JGFly Posted at 5-1 17:40
Yes, I must admit ...

.... But I would like to see DJI publish an official warning about defective propellers and at the very least apologize for the problems incurred and reimburse all those who had to obtain other propellers.

We can hope.

"It may be that the propellers have been modified in the meantime"

...no visible/tangible diffs...?

Myself have original and extra props from Flymore Combo (January)...and 3 new DJI sets from early March....

Cannot see or feel any diffs in texture or pliability or visible deformities.

I cannot rule out props problems in RPM diffs but I guess it's a combination of multiple factors.

hallmark007 Posted at 5-2 00:35
First I was right you are head of the Stasi and a bigger fool with it, you never asked me twice about anything. When you want to address someone on this forum there is a reply button, check again you’ll find you are once again wrong. You’re now trying to say others are posting stuff that’s wrong something you constantly do yourself.
Again making up lies to try prove your idiotic points , I NEVER told anyone to ignore any warning, I don't go around here posting complete rubbish about changing your Rth settings in the middle of Rth just in case the sky falls in, which was complete utter rubbish, then again you're advising others that Rth now works on mini in C mode, you're a total crack pot who thinks he can foist his jaundiced ideas on others and the troll them .

You say, "you never asked me twice about anything".   Are you blind?  It's right here on the same page!

• Post#126 - "Do you have a link to DJI's response?  Where can we see that?"
• Post#149 - "And I again ask, where can we see for ourselves this reply from DJI?"
  

You wrote that I, "go around here posting complete rubbish about changing your Rth settings  in the middle of Rth just in case the sky falls in, which was complete  utter rubbish, then again you're advising others that Rth now works on  mini in C mode, you're a total crack pot"


I won't even try to explain to you again why there are situations in which one might choose to  alter the Failsafe setting, since you have so amply demonstrated yourself to be completely incapable of  understanding such simple concepts.  I'll just post the same video here again so that others can decide for themselves who is the total crack pot here.  

Pay close attention at 2:20 in this video.  The Mini is currently in C-mode when I select Return-to-Home.  The video clearly shows that the Mini remains in C-mode.  If you have uninterrupted control signal while the Mini is RTH mode, you can choose to change to P-mode or even S-mode or even back to C-mode.  You can even choose to cancel RTH and resume manual control of the drone.  The point is, as long as the control signal is uninterrupted, it will stay in whichever mode it's currently in when you manually invoke RTH.

Ergo: Rth certainly does work in C-mode.


Furthermore, I demonstrate with actual video proof the fact that the Mini always powers up initially in P-mode (See start of video).  Also, whenever Return-to-Home is automatically invoked as a result of Loss-of-Control-Signal [I switched the controller OFF], it is a fact that the Mini automatically switches to P-mode.  (See starting at 4:40) The Mini is in C-mode when I switch the controller off to trigger RTH.  At 5:35 when the control signal is regained the Mini is now in P-mode.

Which part of that have you failed to understand?

Zbip57 Posted at 5-2 04:38
You say, "you never asked me twice about anything".   Are you blind?  It's right here on the same page!

Post#126 - "Do you have a link to DJI's response?  Where can we see that?"

Ok...we get your point...thank you and let's be constructive here...g'day.

lee82gx Posted at 5-2 04:05
[view_image]

There is much less speed discrepancy between front and rear motors my Mavic Pro. I happen to have logs which are used for power consumption checking but the only big speed discrepancy comes when aircraft has severe tilted due to acceleration or braking. see my second log for hover:

Thank you very much!

The first thing I did was go for another flight with last yesterday's setup, I wanted fresh data for today's testing.
This setup for those who don't want to look back, I'm running the Original front props on the rear motors and a replacement set of props on the front motors. I'm going to be doing a different test from yesterday's hover and up/down/fwd/back, I wanted a flight path that would be similar to a real world example for me.

