I have some questions about damaging a servo by overheating:
(as an example, consider the HS 645 servo)
is it safe to measure the stall current by overloading the robot joint for a few minutes?
the HS 645 is specified as having a “running current” of 0.45 amperes. I have also seen this “running current” described as a “no load operating current”. Suppose I measure the stall current of this servo to be 5 amperes (I’m just guessing). How high a current (that is, how large a load) can I put on this servo and still run it for several minutes without damage? For several hours?
Another way to ask the same question, if the stall torque for this servo is 133 oz-in, how much torque can I put continuously on this motor without damage?
What is the actual destruction mechanism of an overheated servo? Do the magnet coils burnout? Does some electronic component self-destruct?
Dont know but i wouldn’t do that myself.
I suppose there is another way to measure what you are essentially looking for.
Not to sure, but IMHO:
Enough torque to reverse the shaft’s turning motion, makes it go bad. Eventually. If spinning fast enough, it creates a anti-polarized current in the coils. See it as a double shortcircuit.
In my experience (in any motor) the coil wraps simply get hot (very thin wire = to much resistance for high currents) and burnout the isolation inbetween thus causing a shortcircuit kickstarting a vicious melting circle.
And depending on the make, the onboard chips could fry because of overheat / melting. Some have a power regulater onboard that should prevent.
There are no real answers to these questions from the manufacturer. The stall current over time failure will generally show itself as melting of the plastic box that houses the geartrain. Although they are very similar I have melted an HS-5645 servo, but I have never melted an HS-645.
The HS-645 can blow one side of an H-bridge if you abuse them. Try turning on a hexpod robot in the lowered position with an 8 lb payload to see what I mean. On second thought, better not.
The problem is heat and how to get rid of it. There is no way to determine how much abuse is survivable. lol
Thanks for the responses.
I realize that one can’t give exact numbers to answer this question, but I would have thought that someone would have at least a rough idea from their experience. That is, take a servo with a stall torque of 133 in-oz. Is it reasonable to use this servo in a joint that is required to continuously lift a 4 ounce weight at the end of a 1 foot arm? (48 in-oz)
Is the point possibly that hobbyists don’t think in these terms when making their robots?
I think the point may be that most folks don’t do the kind of tests you describe because the result can be a destroyed $30+ servo. Consider that many folks run off batteries so if the robot were going to be standing there doing nothing for 5 minutes they would just switch it off. Another consideration is the loads on the servos of a moving robot vary widely and just moving helps reduce the heating. Last point is that most “normal” hobby servos use a brushed dc motor and applying them in the manner you describe, that is holding a static load for long durations of time, is not really what they are designed to do.
It should be fairly easy for you to test for what you want to know (at the expense of one servo), then publish the results for the particular servo. I think microservos tend to strip gears at stall, and large servos may burn out the h-bridge transistors prior to damaging the motor wiring.
