Archived from groups: alt.video.ptv.tivo (
More info?)
Nice to know motor technology is same today as it was so
many decades ago. Still, where are the numbers? For that
matter, concepts of back EMF do not explain damage on power
up. It explains damage due to too low line voltage. Again,
all that power up transition is nice. But with experience
under your belt, failure during startup is more often a result
of damage done earlier during 'hours of operation'. Damage
created by too many hours of operation or operation under
excessively low voltage can appear later as a startup failure.
Classic junk science reasoning is why light bulbs fail. The
naive assume failure is from temperature transition during
power cycling. If true, then orange stop light signals would
fail as often as red and green. But instead we first learn
both theory and numbers from light bulb manufacturers. At
minimum, you should have first consulted the industry bible -
IES Lighting Handbook, an industry benchmark. This is my
polite way of putting into your face a fact. I have extensive
knowledge on these topics, in part, because I never believed
anyone or any source that does not provide the numbers. It is
probable that I was doing this stuff before you even existed.
The IES does provide numbers. Light bulbs are damaged by
hours of operation, voltage (which determines filament
temperature), and vibration (when filament is hot). IES even
provide the exponential equations to calculate this life
expectancy using numbers such as voltage and hours of
operation. Power cycling only appears on a list created by
junk science reasoning. Notice when a light bulb fails.
Often the vaporized filament is observed hours before that
bulb failure. Dark spots on the inside glass created by
damage from too many hours of operation or excessive voltage.
Damage so extensive that the bulb may even fail due to a very
gentle shock called power on.
Manufacturers for disk drives and other computer equipment
do provide numbers for power cycling. They just don't bother
to make those numbers available where human eyes glaze over
when the numbers appear. Power cycling is destructive only in
myths. Myths because so many 'experts' just know without
first obtaining numbers. All disk drives have a power cycle
spec. They also have an hours of operation number. I cited
the IBM drive because it was a lowest power cycling number.
So low that it significantly caught my attention. Drive was
speced for only 40000 power cycles.
Disk drive bearing wear is mostly from hours of operation.
The constant heat and vibration of use is destructive - not
power cycling. Vibration and the resulting heat being a
problem to life expectancy. Ball park numbers - the drive
that makes more noise will tend to be the drive with earliest
mechanical failure. Vibration being a most destructive
force. But again the paragraph must be tempered by numbers.
If one power cycles every five minutes for 1 year, then power
cycling could be destructive. How many hours of operation is
that drive rated for? If the poster does not provide a
number, then he cannot provide a definitive answer - as others
have done here. No numbers means junk science reasoning.
Will Way Lee's disk drive life be prolonged if he powers off
every day? Yes. For disk drives, the most destructive
effects tend to be hours of operation. However and again,
that answer does not provide numbers and can be 100% different
if his operation (number of hours on and off) is different
from assumptions I am making. Conclusion can vary even for
different model disk drives.
But again, no responsible person here could answer his
question as some here did without those numbers. More often,
using the drive for only one hour every day (one power cycle
and one hour of operation) will prolong most disk drives
significantly. Why? Power cycling is not anywhere near as
destructive as was hyped here. Most often a 30 minute power
cycle every day will not impact the drive's life expectancy
consider other factors such as 'becoming obsolete in 8 years'.
So who cares. Leaving it one or power cycling it will mean
the drive may die well after it has been replaced. The drive
will be obsolete long before it should fail. Should you want
a Tivo drive for 15 years, then the answer must be tempered by
manufacturer's numbers for hours of operation and power
cycling. Generally, disk drive failure is more often due to
hours of operation.
One final point. What so often causes electronics failure?
Power cycling - that is also called normal operation. Those
transistors that drive a motor or operate inside a
microprocessor do not stay at the same temperature. They
temperature cycle greatly during normal operation. The point
of transistor switching consumes so much energy that the
transistor junction even transmits infrared light during
switching. If you believe power cycling to be so destructive,
then turn digital electronics off. Normal operation is
constant power cycling. Destructive thermal transitions are
constant during normal operation. In the early days of
transistors, one specification was the transistor's number of
switching transitions. Once the number of transitions in a
transistor was significant enough to be speced.
Power cycling inside a microprocessor is particularly
extreme. A microprocessor will go from less than 1 amp of
power to tens of amps - in microseconds. Why is this power
cycle so destructive? The time of transition - microseconds -
means normal microprocessor operation has extensive power
cycling stress (and junction temperature changes) during
something called normal operation. Myth purveyors instead
assume transistors stay at same temperatures during normal
operation. They assume because they don't have numbers.
But again, have I provided anything to draw a single and
definitive conclusion? No. Obviously not. I have left out
the numbers. I have only provided enough information for the
enemies among us - the junk scientists - to turn wild
speculation into political statements.
Anyone who provided Way Lee with a definitive answer was
simply lying to everyone.
In the meantime, learn from things observed every day - the
traffic light bulb. Power cycling does not cause light bulb
failure. Bulb that fails most often (and are therefore most
often replaced with LEDs) is the bulb with most hours of
operation. Light bulb life expectancy is measured in same
hours of operation whether it power cycles in a traffic light
or remains on constantly outside the doorway. Only myth
purveyors using junk science reasoning say that light bulb
life expectancy is shortened (significantly) by power
cycling. When one thinks (for so many decades) like an
product person, then one first demands the numbers. Only a
scum politician or junk scientist would rationalize a
conclusion without numbers.
No numbers means junk science reasoning. In this thread
were too many junk scientists promoting power cycling as most
destructive - only because they knew - numbers be damned.
That explanation about back EMF in a motor tells us zero about
when a motor might fail. But then back EMF and motor damage
is more important in operation under excessively low voltage.
Strongbox wrote:
> Motors will experience inrush current when started, since the back-EMF
> generated by the motor windings is not there yet to limit the current. It
> works simular to the cold vs. warm resistance function of a light bulb (most
> light bulb failures occur when turning them on). Current limiters can and
> do help this problem, but there is still inrush current none the less.
>
> The manufacturers don't spec this effect since they know most people fall
> into one of two catagories; 1) Those who use their PCs very little, so the
> HD doesn't collect a lot of on-air time during the life of the PC, and 2)
> Those who are heavy, must have the greatest and latest, users who cycle out
> the PCs and/or HDs long before they reach the end of their useful life.
>
> Although the hard drives are designed to meet their life specs while taking
> into account a reasonable amount of cold starts, Way Lee asked if the
> lifespan of the HD would be prolonged if it was turned off and on to avoid
> the 30 minute cache instead of letting it run continuosly. The answer is
> still no.
Facebook
Twitter