In general:
Once upon a time in my life, professionally, it was my job as part of a
team to review engineering models & prototypes for the ability to be
manufactured and used reliably by the end user. Later it turned to
quality acceptance and quality assurance. I seem to recall something
that was referred to as "rule of 10s". It went like this"
If a $1.00 part was designed into a piece of equipment and the part was
defective, the projected cost of $10.00 was required to replace it with
a new part. If the $1.00 part was allowed to get through production and
end up in a final product it then would cost $100.00 to replace it. (One
must consider all the unpacking, re-packing, re-test of the production
lot and etc. Then if the $1.00 part (from engineering) went all the way
to the end user, it cost $1000.00 to replace it at the end user site.
(Figure the cost of travel, lodging, meals and the Field Service
Engineer to accomplish this task.)
In the end, a poor design or poor selection of parts can get very very
expensive not to mention a very unhappy end user.
An yes, if the user can figure out a way to operate something
incorrectly thus causing a failure....... they always will. I really
don't think it's possible to "totally idiot proof" a design. I'm told
that there are those that can damage an anvil just carrying it across
the street. :-)
73
Bob K4TAX
Peter Chadwick wrote:
>
> Tom says;
>
> >By far the biggest stress in amateur service is the
> > constant thermal cycling of the filament
>
> That's where the rheostat comes in handy - assuming proper proportions, it
> limits surge current.
>
> Of course, so does a suitably designed surge limiting filament transformer,
> but
> that can be a bit more expensive.
>
> >provisions would have to be made for monitoring voltage
> >accurately
>
> I assumed external metering, but point taken, especially on cost.
>
> Of course, there's the BBC(?) experience with the 4CX250B, showing that some
> filament voltage reduction can help life. GW4FRX has chapter and verse on
> that,
> and, I seem to remember, has published the info here on occasion. Whether, in
> amateur service, you'd notice the difference is another matter......
>
> The real cost of a component in a piece of gear can be quite high. I even
> convinced purchasing on one occasion that a 22p cermet trim pot was cheaper
> than
> a 3p carbon one, because 15% of the carbon ones broke or otherwise failed in
> the
> test department - and it cost just 8 times as much as the component cost of
> the
> cermet to replace the failed part. Even adjusting a preset in the factory
> costs
> a lot of money.
>
> 73
>
> Peter G3RZP
>
> --
> FAQ on WWW: http://www.contesting.com/FAQ/amps
> Submissions: amps@contesting.com
> Administrative requests: amps-REQUEST@contesting.com
> Problems: owner-amps@contesting.com
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|