0
\$\begingroup\$

It is good practice to ensure some design margin for the electronic circuit in terms of absolute maximum ratings.

What is the proper way of selecting such a margin for Total Power Dissipation?

Let's assume 20% as a safety margin for any reason and SOT363 NPN Transistor as an example:

On the chart below there is a common power derating curve for BC847 NPN Transistor in SOT363 Package - [Blue].

I calculated the -20% Power alternative [Green] it looks good however the closer we get to the Absolute Maximum Temperature Rating [150C] the smaller the margin becomes. This is why I wonder if it is a proper way of doing this?

As an alternative, I added another line which is basically -10C of the base the value [Green].

Is there a way to define something "in-between"?

enter image description here

\$\endgroup\$
3
  • \$\begingroup\$ I'm not sure that curve is accurate without knowing the PCB thickness, weight, and footprint it's mounted on. \$\endgroup\$
    – DKNguyen
    Feb 28 '19 at 21:28
  • \$\begingroup\$ As well as any margin calculations people advise - note that graphs are usually "typical ratings unless otherwise specified. Look at tabled specs to get some idea of how much typical and max and min values compare. \$\endgroup\$
    – Russell McMahon
    Mar 1 '19 at 1:16
  • \$\begingroup\$ For the Thermal characteristics, the Maximum thermal resistance values are defined, so it should not be worse than what I presented here. \$\endgroup\$
    – KJA
    Mar 1 '19 at 6:18
1
\$\begingroup\$

If you have never done MTBF calculations before there is a general failure rate acceleration factor that often doubles for every 10'C rise (Arhennius Law).

My criteria for a commercial design when I managed this function, was 85'C for junction temperature at maximum ambient of 40'C normally operated at room temp.

  • package thermal resistance, Rjs = 230'C/W from junction to solder point
  • FR4 Rsa= 338'C/W from solder point ambient with free moving air, while sealed air is worse and adding 1 cm^2 exposed copper to collectors reduces by 20'C/W per cm² area of collector copper ²
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.