Analysis of the TVS Z1 Failures in a Power Supply using the TOP242G in a VFD Application
Overview: Our companyis a satisfied user of the TOP242G power supply switcher in variable frequency motor controllers since 2011 using upwards of 40k/year with very few failures of the power supply.
One customer has experienced multiple failures of the primary clamp zener Z1 in the TOP242G circuit shown below.
Our analysis of a few failed returns has not been conclusive in determining the root cause of these failures.
We are requesting technical support in diagnosing the root cause of the Z1 failure and suggested corrective actions.
Detail:
On three returns:
Z1 failed to a short. Z1 is a TVS rated 200v, 600w_pk, OnSemi part number P6SMB200AT3G, our part number C72200.
This customer’s application is powered continuously 24/7. The failures occur after four or five years in operation.
Pin 5 of the TOP242G was not shorted. It measured in the range of 300kR.
The transformer from one failed return was removed and tested. It passed Hi-pot primary to secondary at the same voltage as a new transformer. This transformer functioned OK when put into a known good circuit board.
The power supply schematic and waveforms are included in the attached file.
Comments
For your statement "Make sure that the reflected voltage do not exceed the rating of the TVS.":
Is it OK for the Z1 TVS to conduct every cycle as long as the operation is well within the Z1 thermal rating? I believe that the Z1 conducts for 300nsec out of the 7.7usec. cycle as highlighted by the circled part of the waveform. This conduction is to clamp the overshoot from the transformer leakage inductance. The Z1 TVS appears to be well within its thermal rating measuring a body temperature of 79c in a 25c ambient. Should the Z1 TVS clamp voltage be increased to prevent this short conduction time?
If this is the case then the rating of the TVS should be fine.
I noticed that for a low line input AC your bulk voltage can reach up to 400Vdc. When you say VFD application do you mean this as a variable frequency motor drive? Motors have a lot of voltage surges which could affect yor power supply. Were tests done to simulate any kind of high voltage surge done on the power supply?
Could you send me the PIexpert file you used in designing the transformer?
Hi,
Thank you for choosing Power Intagrations.
The voltage acorss the primary is the most possible reason for the TVS failure.
The reflected voltage which is determined by your transformer turns ratio is impressed across the primary winding when the internal MOSFET of the TOP242 is off. Make sure that the reflected voltage do not exceed the rating of the TVS.
The voltage across the primary when the internal MOSFET is on is equivalent to the input voltage. Lightning surge can cause the voltage across the primary winding to increase when the MOSFET is on. Was the power supply tested for lightning surge? You might need to put a varistor across the input if this is the cause of failure.
You also mentioned that the power supply is continuously on 24/7. And the failures occur after 4 or 5 years. From your schematic I can see that you have R19 and R2 along your output rails.
An open circuit on your secondary could result in a very high voltagae across the primary winding which could destroy the TVS. Check if any of the devices along your secondary has failed open especially the resistors.