testing power supplies

  • Thread starter Thread starter ed jurewicz
  • Start date Start date
E

ed jurewicz

Does anyone know of a detailed how to on testing power supplies. I have some
experience wit multimetters but that was in another life before a long and
grinding recent journey through a doctoral program. Last multimetter was an
old heathkit that was tossed out several years ago.

So the step by step direction for testing the supply would have to be pretty
elementary. Also any reqs for a reasonably price meter?
 
ed jurewicz said:
Does anyone know of a detailed how to on testing power supplies.

The short story is that it isnt easy to do the most comprehensive
testing, particularly checking that the noise level on the rails is
acceptible etc. The only really viable approach is to use a CRO with
the power supply in use in the system its supposed to be powering.
I have some experience wit multimetters but that was in another life
before a long and grinding recent journey through a doctoral program.
Last multimetter was an old heathkit that was tossed out several years ago.
So the step by step direction for testing the supply would have to be pretty elementary.

Its more complicated than it looks, particularly testing whether
the supply can deliver what the label says it can deliver.
 
ed said:
Does anyone know of a detailed how to on testing power supplies.

Here's one: www.driverheaven.net/guides/testingPSU

Here's another:
http://forums.extremeoverclocking.com/showthread.php?t=137886

Some PSUs won't turn on when you connect the green Power-On wire to
ground (any black wire) because they need a load of at least 0.5A on
the +5.0V rail (between any red and any black wires), and a few,
especially those found in major brand computers, also need a similar
load on the +3.3V rail (orange wires, but some have brown wires
instead, and some Dells use blue-white wires and don't have any green
wires but instead the sole orange wire is the Power-On signal).

If you want to test your PSU thoroughly, follow these simple steps:

www.xbitlabs.com/articles/other/display/psu-methodology.html
Also any reqs for a reasonably price meter?

Anything with a resolution of at least 3.5 digits (AKA 2000 count)
should be more than adequate, even those $3 ones that are sometimes on
sale at Harbor Freight.
 
Rod said:
Mindlessly silly, you dont need rubber safety gloves when testing a power supply.

Oh, rubber *gloves*. I thought they meant rubber pants. No wonder I
couldn't find them except at Fetish Fashions.
 
Back
Top