Mysterious behaviour with Gearmotors

[PickledPhil]

Hi All

I have a real mystery on my hands which I'd like to share for someone cleverer to laugh at.

I'm working on a robot project and have two identical gear motors that are supposedly 'reversible'.  Each has a small 6500 rpm motor and a metal gearbox giving a 12rpm output.

So far so good.  I assumed that 'reversible' meant I can swap the polarity of the connections and the motor will run in the opposite direction... even though the motor has a little red dot near one terminal...(suspicious)

I have discovered that one of these gearmotors has dramatically different stalling currents when running in opposite directions; 0.5A in one direction and 2.5A in the other, and the stalling force is much less in the 2.5A direction, which is a problem when driving a robot.

Ok... so I researched and found out that a motor may have an asymmetric geometry which could account for this, but then it gets weirder...;

The other gear motor does not show this behaviour... it has a stall current of 0.5A in both directions...

Ok... so the gearmotor is faulty... but then it gets weirder still...;

I dismantled the suspect gearmotor and tested the stall current on just the motor and it is the same
in both directions...0.5A. it only shows a difference in stall current when connected to the gearbox.

I examined the two gearboxes and the tiny internal components seem identical up to the tolerances of my micrometer.

So, I'm confused... or in a pickle.... Is this a fault or expected behaviour in 'cheap' gearmotors... and how could that be?

Evidently I'm missing something obvious... Any ideas?

Thanks to all

[DC1MC]

First try to switch the gears (or the motors to see if the issue is specific to a gear. Also it could be that the problem appears when the gear is attached to the motor ( position, centering, level and so on), it  could be that "face" of one motors, where the gear gets fixed is uneven, or the gear surface as well.
Detailed pictures will help a lot.

 Cheers,
 DC1MC

[Richard Crowley]

There are some gearboxes that have a slip clutch in one direction. The ones I have seen look like a spring around a hub. The spring tightens in the "forward" direction and loosens in the "reverse" direction.  Perhaps one of your gadgets is "frozen up" and the one-way clutch is locked so it behaves conventionally.

[Kjelt]

And the mechanical load on both gearboxes is identical ? E.g. you have two robotarms and it is in both the same joint ?

[PickledPhil]

Thanks very much for the replies guys - definitely some food for thought here...  I will try swapping the motors and gearboxes over and see what happens.

Each gearmotor is connect directly to a wheel and are mounted on opposite sides of a small robot.  I was initially testing the stall current by applying a breaking force to the wheels - i.e holding onto it -  I noticed the problem because when moving across the ground one motor was stalling for no obvious reason and the current consumption was skyrocketing.

Anyway, now I have the motors completely dismantled from the robot chassis and the wheels removed but still see the same symptoms.

There doesn't seem to be any springs or clutches inside the gearbox - it's not fancy... just gears, spindles and bushings and a lot of grease.  I will see if I can grab a photo later.

I was wondering if some manufacturing issue with the motor windings or geometry could explain this.  I have some more motors on the way... I'm hoping I can find a pair with roughly equivalent performance!

Phil

[PickledPhil]

Actually before I can get a photo of my unit... here's what it looks like externally...:

[Mr. Scram]

There might be a manufacturing imperfection in the gearbox. Maybe parts have warped subtly in a direction you didn't check or something similar.

[PickledPhil]

Yup, I guess so. 

It's an interesting and unfortunate gotcha when buying electric motors that I wasn't expecting... as Dave might say; "A warning for Young Players"... It's a shame I'm not so young anymore 

[Conrad Hoffman]

I'd vote for a problem in the gearbox, something a bit misaligned or a thrust problem. You should also be aware that many motors have a brush scheme that's designed for only one direction of operation. They'll go backwards, but maybe not with as long life or the same performance.

[SparkyFX]

Another way to look at the problem:
They probably did no heat treatment and grinding to the gears to harden them (unusual for cheap devices), which means their surface came probably direct from the casting, which means it can take a while to break in and wear down to an involute gear tooth shape. Same goes for plastic gears.

I reckon "reversible" means that forces are translated backward through the gearbox and could drive the motor. In contrast most planetary gear boxes are self-locking (and have a high amount of backlash), so the are not reversible.

As a first try to compensate such issues:
- letting it run for a while in both directions might help
- exchanging grease might help as well (the grease fills the gaps between gears and therefore has influence on the stall torque).

