How long does a resistor last?
larry moe 'n curly
I'm referring to 1/4 - 10 watt resistors in TVs and PC power supplies,
probably in 30-50 Celcius surrounding air.
I thought that resistors were supposed to be chosen for twice the
actual power, but I've seen many 0.5W to 1.0W resistors run at almost
exactly their rated power continuously.
Lostgallifreyan
4a77-45b7-b88...@h40g2000prf.googlegroups.com:
How long is a piece of string? I think you need to rethink what you're
asking. More to endurance than duration.
John Fields
---
That's because it costs more to use a larger resistor than is necessary.
Let's say that you can get 1/2 watt resistors for a penny in quantities
of a million, and one watters for 1.5 cents.
If the 1/2 watters will do the job like they're supposed to and you buy
a million one watters instead, you've just thrown away $5000.00.
Frank S
problems in electronics is caused by under rated components (not enough
power or voltage headroom). Caps are a big culprit.
"John Fields" <jfi...@austininstruments.com> wrote in message
news:a77uc5hu0idqj9jvj...@4ax.com...
MooseFET
wrote:
The life at rated wattage will depend on many factors. The legs of
the resistor will conduct the heat away. How much copper they hook to
will control how well this works. The brand of the resistor also
matters. One from Uncle Chin's Resistor Company will last much longer
than those from the Shaw Dee Resistor Factory because Uncle Chin uses
better ceramic.
If you use that urethane conformal coating that the US navy seems to
like so much, it will turn dark and look very ugly long before the
resistor fails. The ceramic materials will do their best when clean.
John Larkin
<larrymo...@my-deja.com> wrote:
Depends on what it's made of and how well it's cooled. We use 0603
surface-mount resistors at half a watt, because we heat sink them
well. Half-watt carbon resistors will die at half a watt in a confined
space.
Enameled wirewounds are very tough, up until the enamel melts.
Some resistors will die from temperture cycling stress.
But in general it's good to derate 0.5 maybe.
John
legg
<larrymo...@my-deja.com> wrote:
As an example, expected failure rates for film resistors due to
applied power alone (temp=K) double between 50 and 100% applied (ref
MIL-HDBK-217 - just one of the methods used to estimate failure
rates). It is of course extremely difficult to keep temperatures or
other factors constant in real life, as this single factor is altered.
Calculated temperature effects on failure rate, between 60 and 120degC
are also a factor of two.
It is not always obvious what construction, limiting hot-spot
temperature or rating a component carries; you cannot make assumptions
based solely on body size or apparent style.
Commercial designers have every encouragement not to introduce weak
points in the reliability of a piece of equipment, in spite of your
observations.
Passive component failures, though more easily spotted, are unlikely
to be the initial source of any specific fault, rather they point to
other factors or events.
Just because a component runs hotter than others in it's immediate
vicinity, doesn't identify it automatically as a potential source for
failure. It may indicate pre-existing damage elswhere, or a be a
conscious attempt to shift normal operating or single-fault abnormal
stresses from less rugged circuitry.
There's more to reliability and fault analysis than spot temperature
measurements, but I encourage your interest. It's worth the trouble,
if only for the intellectual exercise.
RL
larry moe 'n curly
What additional information should I specify, other than rated power,
actual power, and ambient temperature? I'm not dropping high
voltages.
Lostgallifreyan
6d11-4fc3-96d...@s21g2000prm.googlegroups.com:
Materials. What I was getting at is that resistors aren't always nicely
behaved metal/ceramic thingers, and carbonisation of an epoxy or urethane or
other coating might affect a high-value resistance enough to consider it
failed even though the resistive element is fine. Basically it can get
complicated, so you need to think of the specific conditions to eliminate
guesswork.
Another way of looking at it is, if it's a ceramic/metal type and isn't
glowing with heat, or melting solder if mounted that way, then it's likely to
outlast everything else in the circuit and is therefore the least of
concerns...
Robert Baer
Robert Baer
widget by 50 cents per one watt resistor used.
John Fields
---
Then your widget will cost more than an identical widget designed to be
lean, and your competition will eat your lunch.
Michael A. Terrell
And steal your dessert! ;-)
--
The movie 'Deliverance' isn't a documentary!
brent
Robert Baer
last up to 50 years...
Lostgallifreyan
news:dsGdnej0YOnJEkzX...@posted.localnet:
Common knowlege to me. :) It's kind of what amazes me about electrolytics. So
much variation in lifetime, I'm surprised that there are so many short-lived
ones, I'd have thought that good answers were so long out of patent that it
would be a no-brainer to choose ways to make them last. Maybe it had to do
with them not pushing for high capacity though, I mean, for a given size, the
capacity was a lot lower than now.
At this point, either the thread will morph into days and perhaps hundreds of
posts on the subject, or everyone will decide it's been done to death before
and there won't be any.
Hal Murray
>Depends on what it's made of and how well it's cooled. We use 0603
>surface-mount resistors at half a watt, because we heat sink them
>well. Half-watt carbon resistors will die at half a watt in a confined
>space.
