1 Jan
2006
1 Jan
'06
11:56 a.m.
On 1/1/2006 at 2:19 PM William Donzelli wrote:
Anyone that has even a shred of experience on a factory
floor would take
the more complex IC, maybe even more pricey, over even a small pile of
discretes, SMT or thruhole.
Maybe, but consider the AT&T business model--there was a lot more to it
than production floor issues.
AT&T leased all of its equipment to customers, so it retained perpetual
ownership. That really skews the ROI curve--just about any reasonable
thing you can do to extend product life (30 year old 2500 series desk sets
in working condition are not an anomaly). At the time of their
manufacture, AT&T had good reason to suspect that transistors may represent
a particular vulnerability in a design. Note also, that the transistor
did not affect the telephone's ability to receive calls or the customer's
ability to flash the operator in an emergency.
AT&T footed the bill for repairs on their equipment. Anything that can be
done to improve reliability pays huge dividends. Passive components like
capacitors and inductors, if propery designed and specified are almost
bulletproof. And AT&T had a guaranteed market in the millions of units.
Finally, AT&T was tightly regulated. They were allowed a certain maximum
profit margin--it really didn't matter if a few dollars could be saved here
or there.
Cheers,
Chuck