Hi All,
In an amp design I am working on, I am contemplating running some 500mW TO92 devices at around 100mW dissipation, This is more than I am generally comfortable with for these devices, but still well within the manufacturer's derating curve. Given it will be running in a class A amp enclosure, the junctions could reach up to 95degC assuming a very extreme case of 70degC ambient conditions and the 250deg/Watt junction to ambient thermal resistance. I note that such conditions would likely see failure of the output devices anyway but I like to think about extremes.
Anyway, I was interested to know how much this might shorten the component life. I have heard the rule of thumb that lifespan halves for every 10 degrees of temperature increase, but this did not help much as I didn't know the starting point. I did some googling and wasn't able to find much except the attached document. Although it seems to be a specific use case and is for power transistors I thought the below graph I have extracted from it might be a good analogue and of interest to you all. The takeaway for me is that a junction temp of 95 degrees is unlikely to shorten the lifespan to anything I need to worry about (MTBF of ~1000 years!). For the detailed specifics refer to the PDF itself.
If anyone has some additional quantitative data like this please share!
Enjoy!
In an amp design I am working on, I am contemplating running some 500mW TO92 devices at around 100mW dissipation, This is more than I am generally comfortable with for these devices, but still well within the manufacturer's derating curve. Given it will be running in a class A amp enclosure, the junctions could reach up to 95degC assuming a very extreme case of 70degC ambient conditions and the 250deg/Watt junction to ambient thermal resistance. I note that such conditions would likely see failure of the output devices anyway but I like to think about extremes.
Anyway, I was interested to know how much this might shorten the component life. I have heard the rule of thumb that lifespan halves for every 10 degrees of temperature increase, but this did not help much as I didn't know the starting point. I did some googling and wasn't able to find much except the attached document. Although it seems to be a specific use case and is for power transistors I thought the below graph I have extracted from it might be a good analogue and of interest to you all. The takeaway for me is that a junction temp of 95 degrees is unlikely to shorten the lifespan to anything I need to worry about (MTBF of ~1000 years!). For the detailed specifics refer to the PDF itself.
If anyone has some additional quantitative data like this please share!
Enjoy!
Attachments
That's 40 years when you use 25 transistors, 40 years of continuous night-and-day use at 70 degrees Celsius ambient temperature. Still pretty good.
You have about 10 % chance that at least one transistor in a 25-transistor circuit breaks within 4 years of continuous day-and-night use at 70 degrees Celsius ambient temperature. (10 % is a rough approximation, I should really be calculating with products and exponentials.) Then again, hopefully you switch it off when it is not in use and hopefully the ambient temperature is lower most of the time.
You have about 10 % chance that at least one transistor in a 25-transistor circuit breaks within 4 years of continuous day-and-night use at 70 degrees Celsius ambient temperature. (10 % is a rough approximation, I should really be calculating with products and exponentials.) Then again, hopefully you switch it off when it is not in use and hopefully the ambient temperature is lower most of the time.
Last edited:
I agree! As I said, I am not worried. Just posting it for the forum's info.Still pretty good.
Thermal cycling is an important factor - spending 24/7 at elevated temperature is much less agressive than cycling between high and low temperatures every few minutes, due to thermal expansion/contraction movements of the package, die and bond wires (all different thermal expansions).
Automotive ratings for semiconductors focus on reliability, particularly in the face of temperature variations.
Automotive ratings for semiconductors focus on reliability, particularly in the face of temperature variations.