If I am dealing with a simple conductor that follows Ohm's law I can take the dissipated electrical power for a DC current running through the conductor as P = U*I.
However, I think that this should be invalid for a semiconductor as they are not really following a linear dependence between U, R and I.
Is there an equation to calculate what the dissipated power of an DC current through a semiconductor should be? I was not able to find a reasonable answer via google.
thanks for your help!
Always is valid p (t) = u (t) .i (t) - from basic definitions of variables. See for example: Research The shortest way to OPA?
p(t)= u (t)*i (t) is valid for any "black box", it is the definition of power, but you should be careful with averaging. The term "DC current" in practice might have different meanings... if it is averaged , then you can not use it... . If the current is indeed constant for a while, then you still need to define u(t), here all non-linearity, time dependencies and opportunity to present the the Elements of the circuit as lumpt elements would be important.
My experience is with FETs. For a given device, on state resistance varies with temperature, current, and gate bias. These are characterized over several graphs on a datasheet. Physically, they are a combination of terminal, contact, channel, and drift regions which all have different characteristics.
Switching losses per cycle are switching time (based on gate drive and gate charge), commutation voltage and commutation current.
Reverse recovery and output capacitance losses per cycle are based on Eoss and Qrr. Qrr is usually poorly specified as it is highly dependent on current, temperature, and di/dt.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Materials
- Semiconductor