24 Feb
2015
24 Feb
'15
10:35 a.m.
I was working for DEC from 1973 onwards. I am mostly familiar with the
manufacture of the VT and LA terminals.
I think they tested all of the active devices and batch tested the
passives.
Once assembled the individual boards were given a resistive test (power
off) before a full power up diagnostic test.
Failures went to manual rework and rejoined the queue at the resistive
stage.
Third party and DEC branded (but subcontractor built) items where 100%
tested on load.
In a lot of VT terminals the tube and its drive electronics (Mainly Ball
Bros manufactured) were tested and rejects went
back whence they came.
All of this meant the vast majority of terminals came alive on switch
on. But that wasn't the end of it.
With systems shipping world wide the cost of warranty and having the
right people and parts in the right place was a big issue.
They found out that the field failure rate was linked to the factory
burn in time failure rate.
So by tweaking the burn in time they could control the field failure rate.
Yes DEC made reliable systems because it cost them less in the end.
On 24/02/2015 17:12, js at cimmeri.com wrote:
On 2/23/2015 11:58 PM, Brent Hilpert wrote:
** 1980's DEC stuff still seems to be OK, although I frequently find
bad caps in the 3rd party power supplies like those from Astec (eg.
BA123, BA23)... but ALL the AECs benefit from reforming -- you can see
right then and there how they do, how they perform before and after.
And you can tell whether a cap does or not by how it RESPONDS during
the reforming process. 1970's DEC PSUs like for an 11/34 also have
bad caps in them by now.... I just rebuilt 6 modules, and all had at
least one bad AEC... now, mind you, most not to the point where the
PSU had actually failed completely, but to where it *would* have.
I don't replace caps but reform them if their performance during the
reforming process indicates that they still have plenty of life left
in them. I reform instead of replace when I can in order to save
money, and because going through the process of finding the right
sizes and right models is a real major pain in the ass.
- J.
For solid-state consumer equipment like
transistor radios, caps
from the 1960s can be a problem but even then not inherently so. For
solid-state digital equipment: I don't think I've ever done a blanket
replacement of caps. The vast majority of my many dozens to hundreds
of calculators, digital test equipment, computers, etc., most of
which date from the 1960s and 1970s, are running with all their
original caps - including specifically electrolytic filter caps.
My experience is
different. 1940's through 1960's -- I replace all
AECs because most are either bad or too far on their way out.
1970's-1980's, depends on too many variables to give simplistic rules,
but I will test and reform all AECs. Some are found to be bad, but
most are still ok -- again, depending on many factors. 1990's
onwards, I don't worry about as there's rarely a failure (unless of
those bad Chinese variety).
I have rarely, if ever, reformed a capacitor for
a solid-state item.
** I do, all the time, as per above.
Keeping in mind that a mildly leaky cap which
could benefit from
reforming will reform during normal operation.
** I don't concur with this
statement because the charge rate is too
high. Countless times, there's been caps I couldn't reform with too
high a charge rate, but when it was slowed down, they came back
around. Many factors are at play.. amount of leakiness to begin
with, level of operational voltage vs. cap's rated voltage, etc.
BUT, AECs *do* benefit from and *do* require periodic use... which
keeps them from losing their oxide layer to a certain degree (within
the circuit's applied voltage range).
(Also, those big screw-terminal filter caps from
the linear-reg days
were called "computer grade" for a reason.) So, in relation to
computers and solid-state digital, I don't understand why people get
all concerned about caps.
** Because they go bad... even those big computer grade
ones. I've got
a whole bucket here full of them.
In my experience, blanket replacement and
reforming just hasn't been
warranted.
** Here, agreed.... *blanket* replacement isn't.. but they should
all
be checked.
For those
of us who are basically software people (or even pure
digital),
this stuff can be a little daunting - not necessarily because it's
actually
hard, it's just out of our comfort zone. I myself am certainly
daunted by the
concept of replacing every electrolytic in all the power supplies of
all the
vintage -11's I've got... (And I don't want to even think about all
the filter
caps on all the boards! :-)
Early (1970s era) switch-mode power supplies of significant capacity
are a category one might be a little leery or cautious around, as the
design and componentry may not always have been up to the task, as
they hadn't benefitted from a couple of decades of experience and
targetting.
Perhaps more owners of DEC equipment from that era could add
real-world experiences.