Dozens of TNY based SMPS blown up: help!
Hi PI Experts,
My company sells SMPS with TNY268GN onboard. One of our clients is using it for magnetic doorlocks and we're a bit confused after receiving dozens of blown up boards.
The circuit looks much like that shown in AN-23, though the snubber is a parallel RC network between DC rail and TNY drain, with a fast diode in the middle, likely to prevent current from flowing through the snubber when the TNY is on. A 275V MOV and a 3A fuse are all the protection before the diode bridge.
Many boards show either the TNY or the clamp diode blown up, MOV cracked, and secondary rectifier blown, and very few input fuses blown.
I'd appreciate any help on finding the causes of failure or, at least, design hints for further circuit upgrades.
Thanks in advance!
评论
Hi,
From your comment, it appears that there have been instances when the MOV was found to have cracked.
The only reason why the MOV would crack is a line surge.
It appears that this design is being subjected to significant line surges with an inadequate surge filter at input which should ideally limit the input voltage as seen by the power supply to a safe value.
It is important to ensure that the PIV (Peak inverse voltage) of the output diode and the clamp diode are not exceeded for the highest input voltage seen by the unit. It is also important to ensure that the highest voltage between the drain and source of the TinySwitch does not exceed 600V during normal operation. (The Tiny Switch will tolerate a higher voltage however it is important to have some safety margin).
You can verify all the above conditions on an oscilloscope.
Is there a possibility that the power supply design that you are using was designed for operation with a voltage as high as 265V but the power supply is being fed a higher voltage from a multi phase system (277V)?
Regards
PI-Sarek
If you are able to measure the kick back voltage on your scope when attached to the power supply, what is it? The use of a 40 V schottky diode on the output could very well be cause of the problem if the solenoid kick back is "in phase" with the reverse voltage seen on the output diode.
What is the reverse voltage on the diode for the design to begin with? If you do not have much margin to the PIV of the MBRS340CT this would only strengthen my hypothesis about failure of the output diode.
If you can also get me any other information about the transformer like the number of primary and secondary turns or at least the ratio that would be helpful.
Let me know what your measurements are and that may confirm or refute the hypothesis. Thanks!
If you could please provide me with a schematic and transformer design it would be helpful in diagnosing the problem.
It sounds like there could be a number of different things going on here. If the device gets destroyed this is due to violation of the SOA of the device. This can happen if the flux in the transformer is not properly reset. This can happen if the secondary rectifier fails. This could also explain the destruction of the clamp diode. It's also possible that the drain waveform is ringing below ground and this can cause the device to be destroyed as
Also you mentioned they are driving magnetic door locks. Is the output of the power supply connected to a solenoid or other inductive element? Depending upon the energy stored, the size of the output caps and other variables it could be possible that this is destroying the secondary rectifier which could lead to further failure.
If you can give me any more details at all under what circumstances these exact failures occur, it would be extremely helpful.
Thank you and I look forward to hearing back from you.