Motor - Balance Data

I noticed the 3.2.0 version of Tower has some diagnostic information that includes a vibration page.

Perhaps use this, in conjunction with the MP motor test to check for the vibes from the Pixhawk 2's own accelerometers? There may also be MP data screens that show the same info from the PH 2.

Also, I wonder if some of what you are measuring are resonances in the Solo innards, shell, or legs, Gimbal, etc?
 
I noticed the 3.2.0 version of Tower has some diagnostic information that includes a vibration page.

Perhaps use this, in conjunction with the MP motor test to check for the vibes from the Pixhawk 2's own accelerometers? There may also be MP data screens that show the same info from the PH 2.

Also, I wonder if some of what you are measuring are resonances in the Solo innards, shell, or legs, Gimbal, etc?
I noticed the Tower diagnostics screen and attempted to view data while on the ground without props, but it appears the bird has to be in flight to start the flow of readings. It is the most valid reference that we would be able to obtain. In flight and at the PixHawk.

Great point about the PixHawk 2 data stream. Should be a part of the t-log data files, will get one of my recent files and have a look. thanks for adding to my research...;)

I agree it could be sourcing from a variety of locations. But again the motors are the point of reference for any resonances created. At least at this time in my discovery. I do believe that all you mentioned has its own tuning and will contribute in various ways - good and bad. Just last night, after FW update, I tested the motors (less props) via FLY and the body started squealing. It turned out to be the battery tray vibrating at the body joint.

I am considering testing a rubberized mastic tape purposed for vibration dampening. Which could be applied in and around the body. The automotive industry utilizes for the same purpose. Again, later once the motors are balanced to a level I feel is reasonable for the effort.

An added thought. When removing the bell housing from the motor I was expecting more resistance to pull the two apart. Magnet pull is one thing, but I was expecting the bearing shaft interface to be tighter. I failed to measure the bearing ID...
 
Last edited:
Finally made time to start balancing motor #4. I started the testing all over since the motor had been taken a part, benchmark was the same as previous test/chart. My weight is electrical tape, I cut about a 1/4 inch and moved it around the motor bell housing. Basically I used the top as the guide dividing into 1/3 around the bell. On my 3rd motor test with tape, my app was showing a very flat line. Ran the test again and the same. I placed another piece over the other and retested, went back up in the readings.

Again the tests were run without props. Running the motor via MP motor test screen at 50% for 2 seconds.

Here are the results for the imbalance isolated. The mean range is now a 0.157 on the Z axis, prior to tape it was 2.00. Still have a lot of work to do, but the overall vibration felt on the Solo's body has been dramatically reduced. Other than typical motor whine, there are little to no harmonic noises being emitted. I call it a success at this time, but reserve the right correct if error was made in my interpretation of the data.

upload_2015-10-9_22-35-54.png
 
  • Like
Reactions: J R
I did get around to balancing all 4 motors this last weekend. Wasn't able to fly until yesterday, so I wasn't sure what to expect. At this point I need to balance props, balancer on order....

Wanted to provide some details of the weighting method and the issues I encountered with the Solo's pods.

The tape was 3M's plastic tape assorted color package. Red, Blue, Green, Yellow and White. It's 3/4" wide. Unlike electrical tape, this plastic tape did not stretch like you would expect. Adhesion is good at this point, one flight of 10 minutes.

I choose three different colors to segment by size, which would then be classified as a given weight. To my surprise the cut segments actual verified to be correctly weighted.

White = 1/8"X3/4" = .01 grams
Yellow = 1/4"x3/4" = .02 grams
Red = 1/2"X3/4" = .04 grams

Motor 1 - .01 grams
Motor 2 - .04 grams
Motor 3 - .005 grams - halved an 1/8 segment
Motor 4 - .06 grams

The Pod issue. Everything improved regarding overall balancing and vibration, further work still to be done. The difficulty was that the motor sits recessed into the motor pod housing. If you need to add weight to the bottom third it was only accessible on the long plate side of the pod. That's the fine tuning part of balancing the motor once you've located the area, move the weight either up or down the bell. With that in mind and the concern whether the tape will adhere. I'd sure hate for the tape to be the cause for a motor stop if it rolled up between the motor and pod housing....
 
