ACWWong wrote on Jun 7^th, 2006, 5:21am:


My understanding is that mobility is increasing (as temp goes up) at low temperatures because the ions will have less effect on a hotter carrier.  Then mobility decreases because the lattice vibrates and there are collisions.  So the curve of mobility vs. temp is an inverted 'V'.

The Vth change is -0.6 mV/K but mobility goes down with T^(-1.5), so yes, mobility is going to really affect your circuit speed more.  Plus Ids is a linear function of Vgs-Vth with short-channel devices, so the temp has less of an effect on current via Vth than for square-law devices.

Cheers,
Marc

P.S. I am using the handle 'Croaker' now.  It comes from the 'Black Company' books.