defective Tienly motor analysis "forensic exam" sponsored by Guges Mk3

Discussion in 'Articles' started by aotsukisho, Sep 9, 2021.

  1. aotsukisho

    aotsukisho Well-Known Member

    766
    423
    Another test sponsored by Guges Mk3, this time of a motor. A Tienly that was behaving un-Tienly-like, to paraphrase Guges.

    [​IMG]
    Motor arrived in its packaging bubble wrapped

    [​IMG]
    25K medium, looks okay at first glance

    [​IMG]
    Looking down at the endbell, springs and brushes look fine

    [​IMG]
    It doesn't appear that there are any visual issues with the setup, although I did find that the hand-turning of the armature was extremly grindy and not 'notchy' like how it's supposed to be (magnetic equilibrium as it turns)

    I've been slammed at work so I haven't had time to do further testing, but I wanted to get up a post as quick as possible.
     
  2. Squad144

    Squad144 Active Member

    158
    32
    Ocean Springs
    I love to read up one this kind of stuff. Can't wait to see what you find!
     

  3. Ben3721

    Ben3721 Well-Known Member

    631
    320
    Some motor commutators aren't really firmly attached to the armature wires. Like under the metal tab weld crimps the wire still has some coating of enamel so it doesn't really conduct well. Some motors have a short in one of the windings casting a short to the armature, too many shorts and the motor will eat power and not give it.

    But all of that can be tested pretty easily.

    Another issue I've see is the endbell brush holder plates are too tight on the brushes, resulting in them hanging up just a bit.

    The only other possibility is a improperly charged set of magnets, improper windings, or bad comm timing.

    I used to have a couple lonex motors that preformed horrible without cause. I gave up on them.
     
  4. Kuratosu

    Kuratosu Member

    81
    13
    Tucson
    Commenting to see what the outcome is. I have built some pretty janky Franken-motors myself, so it'll be interesting to see what happened to this one.
     
  5. aotsukisho

    aotsukisho Well-Known Member

    766
    423
    Thank you for the immediate show of interest everyone. As it turns out I had a chance last night to tear apart the motor.

    [​IMG]
    I opened one of my new Tienly 25K long motors in order to do a direct comparison.

    https://odysee.com/@digitalcomplex:5/36bKGXkRDOM:f

    Battery used was an older, well loved Turnigy Nano-tech 2s 6000mAh 25-50C with 4mm bullets, and using a 4mm bullet to clone Deans adapter. Motor hooked up with a basic LCT stock MOSFET controller.

    Tienly 25K long (new)
    Battery voltage: 8.24v
    Peak amps: 4.53A
    Minimum voltage: 7.51v
    Peak watts: 35.1W

    Tienly 25K medium (defective)
    Battery voltage: 8.22v
    Peak amps: 8.70A
    Minimum voltage: 7.93v
    Peak watts: 69.6W

    First things first there is a strange noise coming from the defective motor.

    Looking at the wattmeter, there is nearly double the peak power likely drawn at 0rpm spinup. These motors aside from shaft length are supposed to be identical, small fluctuations are acceptable due to materials and manufacturing tolerances but this is night and day different.

    [​IMG]
    In order to find out what's causing this, time to crack this motor open.

    [​IMG]
    When tugging on various bits of the motor to verify solid assembly, I found that the front face plate could come off. Because the pinion and tower are preventing me from pulling it off, I will address this later.

    [​IMG]
    Removing the brush springs from the endbell housing

    [​IMG]
    Visual inspection of the brushes shows minimal wear, meaning the motor has not been used very much if at all before I got ahold of it.

    There is a chip on one of the brushes but that might have been from me when removing it from the housing, either way this will not affect performance much after the motor is broken in.

    With the brushes out of the endbell, we can now pull out the armature.

    [​IMG]
    Because the front plate is loose, I did not have to yank off the pinion and tower in order to remove the armature.

    Now it is obvious why the motor was misbehaving. Part of one of the magnets has cracked, which caused mechanical damage to the rotor laminations near the top of the motor.

    [​IMG]
    Closeup of the magnet damage, most of it is that one large chunk but there is many smaller and still strongly magnetized pieces. I will have to clean this out

    [​IMG]
    Using pliers to remove the large chunk, it is still significantly magnetic and did not want to pinch myself

    [​IMG]
    In order to clean the internals out properly, I need to access the entire motor can. Lifting up the stickers on both sides reveals four total screws that tap into the plastic endbell.

    [​IMG]
    Endbell removed, there will be plastic pieces from the self tapping screws that comes out

    [​IMG]
    I also noticed this scratch on the commutator, I do believe this was due to me pulling out the armature. I will have to smooth this out for better performance.

    [​IMG]
    Magnet fragments cleaned out to the best of my ability, I see residual epoxy around where the front plate sits. I can epoxy it back on and it should be fine.

    [​IMG]
    Looking around the house, I did not have any 'regular' epoxy laying around. I was worried about the 'steel' in JB weld being drawn into the magnets but a quick test with the motor can showed me that the tubes were not attracted at all. Strange but that means I can use it with no ill effect.

