how hot is too hot

[evanc]

What is considered OK upper temp. for a metal can t0-3 mosfet?

[Coconuts 500]

Over 9000.

[ilimzn]

Not as easy to say as one would expect. Actually, it's the temperature od the silicon crystal that is the most important parameter, and you want to keep that as low as possible. Most manufacturers specify for 100 deg. Celsius as a maximum. How hot the case will be, depends on the thermal resistance between the crystal and the case, which is specified in the device datasheet, and the actual power dissipation. Fir instance, if a 1C/W resistance is spoecified, then if you measure the case temperature as 50C, a maximum of 50W dissipation is alowed before the crystal reaches it's maximum 100C.
On the other hand, while it's desirable to keep things as cool as possible, a common design rule is to keep any exposed heatsinks or cases below 50C to prevent the user from burning fingers and such. Heatsinks will normally have a lower temperature than the cases mounted on them as there is thermal resistance from case to heatink, resulting in a temperature differential, so that needs to be accounted for...

[KatieandDad]

As ilimzn states there are two things to think about. The absolute maximum junction temperature of the silcon junction. This is about 125 Degrees C. The case of the device, I would keep below about 80 Degrees (but then it wont be very reliable), or below 50 Degrees C for reliability.

We use a common analogy. If you can keep your finger on it for ten seconds without it burning then its OK.

[DUG]

General rule of thumb...

Actually rule of finger:

If you can't keep your finger on it, it is too hot.

[DUG]

KatieandDad

did not see your post

[AndrewT]

Running Tj at or above 100degC probably invites disaster, or an unreliable circuit.

You can work back from Tj<100degC, if you use the Rth j-c-s-a parameters and the actual dissipation for the device.

[evanc]

Thanks for the info. The case of my transistors are at 70c. Seems a bit on the high side. Check my thinking on this. The Thermal resistance is listed as 1c/w. I'm running 1 amp at 30 volts so 30w. my junction will be 30c over the case. The max junction temp is listed as 150c. I am well below the max spec. but probably shouldn't up the bias much from here. Thanks as always for taking the time to help someone who has a lot to learn.
Evan

[pergo]

Papa dixit

[DUG]

evanc

Only one problem with your analysis.

As die temperature increases, the power dissipation rating decreases.

At max temp 150C the power dissipation is 0W

What is the MAX Pdiss @ 25C?

Draw a line to 0W @ 150C

Don't go above the line.