Update: Motor pods arrived! Hopefully this will solve my perceived issue with motor #4. Several steps to perform prior to confirming. Initial inspection indicates that these operate smoother when spun by hand. Less magnetic thumping!?!
 
Ok, I found out I have very little patience. Did record some vibration test and then installed the new motor. Noticeable vibration reduction and did not take time to balance. Flew for about a minute and it appears the video juddering has reduced dramatically, however having an over cast day could have slowed down the frame rate to influence the video. Will wait and see before fully confirming.

So the perceived problem with the original motor is that when turning by hand the magnets grabbed at a course spacing. An example, if you spun the motor manually with a prop on, finger pushing the prop around. The prop would jump out ahead of the timing of the movement. My other three motor would track very closely with my finger. And the magnetic pull was tighter spaced. I'm curious if anyone knows what this actually called?

Its hard to explain, but the motor always sounded off when flying hard or compensating for wind. The bird would always shutter when making a rapid vertical flight. Prior to balancing the motor bells, the back end of the bird would shutter with any throttle change. Balancing did improve things, but still the vibration showed up in any videos. What I've called juddering. I think the vibration was the results of how the motor was built. I'm unsure if it relates to the magnets or the windings. Anyone have an idea which it would be?

Look forward to a sunnier condition and can prove to myself that this source of vibration has been reduced or can even be eliminated with minor effort.
 
  • Like
Reactions: J R
Ok, I found out I have very little patience. Did record some vibration test and then installed the new motor. Noticeable vibration reduction and did not take time to balance. Flew for about a minute and it appears the video juddering has reduced dramatically, however having an over cast day could have slowed down the frame rate to influence the video. Will wait and see before fully confirming.

So the perceived problem with the original motor is that when turning by hand the magnets grabbed at a course spacing. An example, if you spun the motor manually with a prop on, finger pushing the prop around. The prop would jump out ahead of the timing of the movement. My other three motor would track very closely with my finger. And the magnetic pull was tighter spaced. I'm curious if anyone knows what this actually called?

Its hard to explain, but the motor always sounded off when flying hard or compensating for wind. The bird would always shutter when making a rapid vertical flight. Prior to balancing the motor bells, the back end of the bird would shutter with any throttle change. Balancing did improve things, but still the vibration showed up in any videos. What I've called juddering. I think the vibration was the results of how the motor was built. I'm unsure if it relates to the magnets or the windings. Anyone have an idea which it would be?

Look forward to a sunnier condition and can prove to myself that this source of vibration has been reduced or can even be eliminated with minor effort.
Completely off topic, but judder is a very specific term which relates to a skipping effect in video when panning to quickly for a given frame rate. Your posts are fascinating and valuable, but it throws me off when you say judder. Sorry to nit pick.
 
Completely off topic, but judder is a very specific term which relates to a skipping effect in video when panning to quickly for a given frame rate. Your posts are fascinating and valuable, but it throws me off when you say judder. Sorry to nit pick.
bitch...;) What should I call it? You get the idea of what I'm describing, correct? The video blurs on the horizontal plain.
 
bitch...;) What should I call it? You get the idea of what I'm describing, correct? The video blurs on the horizontal plain.
LOL. I don't know. And yes I know what you mean. Some kind of vibration. Maybe "shudder"
 
Since your my mentor here with Solo, hence forth I will call it shudder... Good thing MSWord has word replace...thanks.
 
OK, since I've renamed the vibration to Yudders, I can now describe further without offending.

Did finally balance the new #4 motor and it took a .001g piece of tape. The readings were comparable to the other motors now. Still have Yudders, but it's isolating further.

Lets be clear, I'm unsure at this point that motors have anything to do with yudders. The following is just my discoveries and nothing more. The sky is not falling. ymmv.

The original #4 motor, I replaced the bearings in it and was able reduce the balancing weight to .001g rather than the .006g I had on it originally. Learned quite a bit about brushless motors prior to the swap and then more as I installed the new set of bearings. In the end, I succeeded in what I set out to perform and learn.

