Re: [CAD_CAM_EDM_DRO] Servo heat-sensors or paranoia?

AbbyKatt wrote:

>I was thinking of buying some current sense ammeters - the analog meter
>type that use a shunt at the back, for my power supply, but I decided
>not to in the end as they were quite expensive. They are quite cool
>though if you like to watch the meters bob up and down as the system
>works. :)
>
>But I'm wondering if any of this is really necessary - isn't just
>placing a heat E-stop on the motor going to achieve the same effect, if
>not better? For slow overheating (ie too much torque, not enough lube,
>blunt tool) this will work fine and shutdown the system before damage
>occurs. I'm wondering about fast failures tho.. If the servo motor
>stalls, are we going to be able to sense the overheat quickly enough on
>the outside or will the wires in the core go red hot before there's even
>much chance for the heat to get to the outside of the motor?
>
>If a servo stalls, how long before the smoke comes out? 3 seconds? 2
>minutes? 5 minutes?
>
>

It depends on the servo amp and the way you have the current limit set.
If you set the current limit to the absolute peak current rating, you
will never
get away with 5 minutes. You might get away with 2 minutes on a motor
that started cold. Normally, peak current should never be delivered to
the motor for more than a second at a time, and totalling no more than
a couple of seconds a minute. A motor temp sensor on the motor housing
of a DC brush motor is WAY too far away from the heat source to be
effective.
That's why higher-end servo amps like Servo Dynamics have thermal
integrators
in them to make these decisions. They integrate current over time, and may
actually use current sense resistors in a thermally insulated block to
attempt
to simulate motor heating.

Jon