Stuck Wheel Problem and Fix

[metabflux]

My Neato XV12 generated a "Right Wheel stuck. . ." error as described by OP. The dysfunctional wheel was substantially more difficult to rotate than the functioning wheel as indicated by other posters with the same error.

I First tried running the wheel via USB using PuTTY and the SetMotor command in TestMode as described in the neato programers manual and suggested by glnc222 above. The wheel turned if manually rotated at the start of the command, but was inconsistent in speed and stopped frequently. However, the SetMotor command is limited to a maximum distance of 10 meters and therefore insufficient for the "break-in " method described by aziesemer2.

I then performed the "break-in" method as described in the original post by aziesemer2, with minor modifications. . . Many thanks to aziesemer2.

UPDATE:

After approximately 6 hours of run time, the wheel stuck error returned. The wheel was again very difficult to turn. Repeating the above procedure allowed the robot to work, but only for a single cycle. I presume it worked only while the motor was warm as suggested by SyRenity above. Therefore, I decided to disassemble the motor. I'm sure the right tools exist to make this exercise rather trivial, but this is probably not advisable for the average consumer (or me, really).

Description:

The motor was removed from the wheel housing. The brown RPM sensor on the back of the motor was removed by securing the gear and twisting it off the shaft. The electronic wheel sensor was removed using a soldering iron to release the leads one at a time. The back of the casing is crimped in places over the back plate. These crimps were relieved using needle nose pliers. The back plate was removed (this was not simple). The brushes were removed and appeared to be in good shape. The contacts were also clean. The rear bearing looked smooth. The rotor was not completely removed because I was unable to remove the gear pressed on the front shaft (there must be a tool for this). Inspection of the rotor by shining light into the casing revealed a ~2 cubic mm brown debris adhered to one of the coils. The origin of this debris is unknown. The material, which appeared to be melted into the coil was carefully removed using tweezers. The casing and coil was rinsed with isoproply alcohol which yielded a few more fragments which also could not be identified. The casing and coil was allowed to dry and lubrication was applied to the shaft and magnets. The motor was reassembled. The backing was reinstalled by gently tapping it into the case with a tap and hammer. I noted that the positioning of the backing in the casing is critical to allow clearance between the rotor and magnets. The backing was aligned by tapping on each side to minimize resistance when turning the gear. The backing was secured by crimping the casing as it was before it was disassembled. The wheel sensor was soldered back into place and the RPM sensor was pressed back on. The motor and wheel housing was assembled. Manual rotation revealed that the resistance of the two wheels was indistinguishable from each other. The robot has run for 10-hours without failure. The resistance of the wheel remains normal.

Interpretation:

The origin of the debris is unknown but probably originated from inside of the motor since there is no opening that would allow entry, and this failure seems to be recently relatively common among these robots. Debris, attached to the rotor accounts for the vibration that was noted after initially removing the motor. The "break-in" work around may help by freeing the debris from the rotor or magnets. depending on how the debris settles, this may be a temporary fix. Other factors, such as clearance between the rotors and magnets may also be important. For instance, small increments in the alignment of the backing resulted in the magnets impinging on the rotor. Taking the motor apart is not for everyone, since Neato has an excellence warranty policy, and you can buy the wheel online for ~$40. I had fun with it.

[glnc222]

excess lacquer used to seal magnet coils?
thanks for such fine detail.

[mfortuna]

From what you are describing there was a melted material on the coils (wiring). This could be insulation meaning:

a. Your motor overheated at some point.
b. Segments of the motor wind may now be shorted together becuase of lack of insulation. This could result in the motor having different performance than nominal.

I assume the melted material was enough to cause it to rub against the magnets?

[SyRenity]

Great description - probably exactly what happened to me and to other people suffering this problem.

Pity these Chinese motors kill all the experience.

[glnc222]

What's needed is a USB control procedure to check a new Neato for quality problems in the wheels, so discovered before cleaning many times. Then a unit could be returned right away if received defective. Not sure how to test these mechanical qualities.

[metabflux]

excess lacquer used to seal magnet coils?
thanks for such fine detail.

It looked just like a resin or lacquer, but I have no experience with small electric motors so I'm not sure. Good idea.

From what you are describing there was a melted material on the coils (wiring). This could be insulation meaning:

a. Your motor overheated at some point.
b. Segments of the motor wind may now be shorted together becuase of lack of insulation. This could result in the motor having different performance than nominal.

I assume the melted material was enough to cause it to rub against the magnets?

I know for sure that the motor ran hot during the "break in" work around, but I don't know if it ever got hot during normal operation before the initial failure. The debris on the coil was definitely substantial enough to severely disrupt the balance but was not touching the magnets. Some of the debris that washed out could have been contacting the magnets. If I had to do it again, I would start by adjusting the back plate to see if it improved the friction, then work my way into the motor.

It is absolutely possible that the motor now has impaired performance. I was worried about damaging the coil while removing the debris, and was relieved when it worked. I assumed that a short in the coil would cause a catastrophic failure, but this is far from my expertise. I should check the GetMotor command to see how it is performing, but if I dig into it one more time my wife will kill me for certain.

Great description - probably exactly what happened to me and to other people suffering this problem.

Pity these Chinese motors kill all the experience.

My thoughts exactly . . .

[glnc222]

Digital rpm control probably prevents the overheating when motors are stopped with brushes engaged, for which a polyfuse was needed in the added side brush motor lacking digital control, though more to protect Neato power lines. Overheating from just intense use seems unlikely to melt insulation, formulated for some heat resistance, an obvious need. The current draw is sensed, if they bother to monitor it. It's sensitive to getting stuck after all.

[mikes60]

Does anyone has actual link where to buy wheel motor separately from the whole wheel assembled?

[SweChris]

Hi
The place that i have found where you can buy spareparts fot Neato is http://www.robotshop.com/neato-replacement-parts.html

[kheops]

I have the same problem (On xv-21), I tried the procedure without success actually the motor ran on 7,5 v during around 40 min with substantial vibration and with uneven speed but now nothing work.

The motor seems totally dead....

Is it possible to have a new motor through neato support ?

Kheops

[flyingfarm]

We just received our Xv-21. It is our first robot. Very exciting... until it stop working on it's very first attempt to vacuum the house. It reported left wheel stuck.

I called Neato's support number and explained the issue to the girl that took the call. She immediately said that our unit would have to be replaced and the new one would not have this problem. They are, obviously, very aware of the stuck wheel issue. They are out of stock on the xv-21 model so I agreed to take an xv-12 with a pet/allergy kit. The said it is the same robot but does not have the purple trim.

Meanwhile, I have done some checking on our xv-21. Turning the drive wheels by hand I can tell the stuck left wheel turns with more resistance than the good right wheel. I took the left wheel assembly apart and the drive train spins freely without the motor. I cannot determine anything wrong with the motor when turning it by hand.

I then swapped the right and left motors and put xv-21 back together. Sure enough it now reports the right wheel is stuck.

I will take the bad motor apart and see what it looks like inside.

[flyingfarm]

Here is what I found after disassembling the motor. The commutators were badly gouged and the edges of the brushes were chipped. The spaces between the brushes were packed with carbon and copper dust. I cleaned it up and reassembled the motor and it works fine now. It is hard to say what caused this because I only have a sample of one to work with. No doubt the Neato engineers have it all figured out and resolved.

It is nice to have our Neato working. We will keep using it until the replacement unit arrives our the motor fails again.