Here's the flight path we will be taking:


Here's the motor graph for our first flight, I hover for one minute and then start my flight and everything looks good here.


Let's take a look at the flight log in FRAP to be sure everything is fine.



Now for actual testing, These are my original REAR Props from Day 0. I used a 50lb weight on top of some books to flatten the props for 3 almost 4 hours. Same test as above, it starts with a hover and then 2 laps around my tree over the pond.



This doesn't look to bad, some deviation has occurred but no errors in the logs or any noticeable performance difference.

FRAP of this flight:


I wanted to ramp things up, I suppose if I wanted to keep things more realistic I would of put the props back and let them sit overnight. Instead I put them back between the books and I proceeded to stand on them for 5 to 10 minutes. This had an immediate effect, so much so I couldn't take off because the props were too deformed.




So I tried what DJI suggests and "Smoothed out" my props, this worked and let me take off but the drone has a different sound and was visually performing harder.



The above video ends because I was in P mode and had to stop recording to switch to Sport.

At 22 seconds in this video you will see I fly up to about 60 feet and hover in the wind for a few seconds before dropping the throttle and waiting for the drone to begin descending and ramp back to full throttle to cause the drone to "bounce" but the drone didn't respond and proceeded to "fall" all the way to the ground where after a second I regained control and it lifted back off. In the logs you can see I'm 100% up stick and the drone still descends and the motor RPM is a constant 5k during this descent. Enough talking let's review the video.



Here's the motor rpm graph from the above flight:

In this graph I have enabled Barometer for Altitude readings and RC Throttle to show I was indeed 100% left stick as the drone descended.


Let's dig a little deeper and check FRAP:



Would you look at that, I thought I had one loss of altitude but FRAP clearly shows 2 drops in altitude. I'm not good at analyzing FRAP data so if JJB finds his way into this thread I would be happy if he could take a look and share his 2cents. .Dat Link

This is the whole point of doing research, science. To test, and re test, if you know X equals Y result keep repeating X until Y changes. Figure out why it had a different effect, log the information, and test again. It's easy to say this that or the other on a forum, but how about you actually contribute and test for yourself and find out if it does in fact have nothing to do with Z.

Here's what I know from testing:
The rear motors work harder when flying forward.
Prop deformities can be seen on RPM information.
If props get to deformed uncommanded descents can happen.
Not enough force/Max power load reached has relation to prop deformities.
Deformed props will strike the body.
Deformed props will effect/prevent take off.
Good props have a different sound profile, deformed props have a can be heard.

I'm sure there's more but this is what I've found from my research.

Now let us review the pictures from the props used above, I want you to tell me if you can visually see any difference outside of the "Rounded/raised edges" and a few micro dings/chips. I don't see any difference between the good props and the bad props (Marked with marker) outside normal prop damage from grazing things.




Close up:





Profile View:






Underside:








Weight | New vs Orginal:




I didn't take pictures of all the original props, I started with that set and I can't see any abnormalities even against the new props so I didn't feel it was necessary to take pictures of the others. This is enough to make my point on visually identifying a bad prop or the lack of I should say, this goes to show you can't trust your eyes when the RPM data don't lie.

Here's the .Dat file Click Me and here's the flight log from Air Data and Phantom Help.

ABeardedItalian Posted at 5-2 05:24
The first thing I did was go for another flight with last yesterday's setup, I wanted fresh data for today's testing.
This setup for those who don't want to look back, I'm running the Original front props on the rear motors and a replacement set of props on the front motors. I'm going to be doing a different test from yesterday's hover and up/down/fwd/back, I wanted a flight path that would be similar to a real world example for me.

...freaggin' amazing work...!

Zbip57 Posted at 5-2 04:38
You say, "you never asked me twice about anything".   Are you blind?  It's right here on the same page!

Post#126 - "Do you have a link to DJI's response?  Where can we see that?"

