Hi Gary
Back in the 70’s the component Quality movement was somewhat in its infancy and the accepted practice was to design and manufacture the devices “as best you were able to” and then test them at the end of the process to establish an “Acceptable Quality Level” or AQL according to various US Standards.
Those production runs with a superior AQL were diverted to the aerospace industry and those with a lesser AQL fed into the industrial or commercial streams at much more competitive pricing.
When we purchased the higher level AQL components, we still had to Qualify them by taking a sample from the batch then run a series of non-destructive and destructive tests to assess if there were any potential rogues in that batch. These rogues usually failed very early on in the testing, following the classical bathtub curve.
We also had a smaller sample of the components from the batch, operating to their full specification in a series of Verification Of Quality (VOQ) tests in the laboratory, to establish their performance over a longer period of time.
Then, the Japanese semiconductor industry with its TQM approach began to make inroads into the aerospace programmes. One anecdotal story is that an aerospace company placed a purchase order for some electrical components with a specified AQL of x% and when they received the goods from Japan, they also found a memo saying that here are the 100 good ones you ordered are here are the x% bad ones, although we are not sure why you would want the bad ones!
This was the first time I came across the concepts of TQM to manage, assure and control the end-to-end organisational processes rather than just manufacturing products and then testing and measuring them at the end to categorise and/or pass/fail them.
These were the days of discrete components; IC’s hadn’t found their way into the design process. There were no plastic packages, only ceramic or metal cans with glass/metal seals for the conductors which had to be tested to ensure there was no outgassing due to cracks in the seals.
We also ran Particle Impact Noise Detection tests to see if any of the EBC connections might become loose (chatter) under launch vibration levels, or if there was any loose debris inside the cans that would then float in zero-g and possibly short out the component.
We still had a typing pool and 10:00am and 3:00pm tea trolley back in those days.
Cheers
Dennis