>
>Enameled wirewounds are very tough, up until the enamel melts.
>
>Some resistors will die from temperture cycling stress.
>
>But in general it's good to derate 0.5 maybe.
There is another dimension to consider: duty cycle. I've been told
(but don't have first hand experience) that old fashioned carbon
composition resistors are much better at pulse loads that newer
carbon film resistors.
--
These are my opinions, not necessarily my employer's. I hate spam.
Lostgallifreyan
news:DO2dnZ1gIacDNU_X...@megapath.net:
> There is another dimension to consider: duty cycle. I've been told
> (but don't have first hand experience) that old fashioned carbon
> composition resistors are much better at pulse loads that newer
> carbon film resistors.
>
They're also a lot better at high voltage. I'm just mentioning this because
when I used some for a HeNe supply current limit resistor, they really
impressed me. Most other things I tried eventually burned. Basically, a solid
mass is better than a film for HV, it seems.
Phil Hobbs
Except for the horrible voltage coefficient. Carbon comps' resistances
decrease by about a quarter at their rated voltage. That's okay for a
HeNe, since the tube's load line is pretty steep.
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058
hobbs at electrooptical dot net
http://electrooptical.net
Lostgallifreyan
news:0badnQb6Bo8xok7X...@supernews.com:
> Lostgallifreyan wrote:
>> hal-u...@ip-64-139-1-69.sjc.megapath.net (Hal Murray) wrote in
>> news:DO2dnZ1gIacDNU_X...@megapath.net:
>>
>>> There is another dimension to consider: duty cycle. I've been told
>>> (but don't have first hand experience) that old fashioned carbon
>>> composition resistors are much better at pulse loads that newer
>>> carbon film resistors.
>>>
>>
>> They're also a lot better at high voltage. I'm just mentioning this
>> because when I used some for a HeNe supply current limit resistor, they
>> really impressed me. Most other things I tried eventually burned.
>> Basically, a solid mass is better than a film for HV, it seems.
>
> Except for the horrible voltage coefficient. Carbon comps' resistances
> decrease by about a quarter at their rated voltage. That's okay for a
> HeNe, since the tube's load line is pretty steep.
>
> Cheers
>
> Phil Hobbs
>
>
Well that's cool cos that's about all I ever use them for. :) But they were
the perfect answer in this singular case.
Lostgallifreyan
> Except for the horrible voltage coefficient. Carbon comps' resistances
> decrease by about a quarter at their rated voltage.
Veering swiftly offtopic... this interests me because people often rave about
the sound of valve amps. I don't mean audiophool ravings, I mean those who
like the effect especially in overdriven amps for bass guitar and other
instruments where the sound is appropriate, and very good. It's normally
attributed to the harmonic distortion and saturation effects in the valves,
but now I'm wondering how much of it might be due to that voltage coefficient
in carbon resistors. Maybe designers deliberately tried to reduce the effect
by careful biasing, or maybe there is a strong effect that is overlooked, I
really have no idea...
Comments welcome, so long as they come from either experience or interest in
this subject.
deathproof
Tim Williams
> I'm wondering how much of it might be due to that voltage coefficient
> in carbon resistors. Maybe designers deliberately tried to reduce the effect
> by careful biasing, or maybe there is a strong effect that is overlooked, I
> really have no idea...
Noise?
It's well known that adding rumble and pop noise to a recording makes
it sound more appealing to the phonograph crowd. And apparently makes
them angry when you tell them that's what you've done.
You aren't going to see much nonlinearity, and resistors certainly
weren't well picked -- 470k is close enough, and with 20% resistors,
who cares anyway. What scares me more is, for instance, how a common
Magnavox 6V6 PP amp uses 0.047uF for the coupling cap to one 6V6, and
0.0047 for the other. And the phase splitter doesn't split anything,
it's just a cascaded stage. The only balancing is due to a voltage
divider. Ewwww. I'd be willing to bet they did that intentionally,
since LF distortion makes phantom bass. They used teensy output
transformers, too.
Tim
Colin Trunt
"larry moe 'n curly" <larrymo...@my-deja.com> wrote in message
news:e18113f3-4a77-45b7...@h40g2000prf.googlegroups.com...
Serious question?
It depends on what it is made of.
Also resistors are measured in ohms not watts.
Genereally speaking the 1/2 life of the material is a good clue
as to when it#s resistance will have doubled.
SO don't use anything to radioactive.
A typical value would be a million triilion years, if it
fails before them take it back to the shop :O)
Dave Plowman (News)
Colin Trunt <co...@trunt.com> wrote:
> "larry moe 'n curly" <larrymo...@my-deja.com> wrote in message
> news:e18113f3-4a77-45b7...@h40g2000prf.googlegroups.com...
> > When run at its rated power?
> >
> > I'm referring to 1/4 - 10 watt resistors in TVs and PC power supplies,
> > probably in 30-50 Celcius surrounding air.
