Electrical Engineering Stack Exchange is a question and answer site for electronics and electrical engineering professionals, students, and enthusiasts. It only takes a minute to sign up.
Sign up to join this community
I'm currently working on my first circuit that dissipates enough power that I need to use a heatsink. The component in question is a TS1085CZ33 which I'm running with a maximum load of 2.8A dissipating around 7.5W. I'm just wondering a couple of things:
I think if you used an up-to-date spec you'll see the above in red - this spec has stuff about RoHS compliance and the device being lead-free so it's probably more recent than the one in the OP link.
That datasheet doesn't seem to specifically mention it, but many positive voltage regulators have the tab tied to the output. It's certainly something you'd either want to test with a multimeter or insulate it. There are TO-220 insulator kits available that consist of a mica insulator and plastic spacers for that purpose if you're worried about it coming into contact with something else.
There is no technical reason to ground the heatsink, but you will see many designs made that way when they are mounted in a metal enclosure just for a combination of mounting convenience and so that the enclosure dissipates some additional heat. Certainly don't go out of your way to ground it otherwise.
Update: Edited to incorporate new information from an updated datasheet.
The datasheet for the device makes no specific representation of the heat sink tab, as to whether it is internally connected to anything or not. The newer datasheet for the component specifically mentions that the heat sink tab is connected to Pin 2, i.e. Output. Many voltage regulators do have their heat sink tab connected to the ground or output pin, but this is not guaranteed.
The following generic approach still applies:
In such a situation, one would simply use a multimeter between the heat sink tab and each of the leads, to confirm whether they are internally bonded. To avoid a false positive due to some internal diode junction, one would use both resistance readings, and diode readings in both directions, for each of the 3 leads.
This will establish whether the heat sink is internally connected at all.
Whether the tab is connected to anything or not, as a general practice one would not need to ground the heat sink - it is simpler to leave it not grounded. That way, accidental contact with a test probe or some wire connected to a powered part of the circuit would do no harm.
