say anything about the resistance below -25degc, which seems ominous.? why is that?

Becuase it is rated for commercial, not industrial temperatures. Why are you looking at a commercial temperature range NTC when you are worried about industrial temperatures? Look at industrial temp NTCs if that's your concern.

The thing is,, at the lower temps, eg -25c, the resistance is very high, and as the temp gets higher, the resistance is so low above say 30degs that the dissipation in the ntc is quite low.

I see what you mean and it seems to be at odds with I^2R where reducing resistance increases current which increases heat power even more. Must be missing something.

 

[kubeek]

The datasheet says that at -25°C the perimissible current is zero. That means there is no point asking about reliabilitiy or cyclic fatigue, you should simply not use that part at such a low temperature.

 

[dknguyen]

The datasheet says that at -25°C the perimissible current is zero. That means there is no point asking about reliabilitiy or cyclic fatigue, you should simply not use that part at such a low temperature.

Oh yeah. you're right. I've never seen a component that had power deratings at lower temperatures before...except batteries. Good to know.

 

[gophert]

some reason it doesnt say anything about the resistance below -25degc, which seems ominous.?

The datasheet says that at -25°C the perimissible current is zero.

I guess he can only apply negative power below -25°C

 

[dknguyen]

I guess he can only apply negative power below -25°C

Or provide the NTC with a PTC to heat it

 

[kubeek]

Oh yeah. you're right. I've never seen a component that had power deratings at lower temperatures before...except batteries. Good to know.

Me neither, but it looks like reading the whole datasheet is usually not a bad idea

 

[Flyback]

Can you make a sketch of the circuit, snap a photo and post it?

Thanks, Yes its as attached...the relevant bit....and also ltspice sim

 

[Flyback]

The datasheet says that at -25°C the perimissible current is zero. That means there is no point asking about reliabilitiy or cyclic fatigue, you should simply not use that part at such a low temperature.

Thanks, that really is something....i used to think that because the NTC would heat up due to having such high R at low temps, that it would be OK to use it...but it seems not.......so NTCs really do have issues with low temps.
I believe that we should stick to a 2r7 power resistor (3W).....this i hope will be able to handle the 2us inrush transient which is several kW's in magnitude., as seen on the sim.....the ROX3SJ2R7 datasheet says nothing about what is the acceptabel magnitude of a surge which lasts for 2us.
Does anyone know what its likely to be?