Chip temperatures spiking up/over 100 degrees killing my devices. Query issue or just bad devices

[Wolfieeewolf]

Describe the bug
Not sure if a bug or not but has anyone else noticed spiking chip temperatures?

I have had a few downlights die as a result of temps hitting over 100 Degrees Celsius. My downlights use BK7231N chips so they do run a little hot but normally sit about 65 to 70. Never really took much notice of temps until the last few days. Could be that my devices have always run hot. Could be the almost 40-degree days we have been having I'm not sure.

I have some BK7231T devices as well that can get up into the 60's which is a bit odd as they normally run cool.er

some of these devices are less than 12 months old so shouldn't be an age issue. They are also a somewhat ok brand. Mirabella Genio and Australia have some pretty good laws to stop most of the crap from being sold so I don't think it would be just a device fault.

I also don't think it has anything to do with the type of LED either. I have downlights, floodlights and PAR38 lights that all show a huge temperature range difference. Some use SM2135, some are just PWM-controlled, and downlights are BP5758D's.

The Stockfirmware runs the downlight super hot. Can't touch them for long after they have been running for a bit. Openbeken runs them slightly cooler and ESPhome makes it so you can at least hold your hand on them for a bit. I did some power adjustments with ESPhome so maybe that helped it as well.

Don't matter if the LED is on or off most devices are running hot. Not sure if this is the normal running temps or if there is something else causing the chips to overheat.

Has anyone else noticed any major temp spikes? What are most of your temps running between? Could it be just sucky heat dispersion on these things? I have most of them inside with an AC set to 24 so wouldn't think it's because my house is hot. The roof space might be getting hot but I do have good insulation and it's not been a problem for over 12 months. Maybe slowly dying over time.

Would be good to see a log system like how the energy meter works but for temps. Maybe I should set this up in home assistant.

Firmware:

• Version [e.g. 1.17.754 - 1.17.783]
• Device? mainly RGB lights
• Chip/model: BK7231T/BK7231N

Edit - All devices have Powersave 1 and NTP. Have turned these off and these is no real difference

[Wolfieeewolf]

So I look at my temp history for the devices that keep dying. looks like something went funny around the 18th of September.

I have looked at all the downlights in that room and they all have a spike on the same day. Just need to work out what happened that day. I'm guessing a firmware update or something. Unfortunately, I don't have the history of the build I updated to but looking at the build around that time there were a few builds on the 18th starting with 1.17.698.

No idea if it is the firmware but seems weird that every device has the weird spiking from that date.

[DeDaMrAzR]

@Wolfieeewolf

I can attest to the similar issue with my downlights that i since replaced but still have them around, will test and report back soon.

[NonPIayerCharacter]

There were some adjustments to temperature calculation formula afaik, so that is probably the reason for the spiking.
If you want your leds to live longer/run cooler - look in docs at current commands (sm2135, bp5758d etc.) and use it to lower the current to leds. For example, default for bp5758d in code is 14mA, but for something like gu10 10-12 mA is safer.