The short answer, it appears the shaft was spinning within the bearing. At higher RPM the bell will oscillate within the bearing rather than ride on the bearing. This creates an imbalance of the bell housing. I did have a slight tick in the top bearing as well.

Long answer:

First, I never had any issues with motors from my previous brand. They were smooth running. I presumed they are T-Motors. It appears they are one of the better brushless motors out there, as far as mass produced. So I'll use T-Motors as the benchmark for this conversation.

The ticking bearing, it was slight but was easily felt when spinning the bell off the bearing. Not calling it bad. Just in the discovery phase for bearings.

The bearings, the shaft and pre-loading. So here's the deal with bearings in a brushless motor. The bell's shaft does not fit tight to the ID of the bearings. There is a certain tolerance for the fit to allow a non-pressed fit, they slide through. Typically this is resolved by the tolerance in the fit along the shaft. Bearings are spaced by the base housing, the stator actually fits over the base, so the bearing to bearing spacing has fairly close tolerances when properly fitted. Bearing slop is a variable however.

The shaft and the bell housing on the other hand are another variable. The bell housing actually rides on the inner race of the top bearing. The shaft then passes through the lower housing bearing. There is thrust washer and then a funky cir-clip to hold the assembly together.

At this point the shaft length is very critical to how the motor will run. Too long and the motor will have a abnormal sound to it. Basically it is oscillating within one of the bearings rather than spinning the bearings. You'll sometime hear the motor change pitch at a constant motor RPM. This is the shaft wobbling within the bearing. Not good, the shaft will show polish points on the shaft where the bearing would be. This is wear.

Will assume the shaft was installed properly or it has loosened from the bell, as factory. Either way, the thrust washer should pull the shaft once the cir-clip is installed, compressing slightly the bearings together. As I've read it, this is called pre-loading the bearings. Too much pressure and you bind the bearings. Too little and the shaft can wobble.

Not saying the brass thrust washer used is no good, but motor #4 did have a polish on the shaft. Originally it required the heaviest weight of .006 to be used out of the four motors. It now takes the lightest weight, .001g. At this point I've only done this motor. I'm still working on a method to separate the pod from the motor so that I can reassemble like factory.

I used a T-Motor Washer 3.175mm, which appears to be a nylon material as a replacement thrust washer. It was slightly thicker than the brass washer by about 0.3mm. Understanding nylon like I do, it will compress with a certain amount of pressure. Further is will not deform, melt at operating temperatures or wear easily. It provides the correct compression for the bearings, with all other components being correct. At least based on this first install. Preloaded bearings, no shaft oscillations.

Bought the T-Motor cir-clips as well....;) It all fits...like I knew what I was doing...

I bought the bearings through a local dealer/mfg to me here in Houston, seemed to have a following in the RC car world. The stated spec's were spot on for my set of calipers, so tolerance are very good. I would imagine RC cars are more demanding on bearings than drones, so I choose them for the bearings. That's pretty much what they carry btw... And a heck of a lot cheaper than others...Tell them RichWest sent you....maybe I'll get a hat.

Avid Racing Concepts LLC | RC Bearings and Accessories for radio controlled vehicles - shipped Monday AM, recv'd by Tues.
Top 1/8 x 5/16 x 9/64 Metal
Bot 1/8 x 3/8 x 5/32 Metal

Helidirect - these guys took two days to ship, recv'd on Thur.
T-Motor thrust washer and cir-clip...

Tapping the bearings out was no big deal, plenty of videos showing the process. I made a tool to install the bearings, pretty slick actually. Again plenty of videos showing ways to install bearings. Do your research prior to messing with bearings.

FWIW, Again these are just my notes from my discoveries. If I'm wrong, please feel to correct me. I really don't recommend changing the bearings in the Solo's motor pods just yet. it's fairly involved at this point, more so than many here should attempt. FWIW, bearings or motor pods will eventually need replacing. It's part of the game and a wearable part.

If you've read this far, I bet you can't wait till I start discussing my discoveries about Cogging... Again, another thing that T-Motors appears to have reduced in their brushless motors.

Good night and good luck!
 
Last edited:
  • Like
Reactions: erikgraham
OK, since I've renamed the vibration to Yudders, I can know describe further without offending.

