Unix PC and WD2010B

Thad P Floryan thad at cup.portal.com
Sun Aug 19 10:02:18 AEST 1990


yarvin-norman at cs.yale.edu (Norman Yarvin) in <25832 at cs.yale.edu> writes:

	In article <32819 at cup.portal.com> thad at cup.portal.com (Thad P Floryan)
	writes:
	>Other differences (on my systems) include the 3B1 power supplies
	>adjusted to 5.20VDC and 12.4VDC, Cramolin-preserved connectors,
	>operation on UPS systems, and everything properly grounded.

	Why the higher power supply voltages?  For higher speed to cure
	flakeys?  Are you normally overloading the power supplies, to bring
	them back to normal?

The specs call for (+/-) 5% on both the +5VDC and +12VDC supplies.  This means
that an "in-spec" system could have:

	 +5.25 VDC through  +4.75 VDC, and
	+12.60 VDC through +11.40 VDC

I used to manufacture 15,000 to 20,000 power supplies a month for several
noted modem manufacturers, and it's my experience that operating at the high
end (within spec) promotes cooler power supply operation and more reliable
overall operation of connected systems.

I still manufacture a (smaller) quantity of power supplies for use with one of
my own products and I still operate the voltages at the high end (and this
product is used in a "Hi-Rel" application by, interestingly enough, many of
the phone companies and the DoD, and is GSA-qualified/-certified).

Resistive losses (PC board traces, connectors, cabling) and "marginal" part
problems are offset by the slightly higher voltage operation.

Considering that I operate my systems and other equipment 24 hours/day (and
have for years) and simply have NO problems should be testimony to my theory.

For the record, I've seen 3B1/UNIXPC systems which operated erratically; a
simple adjustment to the power supply caused the systems to function flawlessly
.
I've seen the "+12 VDC" as low as 10.2 on some UNIXPCs, and those same systems
had erratic HD operation which was cured by the power supply adjustment.

Vibration, shock of transporting, oxidation of the contacts in the adjustment
potentiometers themselves, and component aging and burn-in, can cause the power
supply voltages to drift over a period of time.  I have to check AND adjust the
supplies on my VAX and DEC-20 systems every several months, and I check the
supplies on my UNIXPCs and other systems every 6 months or so when I clean out
the fans.

If anyone IS having problems with their system(s) and hasn't checked the power
supply voltages, do YOURSELF a favor and check/adjust them, especially if
you've bought your system "used" and/or it's over a year old (yes, there still
ARE brand-new, factory-sealed systems to be found (though they were manufacture
d
in late 1986).

CAUTION: do NOT measure the voltages with your power supply free-standing, it
MUST be connected to a load.  To set the +12 voltage it's best to use a "Y"
connector attached to the HD so that the voltages are available on the unused
connector for easier measurement; you don't want to, say, sneeze or slip, and
short the probes against something and destroy your system(s).

If anyone needs the pinouts and specs of the UNIXPC's power supplies again,
I can repost the info (I don't know if the info is archived at osu-cis).

Thad Floryan [ thad at cup.portal.com (OR) ..!sun!portal!cup.portal.com!thad ]



More information about the Comp.sys.att mailing list