LNK306 IC Heating
Dear PI Expertise,
We are using LNK306 for generating 12V,50mA source. Though we have designed schematic exactly as suggested in datasheet and even our load doesn't increase more than 30-40mA, I see that the IC is getting heated up and entering thermal shut-down mode.
I am anticipating that this may be because we have not followed the layout guidelines, particularly the big copper area suggested for source pin.
So I want to make sure before redesigning the PCB that, this IC heating can be controlled my putting a big copper area for Source pin and following other PCB guidelines.
Please give your expert advice on solving this issue and it will be helpful for future designs if you could also share the exact reason behind this.
Awaiting your earliest reply.
Regards,
Arif Gilani
댓글
Firstly, sorry for posting in PIExpert Software. I am new in PI community.
I will check with LNK304. Also will try to analyse every other points that you have mentioned.
Thank you!
Regards,
Arif Gilani
Hi Arif,
FYI, this section of the forum is usually for PIExpert Software questions.
I don't really have all the information so I'm going to try and guess a bit (and assume you are either using buck or buck-boost). There are a few possible reasons, so I'm going to list the ones I can think about now, mixed with some general advice. For that output current level your main cause for losses inside the IC is going to switching losses (by far)
1 - Can you use a smaller LNK-TN (like the 302 or 304)? They have higher RDSon but lower parasitic capacitance, and therefore lower switching losses. From experience I can tell you that a bigger mosfet is not alway better for efficiency/thermal. On top of that you will be able to use an inductor with lower saturation current (smaller and cheaper), and EMI will be easier to tackle.
2 - Is your inductor saturating? The peak drain current of the LNK306 is high, so if it is indeed saturating the drain current will increase further (unless it is a severe it may not destroy the IC, but it will cause it to heat up).
3 - Are you using a big enough inductance? We usually recommend at least 470uH. A bigger inductor will have lower di/dt and therefore will slightly reduce the peak drain current (check Figure 13 of the LNK-TN datasheet). this will reduce the switching losses at the IC.
4 - Check that your freewheeling diode is indeed ultrafast (max 75ns trr).
5 - I don't know your max ambient temperature or your max input voltage, but assuming 265VAC, 50C max at board level, and the D pakcage (SO-8), I guess you could survive with a copper area around 3 times the size of the IC (at least with a 302/304). Keep in mind this is a guess though, you'll have to check. If you have a board and you want to test quickly you can always solder a bit of copper directly onto the source pins (once I did it by stripping a few wires, single strand of copper).
I hope this helps you a bit. Let me know if you have additional questions