Again maybe I’ll have to excuse you because it seem you’re fond of getting things wrong post 126 you should recheck, your ridiculous spouting about Rth in C mode is ridiculous, if you lose signal in any mode your craft will return in P mode and use the speed of P mode not C mode and that’s why it’s both wrong and careless to tell others that their craft will return in C mode unless the switch off RC. That’s just plain wrong. And it’s the same if you’re flying I C mode hit Rth you’re craft will return showing the C but using the speed and pitch of P mode.

hallmark007 Posted at 5-2 05:48
Again maybe I’ll have to excuse you because it seem you’re fond of getting things wrong post 126 you should recheck, your ridiculous spouting about Rth in C mode is ridiculous, if you lose signal in any mode your craft will return in P mode and use the speed of P mode not C mode and that’s why it’s both wrong and careless to tell others that their craft will return in C mode unless the switch off RC. That’s just plain wrong. And it’s the same if you’re flying I C mode hit Rth you’re craft will return showing the C but using the speed and pitch of P mode.

...let it be Hall...lol...we need your knowledge here bud

ABeardedItalian Posted at 5-2 05:24
The first thing I did was go for another flight with last yesterday's setup, I wanted fresh data for today's testing.
This setup for those who don't want to look back, I'm running the Original front props on the rear motors and a replacement set of props on the front motors. I'm going to be doing a different test from yesterday's hover and up/down/fwd/back, I wanted a flight path that would be similar to a real world example for me.

These uncommanded descents are a bit different than others that I've been investigating. I think yours are better described as the actual ascent/descent lagging behind the commanded ascent/descent. I assume you referring to theses points in the flight where the the MM is descending but the throttle is full up?


But, if you look at the CtrlVelVert vel_cmd and vel_fdbk signals it can be seen that although the vertical velocity is negative it is increasing and catching up to the commanded and just lagging behind the commanded.



This also shows that the actual actual ascent/descent lags behind by about 1 sec the commanded ascent/descent. This lag isn't normal - I suspect that your deformed props have a lot to do with it.


Here is an example of an uncommanded descent where the altitude is going opposite from the commanded. I.e., the actual isn't lagging behind the commanded.

fans1cafe718 Posted at 5-2 05:54
...let it be Hall...lol...we need your knowledge here bud

I think the right person is above and he’s Bud. Lol.....

hallmark007 Posted at 5-2 05:48
Again maybe I’ll have to excuse you because it seem you’re fond of getting things wrong post 126 you should recheck, your ridiculous spouting about Rth in C mode is ridiculous, if you lose signal in any mode your craft will return in P mode and use the speed of P mode not C mode and that’s why it’s both wrong and careless to tell others that their craft will return in C mode unless the switch off RC. That’s just plain wrong. And it’s the same if you’re flying I C mode hit Rth you’re craft will return showing the C but using the speed and pitch of P mode.

Hallmark wrote '' ....If you're flying C mode it Rth you're craft will return showing the C but using the speed and pitch of P mode''

Taken note and registered on my '' Good to Know '' list. (I can't trust my memory cause my age)

Thanks!

hallmark007 Posted at 5-2 07:52
I think the right person is above and he’s Bud. Lol.....

...your exp opinions are always of intersest...if you no mind...humbly said, I am an old salt pilot...but noob to drones and your inputs are well appreciated...

ABeardedItalian Posted at 5-2 05:24
The first thing I did was go for another flight with last yesterday's setup, I wanted fresh data for today's testing.
This setup for those who don't want to look back, I'm running the Original front props on the rear motors and a replacement set of props on the front motors. I'm going to be doing a different test from yesterday's hover and up/down/fwd/back, I wanted a flight path that would be similar to a real world example for me.

Very good test, thank you for taking the time to do it. Grazie.

120ccpm Posted at 5-2 17:52
Very good test, thank you for taking the time to do it. Grazie.

I agree also.

