We have some hundred devices with the LM5166 in the field. Somehow on ~10% of the devices the LM5166 suddenly stops working. Only a powercycle can bring it back. The LM5166 works with a current limit of 300mA and has a temperature rise of about 20K. You will find the shematic below.
The pins have the following states while the fault occures:
U BAT= 3,8V (internal lipo battery)
U MainBat_Power= 40,9V
U 4V15=0,6V
Q6A off
Q6B on
SS = 0,365V
ILIM= 0V
SW= constant 1,6V
PGOOD= low
FB=0,36V
Hello Ambreesh,
the small internal battery is drained from system and then disconnects at 3V. Nothing happens in that moment or after that. Only a powercycle at the input of the DC/DC converter recovers the function of the converter. After that, it works maybe without any fault for days.
It is hard to replicate the fault on the boards, because it happens for us randomly. Sometimes the boards work for days and nothing happens, than some boards show the fault on one day.
In the stuck mode all signals are static at their voltage:
U BAT= 3,8V (internal lipo battery, drains slowly within hours)
U MainBat_Power= 40,9V
U 4V15=0,6V
Q6A off
Q6B on
SS = 0,365V
ILIM= 0V
SW= constant 1,6V
PGOOD= low
FB=0,36V
Here are the scope pictures after a powercycle:
SW node
vout at startup
Vout:
Vin:
SS pin
Hi Timothy,
I finally could probe the inductor current over a 120mOhm resistor. You can see it in the pictures below. One with full load, the other with about 5% load.
I have some distortion, because I had to extend the paths to the inductor.
I looked for the residual current and found out, that we have some voltage feedback from another runinng controller. If this controller is off, the output voltage is 0.
What can be the reason, that the LM5166 is shuting down?
