3W power resistor at 100degC, 8 hours per night, 365days per year

Hello,

We have an offline 60W Flyback SMPS. The primary clamp resistor is dissipating 450mW.
This resistor is a 3W rated ROX3SJ39K.
The power supply will be at full power 8 hours through a night, in Europe.
The power supply simply turns on/off once per day.
A thermocouple (k type) attached to its body showed a temperature of 100degC, whilst the internal ambient in the power supply enclosure was 50degC. (external temp was 25degc)

Despite the fact that the datasheet appears to show that this is fine, we cannot believe that such a resistor will last for many years.
Do you agree that the part won’t last for years?

I have requested failure rate details from TE connectivity, but didn’t receive any.

Datasheet 1 below shows resistance change data which shows the power being repetitively pulsed on/off in the resistor with a 2 hour period……..does this suggest that rather than the actual resistor temperature, it’s the resistor power cycling (temperature change) that causes the most problems?

ROX3SJ39K datasheet 1
https://www.te.com/commerce/Documen...rtrv&DocNm=1773271&DocType=DS&DocLang=English

ROX3SJ39K datasheet 2
https://www.farnell.com/datasheets/...MI4LOA0c_P3wIVWflRCh2r2A9_EAAYASAAEgJgR_D_BwE

Hi,

reduced air flow?

And: do you have the 7mm space between PCB and R?

Klaus

Ask yourself the question:
"Why should a hot resistor have a shorter lifetime?"

There is no reason, an electrolytic capacitor has a fluid content that will eventually leak to the environment so some longevity is predictable but resistors are dry solid things. The only short term (meaning decades) failure is caused by mechanical stress during expansion and contraction. They actually like being hot because it keeps moisture and therefore corrosion at bay. The number of temperature cycles rather than the absolute temperature will be your problem but with x6 over-power rating the rate of temperature change should be relatively slow and not 'shock' it too much. A year would only be at most 732 temperature cycles but you should expect it to last for many thousands before failing. The 'Load life' figures in the data sheet are for 1.5 hours on and 0.5 hours off which you can assume allows full heating and cooling, you can treat those as one daily cycle but bear in mind you are not running at 3W/0W but 0.45W/0W so the figures should be much better than shown.

Brian.

Very high temperatures are generally demaging not to the resistor, but to many parts of the board, particularly those made of polymers, such as cables for example, becoming brittle. In the case of devices that radiate an extreme amount of heat concentrated in a small volume, precautions must be taken such as for example forced air circulation away from the other components, and of course keep it as far from other devices as possible. I have seen for example an industrial camera often showing internal temperatures as high as 70oC degrees whose flat cable had a slight discoloration near the chipset, the source of the heat.

100deg C is about the limit for the solder for 10-15 years ... the resistor will be fine at this temp.

reduced air flow?
Thanks, yes, the PCB that holds the resistor is inside a plastic enclosure with no vents, except for a tiny gap round the connector, and there is no fan inside.
And: do you have the 7mm space between PCB and R?

Click to expand...

Thanks but no unfortunately not….this resistor is mounted vertically on the PCB, because of lack of space…so we will have to have silicone squeezed around it to stop its body melting into the adjacent X2 capacitor bodies. We haven’t used one of those plastic vertical mount resistor holders, due to cost.

100deg C is about the limit for the solder for 10-15 years ... the resistor will be fine at this temp.

Click to expand...

. Thanks, I am wondering if we need some copper pour around the PTH pads so that the solder can run a bit cooler?

yup, the more copper pad etc the better ... also check your resistor, steel leads vs Cu leads ...

I do not consider Metal oxide film and ceramic body as sensitive materials when operating at 100C.
And I am of the ones which they did read product datasheet before delivering my opinion.

- - - Updated - - -

treez said:

Thanks but no unfortunately not….this resistor is mounted vertically on the PCB, because of lack of space…so we will have to have silicone squeezed around it to stop its body melting into the adjacent X2 capacitor bodies. We haven’t used one of those plastic vertical mount resistor holders, due to cost.

. Thanks, I am wondering if we need some copper pour around the PTH pads so that the solder can run a bit cooler?

Click to expand...

You could get better answers if you post product pictures.