Power resistor in a DDPAK

[LoveLaika]

I'm planning to use a high-power DDPAK resistor in my application because it's the only one that has low resistances and offers a high power rating. I've never used them before, so in this DDPAK configuration, how is it meant to be laid out? You have two terminals for the resistor (I assume), but what is the large pad meant for? Where do you solder it to?

[canadaboy25]

The backplate would just be for mounting/thermal dissipation.  In the features section at the top of the datasheet, it says the backplate is electrically isolated from the resistor element.  So it has no electrical function.

[LoveLaika]

Thanks. I see that now. Given that, I guess it goes on its own isolated copper island, or would it be better to mount it to a ground plane if doing so will help with heat dissipation?

[srb1954]

Thanks. I see that now. Given that, I guess it goes on its own isolated copper island, or would it be better to mount it to a ground plane if doing so will help with heat dissipation?

You will need extensive heatsinking to this resistor if you are hoping to achieve anything like the specified power dissipation. The power rating of 35W is misleading as it is specified at a fixed 25℃ case temperature so effectively can only be achieved with an infinite heatsink or active cooling.

You can calculate a more realistic power rating if you know your actual heatsink thermal resistance, the R_thj thermal resistance of 3.7℃/W and the maximum element temperature of 155℃. For example, if you can somehow attach this package to a discrete heatsink of 10℃/W you would only be able to achieve about 9.5W dissipation at a 25℃ ambient temperature.

The recommended PCB pad alone would be hopelessly inadequate for transferring heat away from the package so you may need to use an aluminium core PCB to spread the heat out, possibly with an additional heatsink affixed to the back of the PCB.

[LoveLaika]

Really, it's only temporary for limiting the inrush current. It's not going to consistently pass a high current load for a long period of time. Looking at my sim, the max current passing through the resistor is approximately 1.7 amps for ~10 microseconds. Given a resistance of 8-ohms, that's a bit over 23 watts, well within the rated 35-watt limit. I wasn't planning to attach it to a heat sink. Worst case, I just use a larger resistor.

[langwadt]

Really, it's only temporary for limiting the inrush current. It's not going to consistently pass a high current load for a long period of time. Looking at my sim, the max current passing through the resistor is approximately 1.7 amps for ~10 microseconds. Given a resistance of 8-ohms, that's a bit over 23 watts, well within the rated 35-watt limit. I wasn't planning to attach it to a heat sink. Worst case, I just use a larger resistor.

at that time scale a heatsink doesn't matter

[LoveLaika]

Glad to hear it. I just wanted to cover my bases. Hopefully that lone unconnected pad will be fine by itself. It's really odd to think of things in terms of such a time scale where milliseconds can be very long or very short, or both depending on how you look at it.

[langwadt]

Glad to hear it. I just wanted to cover my bases. Hopefully that lone unconnected pad will be fine by itself. It's really odd to think of things in terms of such a time scale where milliseconds can be very long or very short, or both depending on how you look at it.

page of the datasheet has an overload rating with time and joules

[LoveLaika]

Thanks for that. The chart was very helpful.