Are FlexMod P3's still available or is there an equivalent that is ready to ship?

[Draco]

I do the redneck calculation of heat sink size, use the biggest one that still fits, lol and if that fails i used a larger mosfet. or pop a TEC on it, joking of course

[neilhooper]

Increasing the heat dissipation also gives a negative C/W. What is surprising, the results of actual measurements have nothing to do with the results of this formula. I am inclined to think that Øjc value has no practical application because in the datasheet is not shown actual value, max only.

I'm interested in what measurements you are able to take as the formula only works on the junction temperature within the chip which I would imagine is very hard, if not impossible, to measure outside of a lab. In my example the case temperature is likely to be approximately 84ºC with a heat sink temperature of 82.5ºC while the junction temperature is 138ºC.

Whether the Øjc value has a practical application or not is a matter of opinion. I am only pointing out, from my research, the industry standard calculation for heat sink sizing.

If the formula is re-arranged to

Tj = Pd(Øjc + Øcs + Øsa) + Ta

This allows you to insert the heat sink value that you may want to use to see if it is possible to use it.

In my example above

Tj = 15*(3.6 + 0.1 + 3.5) + 30 = 15*(7.2) + 30 = 138ºC

Meaning my junction temperature should be within limits.

However if I were to use this MOSFET, http://www.farnell.com/datasheets/136333.pdf which has a Øjc of 0.75 and possibly a better candidate as it a lower RDSon and very similar characteristics, although it's not my design, then suddenly my junction temperature drops to 95ºC which is more acceptable with an approximate case temperature of 84ºC and heat sink temperature of 82.5ºC. Notice that the case and heat sink temperature remain the same but the junction temperature is significantly lower; this is why the Øjc value is important.

Of course all of this is only theory and actual measurements may be different.

- Neil