Question with Thermal Shutdown and Junction Temperature

Question

I'm using this Battery Charger IC - MCP73831.

So, I am using a SOT-23 package.

My input is 5V, the output Vbat charge regulated voltage on the Vbat pin is 4.2V and the maximum charge current is 340mA.

So, by calculation of power dissipation as per section 6.1.1.3 on page 18, it says that, for my condition,

Power Dissipation = (5V - 2.7V) * 0.34A = 782mW

Maximum junction temperature = (0.782 * 230degC/W) + 25deC (ambient) = 204degC approx.

But on page 3, absolute maximum ratings for maximum junction temperature are given as "Internally limited".

My questions:

1. What does "Internally limited" mean for junction temperature?

2. While searching for this answer, I searched for thermal shutdown and found section 4.10 on page 14 which says that the device will go to thermal shutdown if the die temperature exceeds 150degC.

So, according to my calculation, the junction temperature (204degC approx) exceeds the 150degC. So, for my application, when the die temperature or that is, the junction temperature comes close to 150degC and exceeds it, the device will go into thermal shutdown and my device would start to cool. Only if the temperature goes 10degC than 150degC (as mentioned in section 4.10), the device will start to work again and goes in a loop is it?

So, the device will always be stuck in this cycle and will never fail due to excessive junction temperature which is why they mentioned that the junction temperature will be internally limited, is it?

Answer 1

You are correct: "internally limited" in the absolute maximum ratings means that the device will shut itself down before permanent damage can occur so you don't have to worry about it (though I'm about to tell you that you should). The device should not be damaged by thermal shutdown but designing the circuit so that it is unavoidable it is not a great idea since device lifetime is greatly shortened at higher temperatures. Not to mention that a battery charger that spends a good portion of its time (probably more than half) in thermal shutdown isn't going to be very good at its job. Note 2 states that with a large copper area, you can achieve better than 130 C/W which puts you at ~125 C which is high but not high enough to shutdown the device. Specific layout notes are shown in section 6.4. I would recommend following them.

Even though the equation that gave you 782mW is worst case heat generation during the highest-stress part of the charge, I would take a look at a device designed to handle higher currents so you can run at a lower die temperature. Maybe something in e.g. a SOIC-8 EP or with an external pass element.