I'm new to MM so,  I read everything that is said on the Forum and I put all your analyzes and comments in a file With sub files and keywords and when I will be able to fly the MM (not at this time), in the event of any problem, I will quickly have access to the solutions if they exist.

Thank you very mutch to all of you !

Cheers!

Some of the thread contibutors may find this of interest.
https://mavicpilots.com/threads/mavic-mini-uncommanded-descent-tests.86237/page-6#post-983454

BudWalker Posted at 5-4 06:27
Some of the thread contibutors may find this of interest.
https://mavicpilots.com/threads/mavic-mini-uncommanded-descent-tests.86237/page-6#post-983454

Absolutely wonderful!

What happens now though? I assume compile the information and send it to DJI but do they listen to things like this, have they in the past? I feel like this gives us the answer so we can feel better, but it doesn't actually help anyone if DJI doesn't provide a fix.

ABeardedItalian Posted at 5-4 07:57
Absolutely wonderful!

What happens now though? I assume compile the information and send it to DJI but do they listen to things like this, have they in the past? I feel like this gives us the answer so we can feel better, but it doesn't actually help anyone if DJI doesn't provide a fix.

Kind of a tough call, since it affects less than 1% of the users.

ABeardedItalian Posted at 5-4 07:57
Absolutely wonderful!

What happens now though? I assume compile the information and send it to DJI but do they listen to things like this, have they in the past? I feel like this gives us the answer so we can feel better, but it doesn't actually help anyone if DJI doesn't provide a fix.

Almost certainly DJI knows about this issue. I'm not going to guess what or if they will do anything about it.

Deucalion Posted at 5-4 08:03
Kind of a tough call, since it affects less than 1% of the users.

Is that tongue-n-cheek humor, or is that info DJI has put out (and I missed)?

Through out multiple threads on subject of props (bad blades), several users have reported updating FW of getting warning about bad blades (left rear motor).  After changing prop, they still get a warning, and concluded FW must be bad.

It is possible their spare prop (ccw?) for left rear motor was part of bad batch (or batches).   Other than possibility of looking at whether blade was lost it's twist (pitch) and lays flat on a surface, as pointed out in another thread, user would need to try at least another spare prop (ccw?) for left rear motor.  Maybe moving right front prop (ccw?) to left rear motor?

Again, DJI needs to start communicating with their customers on problem.  Only thing thus far is one user reporting DJI Tech's response to him, in which DJI does not believe FW is problem.   Which would point back to bad batch (or batches) of (ccw?) blades.

HedgeTrimmer Posted at 5-4 08:52
Is that tongue-n-cheek humor, or is that info DJI has put out (and I missed)?

Well, I assume we see a representative number of the uncommanded descents here and on MavicPilots. And DJI sells over a million drones a year, and they must have sold 100k so far, at least, of the mini. And now with. the FW update warning people to change the props, it will be even rarer. Of course, there will be some users who see the "Change props Message" and will be stuck in a exessential crisis, but that will be pretty rare.

Anyways, not tongue in cheek. Back of the napkin educated guess.

Deucalion Posted at 5-4 09:12
Well, I assume we see a representative number of the uncommanded descents here and on MavicPilots. And DJI sells over a million drones a year, and they must have sold 100k so far, at least, of the mini. And now with. the FW update warning people to change the props, it will be even rarer. Of course, there will be some users who see the "Change props Message" and will be stuck in a exessential crisis, but that will be pretty rare.

Anyways, not tongue in cheek. Back of the napkin educated guess.

I think your guess is pretty good, for over 3 months forum was littered with complaints about uncontrolled descent, since new FW two weeks ago, none not a single one, sometimes I think those particularly those who don’t own a mini are not happy until they’re problems, seems some can’t have enough minis falling out of the sky.

I think more reporting on how people are getting on since new FW might make a pleasant change, until something else goes wrong.

It’s much thanks to all who give up their time to try different testing if only just to make hobby more interesting.

