Torture testing servos?

[Foamforce]

I had another crash yesterday due to a servo locking up. The elevator on my Nutball locked and I nosedived. When I retrieved it, I couldn’t move that servo by hand until I forcefully jiggled it. Then it worked again for a bit before locking up again. I replaced it. I’m assuming that a gear tooth broke off. I’m going to do an autopsy tonight to see if I can see any obvious damage.

So my question is, are there any techniques for weeding out bad servos? For example, 10 minutes on sweep mode attached to a small weight?

I’ve had several crashes due to bad servos. The first three were super cheap no-name servos from Amazon, so I switched to Tower Pro servos (SG92R). This is the first Tower Pro that has failed for me, so I think they’re much better, but I’m pretty concerned now that one has failed. Are the metal gear servos a lot better?

 

[Bricks]

Many times the issue is how big the control surface is especially after a crash the torque the servo receives can be very high. I have had both plastic and metal gear go bad many times I think it is luck of the draw. I have some 9 gram cheapies that have flown a 3D foamy for years and not one hickup. Then I have had Hitec and other metal geared servos not make 15 flights.

 

[Foamforce]

Many times the issue is how big the control surface is especially after a crash the torque the servo receives can be very high. I have had both plastic and metal gear go bad many times I think it is luck of the draw. I have some 9 gram cheapies that have flown a 3D foamy for years and not one hickup. Then I have had Hitec and other metal geared servos not make 15 flights.

That’s a little discouraging. I don’t think I’ve crashed this model before and I wouldn’t think that a Nutball would be heavy enough to stress a 9g servo, although the control surface is fairly large.

How often would you say you experience a servo failure? Do you think they’re less likely with metal geared servos?

 

[LitterBug]

I have had extremely good luck with servos over the last 6+ years. 1 metal died due to massive impact, and one plastic gear that failed after dropping my Stratosurfer on the rudder. I fly hard and crash hard.... Guessing now that is about to change since I acknowledged it. Mostly Emax, Hitec, and Bluebird.

 

[Merv]

I agree with Bricks, servo life is the luck of the draw.
I buy the cheap servos, $1 or so for plastic gear and $2-3 for metal gear. I fly them hard, most of the time I will get 2-3 years out of a servo. My servos definitely last longer than my planes do. Every once in a while, there will be a brand new servo that is bad.

I never give a servo a second chance, if it ever locks up, I change it.

I do test each servo before I put it in a new plane. Have it sweep back and forth with a servo tester, while applying some pressure to the arm with my finger. You are not looking to stall it out, you just want to see if there are any bad teeth in the gears. I have the servo push my finger one sweep, then I will push on the servo arm as it returns. Then repeat for the other side. If I feel anything but a smooth sweep, the servo goes in the junk bag for parts. Servo parts can be very handy at times.

3-4 sweeps is all that is necessary to see if the servo is going to fail.

Metal gear servo are a bit stronger than plastic. I have found very few planes that absolutely need them. If I have any type of stability, I will always use plastic gear. The constant movement to keep the plane stable, mostly in the same place on the gear, will prematurely wear out a metal gear servos.

The vast majority of my servos are plastic gear.

 

[Foamforce]

I never give a servo a second chance, if it ever locks up, I change it.

I do test each servo before I put it in a new plane. Have it sweep back and forth with a servo tester, while applying some pressure to the arm with my finger. You are not looking to stall it out, you just want to see if there are any bad teeth in the gears. I have the servo push my finger one sweep, then I will push on the servo arm as it returns. Then repeat for the other side. If I feel anything but a smooth sweep, the servo goes in the junk bag for parts. Servo parts can be very handy at times.

3-4 sweeps is all that is necessary to see if the servo is going to fail.

Great info, thanks! I’ll go try that test on my known defective servo to see if I can get it to lock up. Does that test catch <most> of your bad servos before they lock up in the air?

It seems like most servo lockups in the air result in a crash, although I had one servo go bad on my elevator and that one always failed at the center point, so I was able to land on throttle only. Maybe that one was an electronics failure vs a gear lockup.

 

[Piotrsko]

Lock up is typically a mechanical drive failure, not working is typically electrical. Mode of failure is immaterial after a crash, but kudos for the salvage landing.

Sarcasm: testing it means it worked then, MIGHT continue working during flight.

Not really. State of the art is such that the electronics are more reliable than they ever were back in the day

 

[Merv]

...Sarcasm: testing it means it worked then, MIGHT continue working during flight...

Especially when you fly like I do.

 

[L Edge]

In many cases due to a crash, the result is the cheaper servos have plastic gears that strip. Not the electronics.

If you take the top cover off, one or more gears will be stripped or jammed. If you got the mechanical skills, a number of companies sell replacement gear sets that you can replace the gears and be up and running.
For instance, the Tower servo plastic gear replacement set is $4.49.

The same for metal gear set. I designed a wing that moves aft and forward while in flight. The torque stripped 2 metal sets of gears until I got the right size one. So, either a crash or exceeding the torque is why they fail.

Is it worth it or it is time to buy a better grade servo?

 

[Piotrsko]

I suspect the issue is: we are cheap. Why buy the $$$ good stuff when I can get a whole flying plane for the price of such servos then complain loudly when it doesn't hold up to the fence sieve maneuver

 

[Foamforce]

I suspect the issue is: we are cheap. Why buy the $$$ good stuff when I can get a whole flying plane for the price of such servos then complain loudly when it doesn't hold up to the fence sieve maneuver

😂 I don’t think anybody would begrudge a servo for not holding up to the fence sieve maneuver. However, having a way to test and weed out defective servos before they fly could prevent a fair number of crashes! I’m going to be trying Merv’s technique.

The expansion to that test I was thinking of was to clamp new servos to the edge of a table and then hang fishing weights (equal to the rated pull of the servos) off each one, and then set them at sweep for several minutes. Maybe the weights should be 120% of the rated capacity, not sure. Thoughts?

 

[Merv]

...I was thinking of was to clamp new servos to the edge of a table and then hang fishing weights...

Good idea, but it is a bit complicated.

...Maybe the weights should be 120% of the rated capacity, not sure. Thoughts?

Don’t want to break the servo.

 

[Foamforce]

Good idea, but it is a bit complicated.

Don’t want to break the servo.

Over engineering is the best engineering. 😉

 

[Piotrsko]

AFAIK, servos are rated lite continuous duty so exceeding that rating is going to cook some electronics. Otoh, in 50 years of doing this, I can think of only a couple of actual servo failures I have had and they were pretty much a mechanical failure like fluttering apart the output arm or the previously mentioned fence sieve maneuver