Did finally balance the new #4 motor and it took a .001g piece of tape. The readings were comparable to the other motors now. Still have Yudders, but it's isolating further.

Lets be clear, I'm unsure at this point that motors have anything to do with yudders. The following is just my discoveries and nothing more. The sky is not falling. ymmv.

The original #4 motor, I replaced the bearings in it and was able reduce the balancing weight to .001g rather than the .006g I had on it originally. Learned quite a bit about brushless motors prior to the swap and then more as I installed the new set of bearings. In the end, I succeeded in what I set out to perform and learn.

The short answer, it appears the shaft was spinning within the bearing. At higher RPM the bell will oscillate within the bearing rather than ride on the bearing. This creates an imbalance of the bell housing. I did have a slight tick in the top bearing as well.

Long answer:

First, I never had any issues with motors from my previous brand. They were smooth running. I presumed they are T-Motors. It appears they are one of the better brushless motors out there, as far as mass produced. So I'll use T-Motors as the benchmark for this conversation.

The ticking bearing, it was slight but was easily felt when spinning the bell off the bearing. Not calling it bad. Just in the discovery phase for bearings.

The bearings, the shaft and pre-loading. So here's the deal with bearings in a brushless motor. The bell's shaft does not fit tight to the ID of the bearings. There is a certain tolerance for the fit to allow a non-pressed fit, they slide through. Typically this is resolved by the tolerance in the fit along the shaft. Bearings are spaced by the base housing, the stator actually fits over the base, so the bearing to bearing spacing has fairly close tolerances when properly fitted. Bearing slop is a variable however.

The shaft and the bell housing on the other hand are another variable. The bell housing actually rides on the inner race of the top bearing. The shaft then passes through the lower housing bearing. There is thrust washer and then a funky cir-clip to hold the assembly together.

At this point the shaft length is very critical to how the motor will run. Too long and the motor will have a abnormal sound to it. Basically it is oscillating within one of the bearings rather than spinning the bearings. You'll sometime hear the motor change pitch at a constant motor RPM. This is the shaft wobbling within the bearing. Not good, the shaft will show polish points on the shaft where the bearing would be. This is wear.

Will assume the shaft was installed properly or it has loosened from the bell, as factory. Either way, the thrust washer should pull the shaft once the cir-clip is installed, compressing slightly the bearings together. As I've read it, this is called pre-loading the bearings. Too much pressure and you bind the bearings. Too little and the shaft can wobble.

Not saying the brass thrust washer used is no good, but motor #4 did have a polish on the shaft. Originally it required the heaviest weight of .006 to be used out of the four motors. It now takes the lightest weight, .001g. At this point I've only done this motor. I'm still working on a method to separate the pod from the motor so that I can reassemble like factory.

I used a T-Motor Washer 3.175mm, which appears to be a nylon material as a replacement thrust washer. It was slightly thicker than the brass washer by about 0.3mm. Understanding nylon like I do, it will compress with a certain amount of pressure. Further is will not deform, melt at operating temperatures or wear easily. It provides the correct compression for the bearings, with all other components being correct. At least based on this first install. Preloaded bearings, no shaft oscillations.

Bought the T-Motor cir-clips as well....;) It all fits...like I knew what I was doing...

I bought the bearings through a local dealer/mfg to me here in Houston, seemed to have a following in the RC car world. The stated spec's were spot on for my set of calipers, so tolerance are very good. I would imagine RC cars are more demanding on bearings than drones, so I choose them for the bearings. That's pretty much what they carry btw... And a heck of a lot cheaper than others...Tell them RichWest sent you....maybe I'll get a hat.

Avid Racing Concepts LLC | RC Bearings and Accessories for radio controlled vehicles - shipped Monday AM, recv'd by Tues.
Top 1/8 x 5/16 x 9/64 Metal
Bot 1/8 x 3/8 x 5/32 Metal

Helidirect - these guys took two days to ship, recv'd on Thur.
T-Motor thrust washer and cir-clip...

Tapping the bearings out was no big deal, plenty of videos showing the process. I made a tool to install the bearings, pretty slick actually. Again plenty of videos showing ways to install bearings. Do your research prior to messing with bearings.

