23 Feb
2009
23 Feb
'09
1:25 p.m.
I have to say, for all the talk of failing caps in
power supplies
I've only ever seen one electrolytic cap fail *ever*, and that was
last week in a one-year-old graphics card that has hardly ever been
powered off...
23 Feb
23 Feb
1:36 p.m.
My gut feeling is that electrolytics got a lot lot
better in the 70's
compared to earlier generations. It's not just that they're 20 years
newer than the ones from the 50's, they really were better
quality to begin with.
The big changes came in the mid 1950s and 60s, and continued into the
70s. After World War 2, when the military had some time on their hands
and a big pile of statistics concerning component failures, the
industry started to really look at component quality carefully.
Basically, the data from the designs of the 1940s led to a lot of
experimentation and trials in the 1950s, which led to better parts by
the 1960s. The aero industry really pushed things along.
--
Will
2:58 p.m.
Electrolytics caps: standard cctech recurring topic #2
I swear this concern about electrolytic cap failure is a legacy myth that has
filtered down through the antique radio community, going back to issues of wet
electrolytics and *early* dry ones.
For pre-1950's tube equipment and later *consumer-grade* *tube* equipment I may
replace electrolytic (and paper) caps as a matter of course. Just about
everything else is left as is.
My (industrial-grade) tube frequency counters from the late-1950s are
functioning fine with the original electrolytic caps in them. In one of these
counters there was a bad oil-filled bathtub cap (one of the most reliable types
of that era). The electrolytics were fine.
Pretty much all (I can't think of any exceptions offhand) of my dozens of
transistor and IC counters, calculators (and computers) from the 60's and 70's
are functioning with the original electrolytic caps in them.
Occasionally, I will bother to use a variac.
I have had one 'lytic cap explode on me. That was in a home-made tube radio
power supply from the 1950's that I received in the 90's. Made quite a mess.
The mushy innards of the capacitor were splattered around the entire underside
of the chassis. That situation would have been worthy of initial replacement or
a variac, but I didn't have one at the time.
The issue of poor-quality 'lytic caps from the late-1990's/early-2000's is
well
known in the industry, but that was a quality issue related to several
particular manufacturers (example article:
http://crppit.epfl.ch/documentation/Computer/Hardware/Problem/Defect%20Main…).
Aside from such particular issues as this, IMHO, concern over modern
electrolytic caps is vastly overrated.
1:50 p.m.
>>>> "Tim" == Tim Shoppa
<Tim> writes:
Tim> Other times they literally explode, or they cause other
Tim> components in the circle to fail catastrophically... my
Tim> experience is that switching supplies are far more sensitive to
Tim> out of spec ESR's in electrolytics than any old radio ever was.
That makes sense. Switching supplies tend to be designed far more
closely to the margin. (And that assumes the original good design
hasn't been butchered by unscrupulous manufacturers; for an example of
that read the article in QST a few years ago about modifying a
switching supply to be a ham radio supply.) I remember reading
articles in the trade press about how to size inductors, and
transformers, in switching supplies. It's all about making them as
small as possible while still able to handle the load. This is why
modern switchers get noticeably warm when in use. So the margin of
error is less than it used to be.
paul
2 p.m.
Paul Koning wrote:
That makes sense. Switching supplies tend to be
designed far more
closely to the margin. (And that assumes the original good design
hasn't been butchered by unscrupulous manufacturers; for an example of
that read the article in QST a few years ago about modifying a
switching supply to be a ham radio supply.) I remember reading
articles in the trade press about how to size inductors, and
transformers, in switching supplies. It's all about making them as
small as possible while still able to handle the load. This is why
modern switchers get noticeably warm when in use. So the margin of
error is less than it used to be.
And since the market for most switchers seems to be PC's with a
life span of 6 months before they are out dated, I don't see them
needing a long lifespan.
paul
2:03 p.m.
On Feb 23, 2009, at 2:00 PM, bfranchuk at jetnet.ab.ca wrote:
Nah...Switchers are in everything, not just PCs. PCs make up a
relatively small portion of electronics produced in the world today.
-Dave
--
Dave McGuire
Port Charlotte, FL
That makes
sense. Switching supplies tend to be designed far more
closely to the margin. (And that assumes the original good design
hasn't been butchered by unscrupulous manufacturers; for an
example of
that read the article in QST a few years ago about modifying a
switching supply to be a ham radio supply.) I remember reading
articles in the trade press about how to size inductors, and
transformers, in switching supplies. It's all about making them as
small as possible while still able to handle the load. This is why
modern switchers get noticeably warm when in use. So the margin of
error is less than it used to be.
And since the market for most switchers seems to be PC's with a
life span of 6 months before they are out dated, I don't see them
needing a long lifespan. 
2:28 p.m.
