I am using the ISL8026A chip for voltage regulation from 5 to 3.3V. The datasheet is – datasheet.
Now, this is merely an example circuit. Now, I would like to design a power supply for a hi reliablity application. As part of this I presented a schematic as shown in the datasheet. This is as shown below –
.
Now, as part of this design I presented the schematic above. As shown I am having 2 22uF capacitors for filtering out ripples from the power cable. Now, one of the questions I was asked is ,since this is a high reliability application, I must add redundancy management in this circuit. For eg – the input capacitors. If they are shorted due to some discrepancies the whole system can be destroyed thereby causing critical failure.
This is just an example. There could be many such redundancy scenarios. How can I be aware of all such points where I will be needing redundancy checks ? Is there a checklist for such things ?
This is just an example of a redundancy check on an SMPS. What are the mechanisms deployed on other monolithic and MMIC ?
any pointers would be appreciated.
I agree it could be quite broad. But even pointers on the smps will be deeply appreciated.
Best Answer
It's impossible to do redundancy management at the microscopic level and this includes deciding whether a 22uF capacitor is working or not - to work at this level is impractical and requires more components than the basic design making reliability even worse!
So you make a "module" design and you determine its reliability (possibly by using MIL-HDBK-217F) and if it needs to be more reliable you engineer two "modules" in parallel - input power and output power are common to both and you'll need a switching scheme to ensure that if one fails the other one takes over. You can nearly double the MTTF of the overall module in this way.
Large passenger aircraft can still fly if one engine has failed and this is the idea behind redundant systems.