FWIW, Again these are just my notes from my discoveries. If I'm wrong, please feel to correct me. I really don't recommend changing the bearings in the Solo's motor pods just yet. it's fairly involved at this point, more so than many here should attempt. FWIW, bearings or motor pods will eventually need replacing. It's part of the game and a wearable part.

If you've read this far, I bet you can't wait till I start discussing my discoveries about Cogging... Again, another thing that T-Motors appears to have reduced in their brushless motors.

Good night and good luck!
Points for humor.
 
There appears to be some confusion regarding bearing sizes for the Solo Motor Pods. Here are the dimension for what I purchased. They are classified as Standard or Fractional bearings;

Top bearing -
  • STANDARD: 0.125" (Inner) x 0.3125" (Outer) x 0.140625" (Width)
  • FRACTION: 1/8" (Inner) x 5/16" (Outer) x 9/64" (Width)
  • METRIC: 3.175mm (Inner) x 7.9375mm (Outer) x 3.57188mm (Width)
Bottom bearing -
  • STANDARD: 0.125" (Inner) x 0.375" (Outer) x 0.15625" (Width)
  • FRACTION: 1/8" (Inner) x 3/8" (Outer) x 5/32" (Width)
  • METRIC: 3.175mm (Inner) x 9.525mm (Outer) x 3.96875mm (Width)
I went with Caged Metal bearings from Avid RC. YMMV

Metal Bearings
Top - Avid Racing Concepts LLC | RC Bearings and Accessories for radio controlled vehicles
Bottom - Avid Racing Concepts LLC | RC Bearings and Accessories for radio controlled vehicles

Ceramic Bearings

Top - Avid Racing Concepts LLC | RC Bearings and Accessories for radio controlled vehicles
Bottom - Avid Racing Concepts LLC | RC Bearings and Accessories for radio controlled vehicles

No need for Sealed bearing unless you fly in a dirty environment, they add friction to their running.. Even though Ceramic bearings are better, they require constant maintenance to perform as intended. Pods prevent ease of maintenance. Again, your mileage may vary.
 
Last edited:
  • Like
Reactions: CS223
@Doug Benson - Wow, you caught me reading through the thread.... Same App, seems to work well but pain in the butt to record the data. Wish someone would produce a spreadsheet to filter the recorded data into a convenient chart and calculate the Z RMS...;)

I'm on a huge learning curve with much of this. The notes allow me to redirect back on point, when I'm ready. The motor swap thread by @RolandS888 has got me in a hold pattern regarding motors. I have nearly everything to swap bearings in all the motors that I have, just need to build the tool for reforming the plastic board tabs to factory.

Good luck!
 
In the end here is what I found.
Test each motor and balance
if you have a motor that is just way out and will not balance replace it.
Set a reference on the frame itself with all four motors at 50
then try and get that number as low as you can by balancing and or replacing really bad pods
It worked for me
 
In the end here is what I found.
Test each motor and balance
if you have a motor that is just way out and will not balance replace it.
Set a reference on the frame itself with all four motors at 50
then try and get that number as low as you can by balancing and or replacing really bad pods
It worked for me
Would you be willing to sell off a rejected motor or two to the cause? All I want is the motor, no ESC or plastic housing....still gives you the ability to upgrade a motor later....

I'm wanting to tear down a motor completely and would like to keep my #4 intact as the reference since so much data has been collected on it.
 
Would you be willing to sell off a rejected motor or two to the cause? All I want is the motor, no ESC or plastic housing....still gives you the ability to upgrade a motor later....

I'm wanting to tear down a motor completely and would like to keep my #4 intact as the reference since so much data has been collected on it.
For now I am holding on to them. If I find the time I want to try replacing the bearings just as a project
 
  • Like
Reactions: RichWest
For now I am holding on to them. If I find the time I want to try replacing the bearings just as a project
If you could do it, let us know and what tools need to be changed. And the exact links to these Bearings. We will be very grateful!
 

Members online

No members online now.

Forum statistics

Threads
13,093
Messages
147,741
Members
16,047
Latest member
pvt solo