When to replace motors?

[JoelNewman]

There are some motor balancing videos on YouTube where the guy uses some ubur-technical software to test the motors. Might be what your looking for. I'll try to find one when I can.

Yeah that would be interesting to see what they're measuring.

Joel I don't think vibration is a reasonable indicator for a failing motor or bearings for that fact. You'd have to have a benchmark reading to compare from the same motor when new. But even then you'd have to perform the test off the body to insure no further influences to slight further the test.

Yeah that's a good point actually - I should've known this as I (well actually Roland) have been through the process of balancing them. Luckily I do have some data from when Roland did the balancing, and motor 1 was in the same range as all the others when he was finished.

Amperage would be ideal, but again at the refresh rate and resolution from MP it would be questionable to see slight issues that would indicate failing motor or bearing. However I have never tried.

I vaguely recall seeing a log file that recorded current draw, or at least RPM of each motor, but I haven't been able to find that in any Pixhawk logs. I suspect that may have been a MicroDrone that I used to work with.

Does (or could) the Pixhawk log the desired and actual RPM of each motor and/or the current draw? Does the ESC give feedback on anything?

Would it work to feed a controlled voltage and current into the motor and measure the RPM it's achieving?

In the commercial realm, it will be important to have such a means to determine motor or bearing life. If they can't validate a maintenance program, then the industry is doomed. I'd imagine at present, chinese built motors have very little quality control once the motors are built.

Yeah and that's exactly why I'm interested in this. I am about to get the certification for my own company in Australia, and I have a Solo that may or may not drop out of the sky at any point. It's fine for me to keep an eye on it at the moment, but when my plan for global domination kicks in and I have 100 aircraft buzzing around I'll need something a bit more reliable to tell me when key components like this are going to fail.

 

[ChrisBird]

Yeah that would be interesting to see what they're measuring.



Yeah that's a good point actually - I should've known this as I (well actually Roland) have been through the process of balancing them. Luckily I do have some data from when Roland did the balancing, and motor 1 was in the same range as all the others when he was finished.



I vaguely recall seeing a log file that recorded current draw, or at least RPM of each motor, but I haven't been able to find that in any Pixhawk logs. I suspect that may have been a MicroDrone that I used to work with.

Does (or could) the Pixhawk log the desired and actual RPM of each motor and/or the current draw? Does the ESC give feedback on anything?

You will have the total current (not per motor) and the pwm output for each motor. I believe they are CANBUS esc so they have ability to feedback info but at present im not sure how much of that is logged in the apm dataflash logs.



Sent from my SM-G930F using Tapatalk

 

[Maddog]

I found the videos I was thinking about but they are dynamic prop balancing, not motor balancing.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[JoelNewman]

I found the videos I was thinking about but they are dynamic prop balancing, not motor balancing.

That's pretty amazing stuff, but yeah not much good for the problem at hand.

 

[JoelNewman]

You will have the total current (not per motor)...

Thank you, that comment got me thinking and I graphed the total wattage draw, but while testing just a single motor at a time at 100% throttle with no props.

The graph is below - motors 1, 2, 3, 4, and motor 1 at the end again as a control to make sure that battery drain wasn't affecting results.

Background power consumption is 9 watts, the other three motors pull 9 watts, and motor 1 pulls 12 watts, or about 30% more.




The same relationship is shown in the graphs of current but the numbers are nicer in wattage.

 

[RobertT]

Yeah that would be interesting to see what they're measuring.



Yeah that's a good point actually - I should've known this as I (well actually Roland) have been through the process of balancing them. Luckily I do have some data from when Roland did the balancing, and motor 1 was in the same range as all the others when he was finished.



I vaguely recall seeing a log file that recorded current draw, or at least RPM of each motor, but I haven't been able to find that in any Pixhawk logs. I suspect that may have been a MicroDrone that I used to work with.

Does (or could) the Pixhawk log the desired and actual RPM of each motor and/or the current draw? Does the ESC give feedback on anything?

Would it work to feed a controlled voltage and current into the motor and measure the RPM it's achieving?



Yeah and that's exactly why I'm interested in this. I am about to get the certification for my own company in Australia, and I have a Solo that may or may not drop out of the sky at any point. It's fine for me to keep an eye on it at the moment, but when my plan for global domination kicks in and I have 100 aircraft buzzing around I'll need something a bit more reliable to tell me when key components like this are going to fail.

Joel - going through same process in NZ. As the OEM manual says they are lifed at 150 hours I think you will be hard pressed to get your Regulator to accept any sort of unproven analysis programme. I have dealt with the Aussie regulators in the Oil and Gas sector and doubt CASA is different. I guess what I'm saying is that if you want to get your docs approved quickly just follow OEM lifed time......

 

[RobertT]

Oh and good luck with the world domination!

 

[JoelNewman]

Joel - going through same process in NZ. As the OEM manual says they are lifed at 150 hours I think you will be hard pressed to get your Regulator to accept any sort of unproven analysis programme. I have dealt with the Aussie regulators in the Oil and Gas sector and doubt CASA is different. I guess what I'm saying is that if you want to get your docs approved quickly just follow OEM lifed time......

Yeah the CASA model is the same - they expect all of the operator maintenance manuals to just refer back to the manufacturer manual.
This analysis (and subsequent monitoring) is more for my own piece of mind. I don't want my Solo to fall out of the sky on my first job under my own OC!