    [​IMG]
    Applied sparingly on mating surfaces

    [​IMG]
    Found a 16mm socket that fits perfectly around the pinion and tower so I am able to clamp down on the front plate itself until the JB weld cures.

    Due to Hawaii weather conditions I found that JB weld takes more than its recommended 24 hours to cure properly, so we will revisit this project on Sunday. With part of the magnet missing the performance will be reduced (lower power) and I will take a spinup measurement video like how I did in the beginning of this post. I do not however have an AEG that accepts medium-type motors so I will have to improvise how to test it out.
     
    Last edited: Sep 11, 2021
  6. Ben3721

    Ben3721 Well-Known Member

    631
    320
    I hate to say this. But I'm willing to bet that the motor was dropped. Not only is that magnet damaged. But the magnetic field could have been knocked off axis by the drop.

    If it was my motor, I'd take the magnets out and replace them with some shs/rocket ones and call it a day. Those looked glued in however, which makes it much harder to swap.
     
  7. aotsukisho

    aotsukisho Well-Known Member

    766
    423
    Unfortunately I don't have equipment that can measure magnetic flux so I am unable to confirm. I don't think it will be unusable especially since the motor still ran when power was applied, but I don't doubt performance decrease mainly due to a significant chunk of magnetic material missing.

    I don't know if those magnets are physically compatible and as you said at least part of the magnets are epoxied into the can so I will not attempt a swap. Tienly motors have a much higher than average flux rating so even swapping with intact 'normal' neo magnets the performance may be lower.

    I would like to have a motor dyno or at the very least a stall torque measurement device but for the time being I will just have to find an AEG that this motor fits into in order to show actual performance.
     
  8. Squad144

    Squad144 Active Member

    158
    32
    Ocean Springs
    How did you get the glue off that was originally holding the can together?
     
  9. Guges Mk3

    Guges Mk3 Administrator Staff Member Administrator Supporting Member Lifetime Supporter

    8,123
    4,621
    Minneapolis
    You can shim a medium into the grip of an AR type AEG. measure motor height difference, cut a spacer and slip it under the motor and above the adjustment.
     
  10. aotsukisho

    aotsukisho Well-Known Member

    766
    423
    After three days the JB weld is guaranteed cured. I loosened the vice and the motor can is solid again.

    [​IMG]
    JB weld visible in the small cap between the can itself and the front plate

    [​IMG]
    Reinstalling the screws that hold the endbell on

    [​IMG]
    Reassembling the endbell components

    [​IMG]
    Break-in time. Running the motor in the intended direction at a low voltage will help mechanically seat the brushes onto the commutator.

    I used a salvaged 1s 4200mAh lipo to do the breakin, ran it in three 20 minute sessions and cooled the motor completely between them. Used an electric duster (think reverse vacuum cleaner) to blow out any accumulated carbon dust from the open endbell.

    [​IMG]
    Nice shiny spot on the commutator now. Those irregular rings from before are gone

    [​IMG]
    https://odysee.com/@digitalcomplex:5/gQKVpEITKx0:2
    youtube.com/watch?v=gQKVpEITKx0
    Reran a similar test as the one before in order to capture peak draw, this time I wanted to show realtime data that I should have done in the first test.

    Tienly 25K medium (repaired, broken in)
    Battery voltage: 8.27v
    Peak amps: 4.71A
    Minimum voltage: 8.17v
    Peak watts: 38.4W

    The battery setup is the same one as before but it wasn't sagging as badly this time which is good. We can also see that the motor draws around 2.5A unloaded. Peak amps (4.71) are slightly higher than the brand new 25K (4.53A) - I attribute this to the brush break-in, which I will do to the 25K long and rerun the wattmeter testing.

    [​IMG]
    Repacked the repaired Tienly until I am able to test it under load.

    In the meantime, I will try to run a similar break-in procedure on the formerly-new 25K long that I have in order to see how it compares to the broken-in repaired 25K medium.

    If you mean the front plate being epoxied to the motor can/magnet housing, I didn't. I wiggled around the motor components as part of a build quality check and found that it was loose. They aren't supposed to separate under normal circumstances as there is no other method of attachment (crimps, screws, etc)

    It's possible yes, but I'm considering chopping the motor shaft and sticking it into a short-type AEG instead. Problem being, I don't have any working test platforms.

    I do have a v3-based Marui MP5K (long motor), do you have any Marui-compatible AK cages laying around Guges? If so I can buy one off of you to throw into my next shipment. This way I can do both long (installed in receiver or with AR grip) and short (gearbox removed with AK cage) testing on the same sealed gearbox.
     
    Richard Paoletti likes this.
  11. Guges Mk3

    Guges Mk3 Administrator Staff Member Administrator Supporting Member Lifetime Supporter

    8,123
    4,621
    Minneapolis
    I will have to forage for the cage due to storage consolidation by the wife....give me a few days, my schedule is 16 hour days until Friday.
     
    aotsukisho likes this.
  12. OutlawAirsoft

    OutlawAirsoft Active Member

    237
    68
    Hamilton
    Fascinating. At least the problem was obvious. There's nothing like an elusive issue.
     
    aotsukisho likes this.