Deucalion Posted at 5-4 09:12
Well, I assume we see a representative number of the uncommanded descents here and on MavicPilots. And DJI sells over a million drones a year, and they must have sold 100k so far, at least, of the mini. And now with. the FW update warning people to change the props, it will be even rarer. Of course, there will be some users who see the "Change props Message" and will be stuck in a exessential crisis, but that will be pretty rare.

Anyways, not tongue in cheek. Back of the napkin educated guess.

I'm sure DJI has sold that many to retailers but I highly doubt that's how many own a mini, watching stores like B&H around Christmas, they weren't shipping thousands but hundreds. Same for the other retailers who's orders came in on time, I'd say the mini has a user base of 30/40 thousand, it's not the end all be all drone and while it's price point does make it more entry level DJI got to where they are from drones priced like the Mavic and Phantom so I doubt even with the flood of entry level flyers that it's anywhere near 100k otherwise Ebay would be crazy saturated.

Moving on, even if it effects 1% of the flyers this isn't a hiccup anymore. The Flight Controller is programed to work like this, so at some point it could effect Anyone. That's a lot more of a risk then 1%, for every mini flying has this looming over head that the FC could just decide Nope.

I did a test with my MM, hovering indoor for about 10 minutes, without ever touching the controls (literally: on > auto takeoff > hover > auto landing > off). I then looked at the DAT and averaged out the RPMs for the 4 motors, to have a baseline. NOTE: I'm running the original props and I never had the Motor Speed Error warning, but my props are definitely not new, so this should be considered my own personal baseline, not a reference for all MMs.

The chart is below, but here some interesting data points, if someone wants to do the same test:

Right Front Average: 9363 RPMs
Left Front Average: 8995 RPMs
Right Back Average: 9479 RPMs
Left Back Average: 9913 RPMs

[EDIT 05/14/20: More accurate averages are at post #194]

120ccpm Posted at 5-13 15:54
I did a test with my MM, hovering indoor for about 10 minutes, without ever touching the controls (literally: on > auto takeoff > hover > auto landing > off). I then looked at the DAT and averaged out the RPMs for the 4 motors, to have a baseline. NOTE: I'm running the original props and I never had the Motor Speed Error warning, but my props are definitely not new, so this should be considered my own personal baseline, not a reference for all MMs.

The chart is below, but here some interesting data points, if someone wants to do the same test:

That's some interesting information you have, thank you for sharing. I noticed that like a lot of mini's the Rear Left always seems to be the outlier in these sorts of graphs, not enough to be concerned but enough to show that something is clearly happening to the Rear Left especially.

ABeardedItalian Posted at 5-13 16:57
That's some interesting information you have, thank you for sharing. I noticed that like a lot of mini's the Rear Left always seems to be the outlier in these sorts of graphs, not enough to be concerned but enough to show that something is clearly happening to the Rear Left especially.

Yes, interesting numbers... CCW motors (Right Front and Left Back) seem to always spin faster than the CW ones (+4.3% in my case). Same goes for rear motors vs. front (+5.6% in my case). The combination of these two factors causes the Left Back motor to spin faster than the others, and maybe that's why we see more Motor Speed Error warnings on it. Hard to tell, especially because - as I said - my props aren't new, so some of these RPM differences might be the effect (of a flat prop), not the cause.

120ccpm Posted at 5-13 15:54
I did a test with my MM, hovering indoor for about 10 minutes, without ever touching the controls (literally: on > auto takeoff > hover > auto landing > off). I then looked at the DAT and averaged out the RPMs for the 4 motors, to have a baseline. NOTE: I'm running the original props and I never had the Motor Speed Error warning, but my props are definitely not new, so this should be considered my own personal baseline, not a reference for all MMs.

The chart is below, but here some interesting data points, if someone wants to do the same test:

Here's another graph for comparison.  Same thing, flying indoors, auto-takeoff, hover for 2 minutes, auto-land.    Never touched the control sticks.

