TPS54160A: TPS54160A blowing up after 10 years series production since 2022

Hello,

Can you please share the schematic and layout for review?

Fan_Enable (external on/off 3,3 V) / Fan_Ref (a external resistor) work both fine and are not the problem.

Other schematics without this two elements show the same behavior.

Hello Dieter,

A few observations I had from the review

Schematic:

1. Input capacitance derating is substantial. With 50Vin (nominal) and 50VDC capacitance, the effective capacitance is very small. There is a similiar situation with the output capacitnace

2. The catch diode on SW current raiting is below the DC rating of the device. This is low.

Q1: Is the FB node being driven to modulate the output voltage? It seems like it is being routed far away from the device.

Q2:  How is the device protected from seeing > 60V on the input? (I am assuming some solerance on the 50V input.)

Q3: What is the part number of the buck inductor?

Layout

1. The comp components are far away from the device. These components are critical to be placed close.

2. FB routing appears to be large

Q4: Is there an inner layer ground poor? If not, the input loop is large.

Hi Marshall,

thanks for your review.

A1: Yes ist is adjustable, but also not adjustable/static layout makes same problems.

A2: Input Voltage is always Output of a Voltage regulator (e.g. IRM-60-24 : MEAN WELL). We cannot generate transients above 55 V whatever we do

A3: Bourns SRN8040-101M

The capacitance values, inductors and also the diode is used since 10 yeras never having problems like that.

Hello Dieter,

You are welcome

Those would be my recommendations based on things I saw deviating from the datasheet.

In addition, the inductor part number seems to show a low saturation current (1A) in comparison to the recommendation of it being equal or greater than the peak current limit of the device (2.7A).

I would see if any recommendations are possible to implement and trying to capture the failure on an oscilloscope.

Traces: Vin, Vout, Vph

Ok, I'll try this. Thanks!

Thank you.

Hi Marshal,

I can reproduce blow up with a switch that chatters or by step-changing Input Capacitance (see capture) during normal operation.

Please see the captured Shot of a failure:

green = in / red = out / blue = vph

The regulator survives with a TVS diode placed at the input.

But how ca this happen integrated in a circuit behind the output of a power supply like IRM-60-24 from MEAN WELL?
Do you see any other possibility to reporoduce such kind of failure?

Didi

Hello Didi,

Seems like that path is rather inductive, so you are seeing an inductive ring with the hot-plug test you are preforming.

A TVS seems to be potential solution.

Additionally, you can try a damping capacitor on the input that can be selected with the aid of the below resource:

www.ti.com/.../snva538.pdf