8 Sep
2010
8 Sep
'10
1:01 a.m.
Folks;
The 5V rail in my Sun 2/120's PSU recently failed suddenly while the
system was running. No pop, but the rail measures out around 2.5VDC. I
believe I have tracked the fault down to one of a pair of 26000uF
electrolytics which are in parallel across the 5V output studs. These
are screw-lead, rated for 7.5V (12V surge), 85 deg. C, -10 +75% tol.
They are 1.375" dia, and 2.125" tall. Mechanically, there is no room
for anything even slighlty larger.
Mouser stocks a part which is physically too tall---at 3.125", too
tall by an inch. This is a 10V (12V surge) part. They also stock a 50V
part from a different manufacturer which is even taller (>4"). There
are 10V (12V surge) parts that fit, but have too low a capacitance
(18000uF). There are 7.5V (9V surge) parts which will fit, but have
too high a capacitance (49000uF), and I'm not totally sure how to be
certain that a 9V surge rating is adequate in this application.
Digi-key stocks nothing even close.
Any suggestions?
ok
bear
11:02 a.m.
On 2010 Sep 8, at 1:01 AM, r.stricklin wrote:
The 5V rail in my Sun 2/120's PSU recently failed
suddenly while the
system was running. No pop, but the rail measures out around 2.5VDC. I
believe I have tracked the fault down to one of a pair of 26000uF
electrolytics which are in parallel across the 5V output studs. These
are screw-lead, rated for 7.5V (12V surge), 85 deg. C, -10 +75% tol.
They are 1.375" dia, and 2.125" tall. Mechanically, there is no room
for anything even slighlty larger.
Mouser stocks a part which is physically too tall---at 3.125", too
tall by an inch. This is a 10V (12V surge) part. They also stock a 50V
part from a different manufacturer which is even taller (>4"). There
are 10V (12V surge) parts that fit, but have too low a capacitance
(18000uF). There are 7.5V (9V surge) parts which will fit, but have
too high a capacitance (49000uF), and I'm not totally sure how to be
certain that a 9V surge rating is adequate in this application.
How about using just one of those new 49000 uF units to replace both of
the 26000 uF units (*2=52000 uF)? The 3000 uF diff isn't likely to
matter, but it can be buffed up with a little 3000 uF cap in parallel
if you feel compelled.
2:30 p.m.
How about using just one of those new 49000 uF units
to replace both of
the 26000 uF units (*2=52000 uF)? The 3000 uF diff isn't likely to
matter, but it can be buffed up with a little 3000 uF cap in parallel
if you feel compelled.
I wouldn't think that 3000uF was necessary. Aluminium electrolytics are
wide-tolerance devices at best, so that 49000 may be way off anyway (why
isn't it a 47000, which is a prefered value, I wonder...)
Which reminds me of what I call the 'silly capacitors' in HP machines.
Things like 8100uF (why not 8200uF), 21000uF (why not 22000uF), and the
classic of a PSU with both 5600uF and 6000uF devices on the same PCB...
-tony
11:22 a.m.
On 2010 Sep 8, at 1:01 AM, r.stricklin wrote:
The 5V rail in my Sun 2/120's PSU recently failed suddenly while the
system was running. No pop, but the rail measures out around 2.5VDC. I
believe I have tracked the fault down to one of a pair of 26000uF
electrolytics which are in parallel across the 5V output studs. These
are screw-lead, rated for 7.5V (12V surge), 85 deg. C, -10 +75% tol.
They are 1.375" dia, and 2.125" tall. Mechanically, there is no room
for anything even slighlty larger.
Mouser stocks a part which is physically too tall---at 3.125", too
tall by an inch. This is a 10V (12V surge) part. They also stock a 50V
part from a different manufacturer which is even taller (>4"). There
are 10V (12V surge) parts that fit, but have too low a capacitance
(18000uF). There are 7.5V (9V surge) parts which will fit, but have
too high a capacitance (49000uF), and I'm not totally sure how to be
certain that a 9V surge rating is adequate in this application.
Forgot to mention: considering the caps are on the output of the
regulator, I would expect the 9V surge rating to be fine. If they were
before the regulator it might be a greater concern.
What is the basis for suspecting the caps? If they are dragging the 5V
down to 2.5 they should be sucking a lot of current through the
regulator and dissipating a lot of power themselves (heating up).
12:19 p.m.
On 8 Sep 2010 at 11:22, Brent Hilpert wrote:
What is the basis for suspecting the caps? If they are
dragging the 5V
down to 2.5 they should be sucking a lot of current through the
regulator and dissipating a lot of power themselves (heating up).
Don't know about the OP, but after a time, a lot of these caps just
dry out and might as well be empty cans. I've got a couple of Lambda
supplies like that--great 120Hz ripple, needing new computer-grade
replacement caps. Something like 10,000 uF at 30V--not easy to find.
--Chuck
5000
days inactive
5000
days old
5 comments
5 participants
participants (5)
-
ard＠p850ug1.demon.co.uk -
bear＠typewritten.org -
cclist＠sydex.com -
david_comley＠yahoo.com -
hilpert＠cs.ubc.ca