On 23 Feb 2009 at 14:03, Dave McGuire wrote:
Nah...Switchers are in everything, not just PCs.
PCs make up a
relatively small portion of electronics produced in the world today.
Yup, they're showing up with increasing frequency in wall warts,
which makes me wonder how really safe that is. It could be that the
days of the iron-core low-frequency power transformer are numbered.
--Chuck
2:51 p.m.
Chuck Guzis wrote:
On 23 Feb 2009 at 14:03, Dave McGuire wrote:
I suspect there are a few (large) ones humming away, since I heard that they stopped
DC power only a few years ago.
Nah...Switchers are in everything, not just
PCs. PCs make up a
relatively small portion of electronics produced in the world today.
Yup, they're showing up with increasing frequency in wall warts,
which makes me wonder how really safe that is. It could be that the
days of the iron-core low-frequency power transformer are numbered. --Chuck
3:08 p.m.
On 23 Feb 2009 at 12:51, bfranchuk at jetnet.ab.ca wrote:
I suspect there are a few (large) ones humming away,
since I heard that they stopped
DC power only a few years ago.
Seriously, how different (in terms of number and type of components)
is the circuitry that goes into the base of a cheap compact
fluorescent lamp from that found in a small switcher? It might be
that a better economy of scale is possible with a SMPSU wall-wart--
and you gain the advantage of 120/220v operation without changing the
design.
Cheers,
chuck
3:24 p.m.
Chuck Guzis wrote:
On 23 Feb 2009 at 12:51, bfranchuk at jetnet.ab.ca wrote:
As well as the switchers in low-voltage halogen lamp fixtures. Apparently the
savings in iron and copper winding has made them more economic for some years now.
I suspect there are a few (large) ones humming
away, since I heard that they stopped
DC power only a few years ago.
Seriously, how different (in terms of number and type of components)
is the circuitry that goes into the base of a cheap compact
fluorescent lamp from that found in a small switcher? It might be
that a better economy of scale is possible with a SMPSU wall-wart--
and you gain the advantage of 120/220v operation without changing the
design.
5:24 p.m.
On Feb 23, 2009, at 2:48 PM, bfranchuk at jetnet.ab.ca wrote:
Wall warts, CD players, televisions, set-top boxes, video games,
server-class computers (BIG switchers!), refrigerators and stoves/
ovens with digital controls...pretty much everything that plugs into
a power outlet these days contains a switching power supply.
-Dave
--
Dave McGuire
Port Charlotte, FL
Nah...Switchers are in everything, not just PCs. PCs make up a
relatively small portion of electronics produced in the world today.
I thought the other stuff was wall warts. :)
How ever, other products don't have a small margin profit as PC's. 
24 Feb
24 Feb
12:39 a.m.
On Monday 23 February 2009 01:25:29 pm Tim Shoppa wrote:
My gut feeling is that electrolytics got a lot lot better in the 70's
compared to earlier generations. It's not just that they're 20 years
newer than the ones from the 50's, they really were better
quality to begin with.
Having worked with stuff starting out with junk made in the 1950s and earlier
and onward, I would agree with this.
> I have to say, for all the talk of failing caps in
power supplies
> I've only ever seen one electrolytic cap fail *ever*, and that was
> last week in a one-year-old graphics card that has hardly ever been
> powered off...
Gordon, you don't mention how many caps you've looked at or tested, or
how you've done so, but I encounter them constantly. Not so much in
the 1980's vintage DEC equipment YET, but it's no myth that AEC's are
electrochemical vats that have a lifespan. The lifespan varies widely
depending on many factors, heat being the big one. I work on old radios and it's pretty much a given
that any set has
electrolytics in need of replacement. If the set was used for a while,
in fact some lytics were probably already replaced in the 50's or 60's, and
maybe the replacement needs replacement today.
"Failed" is a relative term... it's easy to find electrolytics leaky
enough that they no longer meet their original spec, or leaky
enough that they get warm. But the set still works.
Other times they literally explode, or they cause other components
in the circle to fail catastrophically... my experience is that
switching supplies are far more sensitive to out of spec ESR's
in electrolytics than any old radio ever was.
Yes. I've done a number of those, too. A bit of hum in an old radio isn't
necessarily a killer, and it's amazing sometimes how bad those old parts can
be and the radio still seems to work. :-)
--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, a critter that can
be killed but can't be tamed. --Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin
5601
days inactive
5602
days old
12 comments
8 participants
participants (8)
-
bfranchuk＠jetnet.ab.ca -
cclist＠sydex.com -
hilpert＠cs.ubc.ca -
mcguire＠neurotica.com -
Paul_Koning＠Dell.com -
rtellason＠verizon.net -
tshoppa＠wmata.com -
wdonzelli＠gmail.com