experience with zener reliability?

[vthielen]

I need to (try-to) design electronics for 30-years lifespan continuous operation.
I had some trouble with zeners that failed in a couple of years time, even though they where only used at about 30% of specified power ratings at room temperature.

Does anybody else have similar experiance?
Might this have been down to a bad batch?
Does it help to chose a zener of higher power specification, or does this make it worse (larger surface so bigger chance of defects)?

The practical problem is that I need a low-power (50mW) 12V supply (for MCU) in a 220VAC appliance. I have 2 options:
1) using a commercial switching power PCB module supply like SPAC265FC
2) using a capacitor to get a small current from the 220VAC, and run that through a zener.

which would be the most reliable option?

 

[Krithik]

Hi,

Design with Zener for MCU Supply is correct and not preferable.

Because MCU having little bit tolerance +/-1% only.

But in now a days we cant believe the zeners. Regularly we can get 12.1V zener from market. But it will have the tolerance of minimum 5%.

And more over the zener tolerance will change with the life time depending up on the power dissipation.

According to my previous experience with zeners it is not suitable for MCU.

And no need to put that type of large modules(SPAC265FC).

If you want i will give you the compleet solution for your design with proto sample.

And also if you have the quantity i can support from my side.

I have own indipendent company. so, i can give you the best price as compared to DIGIKEY and others..

I

 

[chuckey]

If you want a 30+ years life then put in extra mains filtering, even though the type X and Y capacitors tend to blow up. this is because in most locations the mains suffers from high voltage pulses and even lightning strikes. Is this the reason why zeners seem to be unreliable in cheap power supplies? Also electrolytic capacitors are unreliable especially when run at elevated temperatures.
Frank

 

[ALERTLINKS]

It is the quality of components which is also to be considered. Use componenets from well known manufacturers. Specially avoid fake Chinese components.

 

[schmitt trigger]

In addition to everything else that has been said.......The key for ultra long service life is redundancy.

 

[luis_mendes]

What about Texas Instruments TL431A or Zetex Precision-Shunt regulator capable of supplying up to 100mA.
TI
RS: 527-306
Ref fab. TL431ACDBZTG4

or
Zetex
RS: 738-5333
Ref fab. TL431ASA-7

 

[FvM]

What about Texas Instruments TL431A or Zetex Precision-Shunt regulator capable of supplying up to 100mA

This is probably an answer to post #2, but not actually referring to the original problem, which is about long-term reliability respectively unexpected component failure. There's no indication that IC regulators have higher reliability than basic Z-diodes, we would rather expect the opposite.

To refer to the original question, if we observe medium time failure at 30% rated power, these possible explanations might be considered
- occasional severe overload, e.g. due to inrush current or surge voltages. This is particularly a problem of capacitive power supplies without sufficient protective series resistor
- permanent temperature changes due to intermitted operation with high on-off rate
- bad component quality
-