Re: geckos lifetime...

>
> thanks for the input...
> hmm,....thats what makes me most nervous, maybe running them at

6amps

> is the pain-limit? i currently use jvl drivers 100% flawless hour
> after hour, year after year, rated to 6 amps, and i squeeze out all
> of those amps, and they dont get hotter than a fart, but they cost

3

> times more than the 201´s....
> the only reason why i was thinking of a swap to the geckos is that

i

> thought i could save a some $. however, i dont want to spend my

time

> travelin around and replace popped g201´s...
> does anyone know any commercial, heavy duty, serious machine brand
> out there with geckos that runs ok? (at 6 or even 7 amps)
>

The concept here is pretty simple. Electronic components work better
and last longer the cooler you keep them. All active components have
max ratings. When you read the specs you find that those ratings are
at a given temperature. Assuming you can somehow keep the temps down
at the interior of the component to those levels, the MTBF (Mean Time
Between Failures) is extremely high. For every degree C you go above
that figure the MTBF starts to fall. The more power you run through
a device the more losses (and heat) it develops so obviously running
at lower power levels produces less heat. There are derating curves
that will predict how much less current you can run through a device
at elevated temps to maintain a low MTBF. Designs are based on those
curves.

Here are several MTBF things to consider.

1. Infant mortality. The largest number of failures will occur in
the first X hours of full usage. Proper burn-in and testing removes
those units from the pool.

2. Decreasing heat increases MTBF. This number is predictable. For
components that have MTBF's of many years, the way to test for MTBF
(unless you have years to wait) is to accelerate the test by raising
the temperature.

3. The MTBF of an assembled unit is only as good as its worst
component.

4. The probability of a failure in any given assembly increases with
the total number of parts. Simple designs tend to have fewer parts
and fail less often.

Airflow is your friend. Good heat transfer (thermal coefficient)from
the interior of the device to the heat transfer medium (heatsink) is
essential. Just a little airflow vastly improves the transfer of
heat from the heatsink to the surrounding air. Vertically mounting
heatsinks that do not have forced air (fan) cooling improves the
transfer of heat. Worst is mounting the heatsink horizontally where
the natural convection currents cannot easily flow away from the fins.

All good designs should take MTBF into account. For a manufacturer
the failure of a component in the field is MUCH more expensive to fix
that one in testing so every engineer is expected to design circuits
that will last forever under all conditions ;-0

As to the failure mode of most components: Most active components
(transistors, IC's, FET's, etc) fail shorted. They meltdown, expel
the magic blue smoke, and turn from a semiconductor to a full
conductor. If large amounts of current are available to pump through
the short, then other interesting effects are triggered, some of
which involve load noises and flying projectiles.

Most passive components (resistors, caps,) fail open. There are
exceptions to the rule.

While there could be a scenaro where a failed component could cause
your drive to decide that your gantry needs to be in your neighbor's
yard, the likelyhood is small and a proper set of limit switches that
remove power to the drives is enough to cover that possibility. I
have yet to see a fractional HP loaded motor store enough energy to
keep running more a few milliseconds with power removed. Normally
the things that fail in a power circuit are the power components
(sometimes from the failure of upstream components) and without them
you have a hard time keeping motors spinning.