114 years of *Mean* Time Between Failure? One can only wonder how they tested that.

Reliability engineering, you can change some parameters and multiply the lifetime by the corresponding factors.

I don't know these statistical factors but lets say 100 of your electronic devices are put in a hot chamber of 90 degrees C and a humidity of 80% and the average of those devices last 10 months than with those statistical factors you could for instance multiply that by 50 and get an average expected lifetime of 41 years.

And, since a VFD is a vacuum tube, they laugh at 90C! So, the life at 90C is the same as 20C. Apply the Boltzman curve, and you extrapolate a crazy MTBF. Yes, it is all a house of mirrors.

My favorite was the tantalum capacitor debacle. A well-known failure mechanism of tantalum capacitors is you run a device for a while, put it on a shelf for 2-5 years, turn it back on and several tantalum caps fail shorted when the DC power abruptly turns on. So, the US military specified tantalum caps for a wide range of gear, severe accelerated life testing showed great results, and then they built thousands of radios, radars, computers and such, and after a good test run at the factory, it all went on the shelf at the supply depot! Perfect recipe for this problem.

But, if you know the system for doing this MTBF estimation, it is so TOTALLY bogus. You count the number of each type of component, look up the MTBF for each type, and then combine it all with a formula, and there's your MTBF number! This is all done with a spreadsheet, not with any sort of testing of the complete device. It totally ignores such factors as the load a component carries, the duty cycle of the load, thermal cycling, ESD at external connections, physical handling, and many other issues. But, this bogus system produces totally WHACKY numbers like a whole military field radio with 100+ years MTBF. Oh, RIGHT!

(Now, military and aerospace gear IS tested rigorously, with heat, cold, thermal cycling, vibration, salt spray, humidity, and on and on, and that weeds out marginal parts and designs. I'm not badmouthing that. Just the MTBF calculation craze that DOD and others went through some years ago. They still insist on it, but I think they have come to realize it isn't as predictive as they had hoped.)

Jon