I did receive the Motor Speed warning (left-rear) the first time I flew with the new firmware.  I replaced only the left-rear prop blades, as instructed by the warning message.

This graph is with the original props still on the other three motors, and new blades on the left-rear.  Note the CCW motors, left-rear (blue) and right-front (red), are now spinning slower than the CW motors right-rear (yellow) and left-front (green).

Now I'm tempted to replace the right-rear prop blades as well to see how that affects the right-rear motor speed (yellow).

I went ahead and replaced the right-rear prop blades.  Here's the new graph.  Same deal, flying indoors, auto-takeoff, hover for two minutes, auto-land, never touched the control sticks.

Compare the right-rear motor speed (yellow) now with fresh prop blades installed versus the original blades (yellow) in the previous graph above.

I cannot visibly detect or feel any physical difference between the old vs new right-rear props.  The originals are not knicked or dinged in any way.

When I earlier replaced the left-rear blades I could see the original blades had flattened (lost pitch) near their tips when compared to fresh blades.  But when I look at them again now, I no longer see any difference.  It's as though they've recovered their original shape.

Conclusions?

Very good test! Conclusion? I think it confirms what we already know, which is that there is a significant variance in prop performance. If we assume that "good" props are those with more pitch (i.e. lower RPMs), then this simple hover test (2 minutes, indoor, take off > hover > land) is a great way to measure prop performance. I think RPM averages are even better, as you get a number for each motor, but even visual analysis is useful. All this, without getting into the debate of why there is such variance in prop quality. I really look forward to test good aftermarket props, new out of the box, and see how they do.

Zbip57 Posted at 5-14 10:00
I went ahead and replaced the right-rear prop blades.  Here's the new graph.  Same deal, flying indoors, auto-takeoff, hover for two minutes, auto-land, never touched the control sticks.

Compare the right-rear motor speed (yellow) now with fresh prop blades installed versus the original blades (yellow) in the previous graph above.

I'm going to be trying this test soon, but looking at my old props that I "Destroyed" look to be back in their original shape. All be it the original shape wasn't like the pictures we've been seeing but it's visually back to normal. When I get some time I'll re-evaluate the things.

@120 did you find a quicker way of averaging out the rpm information?

ABeardedItalian Posted at 5-14 12:42
I'm going to be trying this test soon, but looking at my old props that I "Destroyed" look to be back in their original shape. All be it the original shape wasn't like the pictures we've been seeing but it's visually back to normal. When I get some time I'll re-evaluate the things.

@120 did you find a quicker way of averaging out the rpm information?

You can output a CSV file and open in Excel.  It's a lot of data cells but easy to perform average, min, max, etc calculations.

I could not find an obvious way to export data from CsvView, but I found that it creates a CSV automatically, in the C:\Users\your_username\.CsvViewDir\temp folder. Motor speeds are in columns DD to DG, and I recommend using the following formula to calculate averages (remove spaces around ":", had to add them or the forum will treat it as an emoticon):
=AVERAGEIF(DD : DD, ">8000")
This will eliminate values below 8k RPMs, due to motors being idle or spinning up/down. I actually did not do that yesterday, thinking that my flight was long enough (over 10 minutes) but I just did it and the numbers changed quite a bit:

Right Front Average: 9730 RPMs
Left Front Average: 9345 RPMs
Right Back Average: 9843 RPMs
Left Back Average: 10292 RPMs


[EDIT: the export function in CsvView is right there where it should be, under File > Export Csv, in the main window. Ooops.]

So far we have been focussing on prop shape. However we had reports of variation in motor speed with props that look fine, and also reports of props that look flattened but fly well and do not trigger any error. Therefore could it be some variation in plastic quality, such as flexibility? At 10k rpm that would certainly have some incidence on lift too, wouldn't it?

zeb_ Posted at 5-14 14:03
So far we have been focussing on prop shape. However we had reports of variation in motor speed with props that look fine, and also reports of props that look flattened but fly well and do not trigger any error. Therefore could it be some variation in plastic quality, such as flexibility? At 10k rpm that would certainly have some incidence on lift too, wouldn't it?

