At 03:50 PM 7/25/01 +1000, you wrote: > I'm going to be picky with your fet calcs Hi, Roman;- Sure! > remember 65'C/W is for free standing >TO220 pack with no parts around it, and a room temp PCB. For >most situations, especially if in a small enclosed box I would >derate that to say 85'C/W. Well, I was talking the internal Ta, but your box would have to be pretty small (and sealed) to get that hot inside. >Assuming 0.035 ohm Rds, at 5A is 0.875W. So maybe 75'C rise >will be seen over ambient, FET will get to about 100'C. Really worst case, and that is junction temperature, vs. rated 175'C. Remember typically the FET will be more like 0.025 ohms or so with a drive from 5V. >Anything over 55'C total temp at the transistor will cause >failures, probably within months as the solder weakens and >goes through enough heat/cool cycles to get circle fractures. Yes, thermal cycling can kill the die eutectic bond. >This is the stuff I fix all the time in TVs, and having the >infrared thermometer now it's a big eyeopener, a TV with tran >at 50'C runs for years, models with transistors at even 65'C >fail repeatedly after a few months. With some models i've >been clamping bigger heatsinks on, or using better (lowsat) >parts, having to re-engineer their crappy product so the >poor customer gets more than 6 months MTBF. :o) > >I suggest 0.3W max from an un-heatsunk TO220 pack. That is extremely conservative by most people's standards, I think my 600mW for industrial environments (up to 60'C behind the panel) is pretty conservative for 24/7 operation, and it's been shown to be very much ok in some high volume products (> 10 years life 24/7). But I can't argue with being on the safe side (unless I have to pay for the safety factor.. ) There are some TO-92 (!) packages that are sorta ok for consumer applications at more power than that. eg. the 8550, not so conservatively rated at 1W at Ta=25'C (normally, it's 360mW or so). Not that I'd make crap like that, but if you look inside those $10 amplified speakers.. >One last point, FETs made for 5v drive, you won't get 5v >from a PIC output pin. PIC pins source 4.2v typical, down >to 3.5v once you are sourcing 10mA or more. The MOSFET gate draws 0.0mA, the gate-source resistor I suggested draws 500 uA. So, the output will be darn close to Vdd (150mV or so) With darlingtons, transistors etc. you have to pay attention to this. Fortunately the drive goes up at high temperatures where the transistor beta is also higher. For the MOSFET driver, your suggested pull-up resistor is usually a bad idea because it will cause the solenoid to come on if the PIC output is high impedance for any reason (eg. during startup), or if the PIC is unplugged. That's why I suggest a pull-DOWN resistor. The MOSFET can also be damaged if it drifts into the linear region. >Sorry to be picky, but newbies especially should be VERY >conservative with all their ratings. The expert engineers >can start to push the limits... (but then a year later i'll >be fixing their product!) ;o) Not a chance! Best regards, =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Spehro Pefhany --"it's the network..." "The Journey is the reward" speff@interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com Contributions invited->The AVR-gcc FAQ is at: http://www.bluecollarlinux.com =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= -- http://www.piclist.com hint: To leave the PICList mailto:piclist-unsubscribe-request@mitvma.mit.edu