Temperature rise on LNK306
HI PI
We are using the LNK306 in the buck topology. The output power is around 0.5W with a peak load of <1W.
In normal operation everthings works fine. Also conducted emmission are OK. But we have some problems with the fast transient burst tests (IEC 61000-4-4). The microcontroller system which is supplied by the LNK306 is reseting. But what worries us more is that the inductor and LNK306 gets very hot after 1-2 minute and the LNK306 stops working because of its over temperature protection.
When a constant load of 1W at normal operation neither the inductor nor the LNK306 get warm.
We assume that the supply is somehow not working properly which causes the resets of the microcontroller system and temperature rise the inductor and LNK306.
Hope you can give some hints about that.
Comments
Thanks PI. When I use a inductor that has higher current rating, the resets on the microcontroller system disappear. Seems like the disturbence of the burst don't reach the microcontroller system anymore.
But now the over temperature protection of the LNK306 is already reacting after around 1min while bursts are applied to the supply. But only with during burst tests. During normal operation the LNK306 works fine.
I can not use a full wave bridge because the L and GND need to be on the same potential.
Do you have any clue what causes the temperature rise in the LNK306?
You indicate that the power supply does not overheat with a normal 1 Watt load. With a 1 Watt burst mode, the LNK306 overheats. Please describe the burst mode test.
Z. Cochrane
The load is 0.5W in normal operation with 1W peak load. No problem at all during normal operation with this set-up, no reseting of microcontroller system no oberheating of the LNK.
The fast transient burst test is accoring to IEC 61000-4-4, see attachement of the transient burst waveform. These disturbencies are coupled into the connected hot and neutral wires. During this test the load is 0.5W. After using a inductor with higher saturation current, the reseting of the connected microcontroller system disapeared. But the over-temperature protection of the LNK306 acts after around 1 min.
This is a 2000 Volt transient at a high repetition rate. Running this test for over a minute will thermally stress the device. If you need to run this test for such a long period, then you will need to clamp the transient using a resistor in conjunction with a MOV or TVS device.
Z. Cochrane
It sounds like the output inductor is saturating. You need to use a inductor that is rated for 0.515 Amp. You can verify this by measuring the output current with a scope/current probe. When the LNK306 is conducting, the inductor current will ramp up. If the ramp begins to curve up, then the inductor is saturating.
Try using a larger output capacitor to solve the fast transient burst test. If this doesn't solve the problem then try using a full wave bridge rectifier at the input.
Z. Cochrane