[PickledPhil]

Thanks Guys

Yes, the brush/magnet design optimised for one rotation direction is what I was thinking was what they refer to when they say 'reversible' or 'not reversible' but it's the kind of thing that wouldn't be mentioned on sales website...: "It can be reversed but it's nowhere near as good and will wear out more quickly..." 

This particular gearbox uses a worm gear on the motor shaft so it cannot be back driven from the output shaft, so I don't think it's called 'reversible' for that reason at least.

Cheers

[SeanB]

Is the bolt pattern identical on both sides, so that you can reverse the output shaft and have both of them operate in the some motor polarity but have different shaft rotation directions?

[mikerj]

It's hard to see how the gearbox could have any influence on the stall current, provided the motor is a actually being stalled (i.e. there is nothing slipping within the gearbox).  A bad gearbox could allow the motor to be stalled more easily, but the stall current is determined by the resistance of the brushes and motor windings.

[SparkyFX]

This particular gearbox uses a worm gear on the motor shaft so it cannot be back driven from the output shaft, so I don't think it's called 'reversible' for that reason at least.

Ah, my mistake. Thanks for giving this detail.

It's hard to see how the gearbox could have any influence on the stall current, provided the motor is a actually being stalled (i.e. there is nothing slipping within the gearbox).  A bad gearbox could allow the motor to be stalled more easily, but the stall current is determined by the resistance of the brushes and motor windings.

As long as "stalled" means 0 RPM, it does not make much sense.
The brushes could be misshaped and connect both sides of the commutator concurrently in one case vs. the other, but that could happen to both gearboxes. If it is a design issue, it might even depend at which position the motor is stalled and could happen to both, only readable after several measurements.

[PickledPhil]

It's hard to see how the gearbox could have any influence on the stall current...

Yeah... that's what I thought too!...

...provided the motor is a actually being stalled (i.e. there is nothing slipping within the gearbox).  A bad gearbox could allow the motor to be stalled more easily, but the stall current is determined by the resistance of the brushes and motor windings.

...slipping internally...Good thought - perhaps that is exactly what is happening... but wouldn't we see a current up to, but not exceeding the 0.5A that we see on the other gearmotor when fully stalled?

Is the bolt pattern identical on both sides, so that you can reverse the output shaft and have both of them operate in the some motor polarity but have different shaft rotation directions?

Unfortunately not, SeanB... The bolt pattern and output shaft are on one side only.  I was trying to concoct a wacky mechanism to allow both motors on opposite sides to run in the same direction... but it wouldn't be pretty.. urg...  and the problem would still exist when I put both motors in reverse...

I can't really see a way around it, except find a pair of matched motors.

[Gregg]

Since you stated the worm drive is attached to the motor shaft; it could be a thrust problem within the motor.  You may be able to carefully take the motor apart and rearrange any thrust washers to make the two somewhat closer to the same. See how much free play in the thrust of the motors and make sure nothing is binding in reverse with external thrust applied; try pushing or pulling on the worm gear while turning with your fingers. 
Since they are cheap, you may want to buy several for comparison.  Also, better quality lubrication may help [if Chinese lubricants are anything like Chinese flux; maybe they are the same thing]

[Zero999]

It's hard to see how the gearbox could have any influence on the stall current, provided the motor is a actually being stalled (i.e. there is nothing slipping within the gearbox).  A bad gearbox could allow the motor to be stalled more easily, but the stall current is determined by the resistance of the brushes and motor windings.

Yes, the stall current is simply equal to the voltage divided by the DC resistance. It's purely an electrical parameter and is independent of the gearbox. It's more likely to be an electrical, rather than a mechanical problem.

[PickledPhil]

Thanks guys...

So in summary, what we probably here is an electrical parameter of the motor, as we expect, but influenced by the gearbox pushing and pulling on the motor shaft, changing the DC resistance by shifting the alignment of the magnets and windings. 

I imagine that, because the shaft has a worm gear on it that when stalled in one direction the motor shaft will be pulled outwards and when stalled in the other direction it will be pushed back into the motor - and as a result of sloppy bearings in the cheap motor as Gregg suggested.

I will see if I can try and find a way of verifying that.

[MK14]

If 0.5 Amps is the correct stall current. Sometimes there are datasheets or spec sheets, with the motor stall current to verify this.

Then the 2.5 Amps maybe is an internal short or very low resistance path between stuff in the motor, as it moves about, due to the sloppiness.