At 12:15 AM 1/7/2014, you wrote: >On Mon, Jan 6, 2014 at 8:58 PM, Sergey Dryga wrote: > > If PIC is driving the gate, then make sure the MOSFET is "logic" MOSFET= , so > > that 5V completely turns it "on". Your device (STP36NF06L) does not sw= itch > > completely "on" at Vgs=3D5V. This will lead to reduced current (Id) an= d > > higher thermal dissipation. Look at Output characteristics (Figure 3 i= n > > datasheet) and Transfer characteristics (Fig 4). > >Ah, good point. I am open to suggestions if you've got a better >option! I haven't ordered any parts yet. > >Josh Hi, Josh:- IMHO, the best thing is to look for guaranteed Rds(on) at your minimum drive voltage or lower, AND at the maximum junction temperature rating (Rds(on) may increase by >50% at Tjmax). Calculate the temperature=20 rise from this (I^2*R, theta junction-ambient) and verify it's okay at worst-case ambient. BTW, it's not a terrible idea to put a relatively high value pulldown resistor from the gate to source, especially during development. Otherwise, if the driver goes high-Z for some reason, the gate voltage could wander around into 'undesirable' regions. Something like 100K will often suffice. Don't (unnecessarily) drive the gate too hard- it may be a good tradeoff to have more switching dissipation and less EMI (keep within the SOA =3D Safe Operating Area) curve though). If your PWM is at 100Hz, you don't need to drive it like a 100kHz PWM. Cheaper too, if you can avoid drivers. Unlike BJTs, there is a real cost to having a much higher Vds rating than you need, so don't go nuts on the voltage rating (or cut it too close!). --sp --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .