Why are gears made from bulk metallic glass ideal for precision robotics?

[kolleamm]

In an attempt to find ways to reduce the noise generated by my robot, I'm trying to understand why the servo motors themselves make noise. On the web I've read their noise has to do with the gears they have. Even when a servo is not on and you turn it, it makes noise due to the gears.

Can someone please explain why the gears make noise when moving.

Thanks in advance.

 

[Grinkle]

You will get better answers than mine, I am sure. The gear boxes make noise because the have components that are vibrating in the sonic frequency range. Why would they? The components are stiff by human tissue standards (they feel stiff to your fingers) but they are certainly not completely rigid, and the forces generated when the gears turn definitely cause (hopefully elastic) deformation in the components. Since the components are elastic, they will return to their initial shape as the forces are removed, and since the force application is periodic, so is the deformation / relaxation, and this is what causes the vibrations and the noise.

Edit:

Also I imagine non-perfectly fit components will knock against each other also periodically, and these periodic impacts will cause vibrations.

 

[kolleamm]

Vibration seems to be the right answer I was looking for. I'm guessing friction plays a large role as well.

 

[jack476]

I think it happens because a lot of hobbyist servos are driven by a rather cheap DC motor and a cheap gearbox. So the motor itself is making more noise, and the gearbox is less precisely machined so that means more rattling around. I do a lot of amateur robotics stuff and I have definitely noticed that cheaper servos tend to be louder. Opening them up to lubricate the mechanism every once in a while might help as well.

 

[tygerdawg]

Take a marble and roll it on a table top. Does it make noise? Of course. Uneven surface roughness, collision forces between surface variations, sound induced by a soundboard-effect mechanical & percussive action.

A gearbox and the motor that drives it may have dozens of bearings in races creating one component of the total noise.

And each time a gear tooth comes into contact with another gear tooth, there's going to be noise generated and resonated through the housing of the gearbox. Those teeth are designed perfectly on paper or in CAD, but suffer manufacturing tolerances on their profiles. For a general purpose industrial gearbox with helical gears, the quantity of gear teeth in contact is only a few. And the teeth can be assumed to be relatively large (meaning: larger mass, more strength, harder to "ping" to a noise level). Most robot joint gear boxes will be harmonic or epicyclic arrangements due to torque density requirements, have small teeth, and will have dozens of teeth in contact. Hence the whine we hear.

 

[Ketch22]

Your query asks "Why does my robot servo's make noise?" My initial response would be what sort of robot is it and what type of servos? For most hobby robots the servos are simple as well. A major contribution to noise is the gear mesh pattern. A close significant second is the presence and type of lubrication. A straight cut gear is the simplest and most economical to manufacture but also creates the most noise. For a typical plastic straight cut gear with no lubrication one can expect quite a bit of noise and the frequency will be determined by the number of gear meshes per second.
For an industrial type robot with helical or herringbone gears that are operating in an oil bath or spray. Noise has significantly different source and characteristics. Tell us about the project and the noise you are encountering.

 

[Baluncore]

What does the noise sound like?
Does it depend on the external load or torque?

 

[OCR]

What does the noise sound like?

That would help a great deal if form members could hear "the noise".....

Search here, and here... see if you can find a noise that is somewhat equivalent to what you hear.... ?

You might even find an answer to your question... if not, post something that sounds as close as you can find...

 

[CWatters]

Many servos have a "dead band" that turns off the motor once the output arm is close to the demanded position. This dead band is intended to stop the servo hunting or buzzing about the demanded position.

However if the servo is heavily loaded it can cause the output arm to be pulled far enough away from the demanded position that the motor is re-energised to pull it back to the required position. This process can repeat leading to a buzzing effect.

This effect increases power consumption and is one reason why a model aircraft pilots should make sure control surfaces are free to move without any binding or excess friction.

On a robot you might try adding counter weights to reduce the load/torque on the servo?

 

[Nidum]

(1) Most of the noise in these toy size servos comes from motor whine and first stage gear whine .

(2) The plastic gears used are often assembled with zero running clearance so as to eliminate backlash . This tends to make gear noise worse .

 

[Nidum]

(3) You could easily design and build a servo drive which was both quieter and more suited to your purpose .

 

[Tommyboyblitz]

Gears will always make noise, using grease or bathing them in oil reduces noise a lot. Maybe do some experiments?

 

[mechpeac]

Gears always make noise. Cheaper spur gears are noisy because of the quick impact between gear teeth. Noise reduces with better made teeth or different gear shapes such as helical gears (they make less noise because the impact between teeth is more gradual). Cheap gears have more 'slop' in the design and will always have noise.

 

[1oldman2]

On the subject of gears and robotics, JPL is making very good progress.
http://www.jpl.nasa.gov/news/news.php?feature=6684

Gears are essential for precision robotics. They allow limbs to turn smoothly and stop on command; low-quality gears cause limbs to jerk or shake. If you're designing a robot to scoop samples or grip a ledge, the kind of gears you'll need won't come from a hardware store.

At NASA's Jet Propulsion Laboratory in Pasadena, California, technologist Douglas Hofmann and his collaborators are building a better gear. Hofmann is the lead author of two recent papers on gears made from bulk metallic glass (BMG), a specially crafted alloy with properties that make it ideal for robotics.

Hofmann said that gears made from BMGs can "run cold and dry": initial testing has demonstrated strong torque and smooth turning without lubricant, even at -328 degrees Fahrenheit (-200 degrees Celsius). For robots sent to frozen landscapes, that can be a power-saving advantage. NASA's Mars Curiosity rover, for example, expends energy heating up grease lubricant every time it needs to move.

"Mass producing strain wave gears using BMGs may have a major impact on the consumer robotics market," Hofmann said. "This is especially true for humanoid robots, where gears in the joints can be very expensive but are required to prevent shaking arms. The performance at low temperatures for JPL spacecraft and rovers seems to be a happy added benefit."