> >
> > I thought that resistors were supposed to be chosen for twice the
> > actual power, but I've seen many 0.5W to 1.0W resistors run at almost
> > exactly their rated power continuously.
> Serious question?
> It depends on what it is made of.
> Also resistors are measured in ohms not watts.
You must be lucky if you choose them without taking into account the
current they're handling.
--
*Caution: I drive like you do.
Dave Plowman da...@davenoise.co.uk London SW
To e-mail, change noise into sound.
Ian Jackson
writes
>
>"larry moe 'n curly" <larrymo...@my-deja.com> wrote in message
>news:e18113f3-4a77-45b7...@h40g2000prf.googlegroups.com...
>> When run at its rated power?
>>
>> I'm referring to 1/4 - 10 watt resistors in TVs and PC power supplies,
>> probably in 30-50 Celcius surrounding air.
>>
>> I thought that resistors were supposed to be chosen for twice the
>> actual power, but I've seen many 0.5W to 1.0W resistors run at almost
>> exactly their rated power continuously.
>
>Serious question?
>
perfectly 'sensible' question.
>It depends on what it is made of.
>Also resistors are measured in ohms not watts.
>Genereally speaking the 1/2 life of the material is a good clue
>as to when it#s resistance will have doubled.
>SO don't use anything to radioactive.
>A typical value would be a million triilion years, if it
>fails before them take it back to the shop :O)
>
Is this a misguided attempt at humour, or have you really not the
slightest idea what you're talking about?
I suspect that the answer is very similar to that to "How long is a
piece of string", but maybe there is someone who is prepared to venture
a reasoned answer.
--
Ian
Arfa Daily
"Ian Jackson" <ianREMOVET...@g3ohx.demon.co.uk> wrote in message
news:1D0mB9Nm...@g3ohx.demon.co.uk...
I think it depends on many factors, not the least of which are type -
w/wound, c/composition, metal film etc - how close it's being run to its
power rating, how much free circulation of air there is in the equipment,
how close it's being run to its voltage rating, and its value, depending on
type.
Resistors in high voltage circuits have a tendency to go high, for instance
anode (plate) load resistors, screen feed resistors, resevoir cap voltage
sharers or bleeders. Resistors that have a high value in the first place
i.e. above say 220k, have atendency to go high in normal use, although some
types, the old cc ones for example, are worse for this than more modern
types. Low value resistors in low voltage circuits, on the other hand,
seldom fail, unless they have done so in an obvious way - burnt out - caused
by another fault downstream of them. There is no reason at all why a
resistor in say the base bias circuit of a transistor, shouldn't last,
unchanged in characteristics, pretty much for ever. Obviously many
generalisations here, and there will of course be many anecdotal exceptions
to the rule, but I think that gives a fairly useable answer to the question.
Arfa
Colin Trunt
"Dave Plowman (News)" <da...@davenoise.co.uk> wrote in message
news:50b2416...@davenoise.co.uk...
> In article <iDtGm.27$ei...@newsfe22.ams2>,
> Colin Trunt <co...@trunt.com> wrote:
>
>> "larry moe 'n curly" <larrymo...@my-deja.com> wrote in message
>> news:e18113f3-4a77-45b7...@h40g2000prf.googlegroups.com...
>> > When run at its rated power?
>> >
>> > I'm referring to 1/4 - 10 watt resistors in TVs and PC power supplies,
>> > probably in 30-50 Celcius surrounding air.
>> >
>> > I thought that resistors were supposed to be chosen for twice the
>> > actual power, but I've seen many 0.5W to 1.0W resistors run at almost
>> > exactly their rated power continuously.
>
>> Serious question?
>
>> It depends on what it is made of.
>> Also resistors are measured in ohms not watts.
>
> You must be lucky if you choose them without taking into account the
> current they're handling.
Not usually an issue.
Anyway I can allways stick in a resistor to drop the current
down to one it can handle. ;O)
Anyway if it lasts a day it should last a lifetime.
The usual solution is to add a heat sink.
Jon Elson
>> Depends on what it's made of and how well it's cooled. We use 0603
>> surface-mount resistors at half a watt, because we heat sink them
>> well. Half-watt carbon resistors will die at half a watt in a confined
>> space.
>>
>> Enameled wirewounds are very tough, up until the enamel melts.
>>
>> Some resistors will die from temperture cycling stress.
>>
>> But in general it's good to derate 0.5 maybe.
>
> There is another dimension to consider: duty cycle. I've been told
> (but don't have first hand experience) that old fashioned carbon
> composition resistors are much better at pulse loads that newer
> carbon film resistors.
>
Proven! I was using some 100 Ohm resistors to limit current in a
spark-gap ignitor for mercury short-arc lamps. The film resistor just
BLEW off its resistance film on the first shot. I mumbled "oh yeah, I
KNEW it was going to do that" and replaced with a bulk carbon resistor,
and that held up for the life of the unit. You can guess, the pulse
current was 20,000/100 or about 100 A. 200 A ^ 2 * 100 = 4 mega-watts
for a couple microseconds.
Jon