We had an SDS9300 on critical usage. The system came with a test rack, scopes, sig gens, full circuits, and spares. We could reapir any board without recourseOct 7, 2008 1 of 1View SourceWe had an SDS9300 on critical usage.The system came with a test rack, scopes, sig gens, full circuits, and spares.We could reapir any board without recourse to the manufacturer, there was no backup in the UK or Europe.A PDP7 we repaired ourselves, PDP11's were on 4 hour response by DEC and a car load of spares.Mike----- Original Message -----From: Ray SillsSent: Tuesday, October 07, 2008 3:19 PMSubject: [personal] Re: [midatlanticretro] Re: OT Humor
The assumption is that the replacement modules are known to be good.
It that's in question, then all bets are off.
Presumably, in a mission-critical application like SAGE, there would
be good spares on hand. Some large systems
have test bench ways to determine operational quality off-line, so
you'd know that your spare modules are OK.
But, yes... if the spares are questionable, then the problem is a
whole lot bigger.
73 de Ray
On Oct 7, 2008, at 10:06 AM, fairlanefastback wrote:
> Whats to say all the replacement modules are free from any defects in
> such a scenario? One can introduce a new problem unwittingly in such
> a methodology.
> --- In midatlanticretro@ yahoogroups. com, Ray Sills <raysills@.. .>
>> Hi Mike:
>> In a certain sense, that method of troubleshooting does optimize the
>> time it takes to get the unit functional. By "walling-off" half the
>> possible defective units by replacing them with good modules, you can
>> quickly determine where the fault lies. Of course, as you way, you
>> have to have at least 50% spare modules of the total that are known
>> be good. And by replacing half of all the modules, you have a 50-50
>> chance of getting
>> the system back up on the first swap.
>> If the system is still bad after that swap, then the same procedure
>> will work by swapping out half again of the remaining un-swapped
>> modules, etc.
>> It would mean a maximum of 4 swaps to find and replace the bad
>> module. And since it was a SAGE system, involving national defense,
>> you'd want to get it back up ASAP. That's the point, of course.
>> 73 de Ray
>> On Oct 7, 2008, at 9:16 AM, Mike Loewen wrote:
>>> On Mon, 6 Oct 2008, Stan Brewer wrote:
>>>> How did DEC Field Service Engineers change flat tires?
>>>> They took the spare out and kept swapping out tires until they had
>>>> 4 good
>>>> ones on the car! : )
>>> That reminds me of the first time I was exposed to the "shotgun"
>>> technique of troubleshooting. It was on the SAGE system, and one of
>>> systems developed a problem which was quickly narrowed down to a
>>> module of
>>> 20 pluggable units. Rather than taking the time to logically figure
>>> which PU was at fault, the tech replaced half of the PUs and re-ran
>>> tests. He continued replacing half of the remaining PUs until he
>>> was left
>>> with the bad one. He explained that sometimes you don't have time
>>> to be
>>> logical, but instead have to get the system back up as soon as
>>> I don't particularly agree with this technique, but if time is
>>> and you have the spares I suppose it makes sense.
>>> Mike Loewen mloewen@...
>>> Old Technology http://sturgeon. css.psu.edu/ ~mloewen/ Oldtech/
>>> ------------ --------- --------- ------
>>> Yahoo! Groups Links
> ------------ --------- --------- ------
> Yahoo! Groups Links