 

[RobertT]

Yeah the CASA model is the same - they expect all of the operator maintenance manuals to just refer back to the manufacturer manual.
This analysis (and subsequent monitoring) is more for my own piece of mind. I don't want my Solo to fall out of the sky on my first job under my own OC!

I have only ever used vibration analysis on large motors as any sort of indication of failure. Some DP components have Mean Time to Failure indications BUT still a component that has a MTF of 10000hrs MAY fail in hour 1. As Solo is a new airframe there is no experience of time based motor failure (doesn't count if you ram them into a tree on hour 50 as one of my lads did with the DJI Inspire Pro last week). WE have said that for all the manufacturers we will run OEM hours with regular PM and inspections pre and post flight and over time try to build a case to extend the lifed time with the CAA.

 

[RobertT]

I have only ever used vibration analysis on large motors as any sort of indication of failure. Some DP components have Mean Time to Failure indications BUT still a component that has a MTF of 10000hrs MAY fail in hour 1. As Solo is a new airframe there is no experience of time based motor failure (doesn't count if you ram them into a tree on hour 50 as one of my lads did with the DJI Inspire Pro last week). WE have said that for all the manufacturers we will run OEM hours with regular PM and inspections pre and post flight and over time try to build a case to extend the lifed time with the CAA.

And an aside (which might get me flamed) the biggest risk to customer satisfaction with the Solo is not it failing AFTER it is airborne it's getting it airborne in the first place....once it's up I have had no issues with it. Could say that about several things I guess.

 

[SurfnSkate81]

@RobertT how would you say 3DR goes vs DJI for reliability in your experience?


Sent from my iPhone using Tapatalk

 

[RobertT]

@RobertT how would you say 3DR goes vs DJI for reliability in your experience?


Sent from my iPhone using Tapatalk

I will be 100% honest. I started with the Solo in Malaysia as a hobby UAV and it morphed into something else. We now have 3 x DJI 1 x Solo and 1 x Parrot. We fly the units according to the mission. Till we fitted the GPS shield on the Solo the DJI units were way more reliable. Now we can get a good lock repeatably the Solo is back in the field. Last week the DJI Inspire Pro was in contact with a tree. The damage was extensive, camera wrecked, gimbal wrecked.....Solo would have survived WAY better!! Phantoms are the Land Rovers of UAV in my opinion. They do the business with the 4k camera when we are videoing solar panel installations and when my daughter uses her 3 Standard to let her class practice coding and they fly it into walls it just grunts and get up again and again and again.....its horses for courses!

 

[Maddog]

Last week the DJI Inspire Pro was in contact with a tree.

Funny- sounds like something a politician would say after somebody flew it into a tree.

 

[RobertT]

Diplomacy! In fact we were flying a hole at a golf course. Our SOP calls for the Observer to ensure no-one speaks to the Controller or distracts them as well as watch the aircraft. In this case the Observer (me) allowed the Controllers very lovely fiance to be whispering in his (the Controllers) ear as he (the Observer) was talking to two old dudes waiting to tee off.......moral? follow your SOP!! But the contact in no way deserved the damage done! Solo we would have picked her up and soldiered on!

 

[Maddog]

Diplomacy! In fact we were flying a hole at a golf course. Our SOP calls for the Observer to ensure no-one speaks to the Controller or distracts them as well as watch the aircraft. In this case the Observer (me) allowed the Controllers very lovely fiance to be whispering in his (the Controllers) ear as he (the Observer) was talking to two old dudes waiting to tee off.......moral? follow your SOP!! But the contact in no way deserved the damage done! Solo we would have picked her up and soldiered on!

Politician= diplomat, no?
Interesting that the Inspire can't take much of a beating, but I guess it's to be expected with retractable (fragile) landing gear.

 

[JoelNewman]

I sent my logs to support, and they assessed it using the PWM signal to the motors, on the assumption that the flight controller will compensate for a motor that has higher resistance by sending it a higher PWM signal to keep the aircraft stable.
Anyway think what they must be doing is looking at RCOUT and comparing channels 1 to 4. In a hover with no wind the front pair, 1 and 3, should have similar PWM signals and the back pair should have similar signals, depending on how your payload is balanced. My dodgy motor is slightly higher than it's pair under load, but obviously not enough for them to worry about.

So going back to the OP who was looking for a motor replacement protocol, I would do a test flight as part of my periodic maintenance check and do some excel wizadry on the log to determine if there's any difference in the PWM signal. I don't know what sort of numbers are a problem, but at least it gives you some quantitative data to start playing with, and an audit trail.

 

[reelcrazyfish]

Found this log while diagnosing a possible motor failure. Any new info on this? Last post is almost a year old. Are folks replacing motors on a timed basis hoping to replace them before they fail, or is there a better way?

 

[andrew.baker.142]

I think 3DR expected a 150 hour life span for the motors, assuming they are treated well.
That's a bunch of flights!

 

[Maddog]

I haven't heard anything. I think everybody is just "winging it".
I'm not sure how many people are actually tracking their motor hours.

 

[reelcrazyfish]

I had what I presume is a failed motor on Saturday and I have far less than 150 hour flight time on the drone. I submitted a ticket to 3dr but haven't heard back yet.