[Elecraft] KLPA3A power transistor thermal stress concerns

[Don Wilhelm]

Bill is entirely correct.
Once in my prior career, I did a study of Mean Time Between Failure for 
IC circuits (same applies to transistors).  That study had to consider 
the case temperature of the IC (or transistor) with respect to the 
junction temperature.
The specification for the IC or transistor package will include a 
thermal resistance parameter, which must be considered to obtain the 
maximum temperature for the case of the device.

Bill's comment about the screw tightness is entirely correct - the heat 
of the device case must be transferred to the heatsink in an efficient 
manner.

73,
Don W3FPR

On 10/27/2017 6:07 PM, Nr4c wrote:
> Yes. But I believe that is Junction temperature, not case on heatsink temp.  Remember the case bottom is heatsink for those. Make sure screws are tight.
>
> Sent from my iPhone
> ...nr4c. bill
>
>
>> On Oct 27, 2017, at 12:19 PM, N2TK, Tony <tony.kaz at verizon.net> wrote:
>>
>> 50C is not hot for a transistor. Commercial transistors are usually speced to 85C. JANS (space) transistors are speced to 125C and actually baked without bias for 320 hours at 200C. The hFE is measured before and after to check for change.
>>