Absolutely, we can't really know what's going on when a thin piece of plastic rotates at 10k RPMs or more... and visual inspection might tell you something, but not the full story. Dings and scratches (up to a certain limit, of course) are not nearly as important as other factors that can influence the shape of the prop UNDER LOAD. That's what really matters, not how the prop looks when you have it in your hands. I'm not saying it's useless, some photos are quite telling and show clear flattening, but what really counts is what the prop does in flight.  
This is why I cannot stress enough how useful this test is, because RPMs do tell the full story. Even better would be a test under heavy load (like S-Mode full speed ahead), but it would be complicated to isolate the right samples, and you also have wind or turbulence getting into the equation. As far as I'm concerned, indoor hover is good enough, easy enough, and if we all do exactly the same thing, we can properly compare results.

120ccpm Posted at 5-14 13:51
I could not find an obvious way to export data from CsvView, but I found that it creates a CSV automatically, in the C:\Users\your_username\.CsvViewDir\temp folder. Motor speeds are in columns DD to DG, and I recommend using the following formula to calculate averages (remove spaces around ":", had to add them or the forum will treat it as an emoticon):
=AVERAGEIF(DD : DD, ">8000")
This will eliminate values below 8k RPMs, due to motors being idle or spinning up/down. I actually did not do that yesterday, thinking that my flight was long enough (over 10 minutes) but I just did it and the numbers changed quite a bit:

Interesting.  In my CSV file the Motor Speeds are found in columns FI-FJ-FK-FL.

Here are my averages using your formula >8000.

With fresh Left-Rear blades;
R.Front - 9595
L.Front - 10258
L.Back -  9850
R.Back - 10689

With fresh Left-Rear & Right-Rear blades;
R.Front - 9581
L.Front - 10338
L.Back -  9920
R.Back -  9661


Next step, I think I'll go ahead and replace the L.Front blades too.

Zbip57 Posted at 5-14 14:37
Interesting.  In my CSV file the Motor Speeds are found in columns FI-FJ-FK-FL.

Here are my averages using your formula >8000.

Averaged Motor Speeds, not counting those under 8000 while take-off & landing.


With fresh Left-Rear blades;
R.Front - 9595
L.Front - 10258
L.Back -  9850
R.Back - 10689

With fresh Left-Rear & Right-Rear blades;
R.Front - 9581
L.Front - 10338
L.Back -  9920
R.Back -  9661

With fresh L.Front, L.Rear, & R.Rear blades.  R.Front remains original;
R.Front - 9756
L.Front - 9603
L.Back -  9974
R.Back -  9564

It was the L.Rear that triggered the Motor Speed warning on an earlier flight, and it was replaced then.  It flew fine before and after, and probably would have continued to do so.  I've never experienced an Uncontrolled Descent.  If I stuck with the original blades, would they eventually have  degraded to the point of triggering a Motor Speed warning?  Who knows?

Even though I can see no visible difference when comparing the old vs. new prop blades, these Motor Speed plots clearly indicate that fresh prop blades do have a measureable effect on Motor Speed.

Zbip57 Posted at 5-14 14:37
Interesting.  In my CSV file the Motor Speeds are found in columns FI-FJ-FK-FL.

Here are my averages using your formula >8000.

No idea why you get motor speeds in different columns... maybe a different version of CsvView?

Did you replace also the Left Back prop, for your second flight? From the numbers, it doesn't seem you actually needed it... you were at 9850 RPMs, and went to 9920... very similar.

I think it's fair to say that the best props will hover at about 9600 RPMs, so we can use that number as a baseline. That doesn't mean we should all jump to replace props if we see higher numbers, I think we still to understand how much of a higher number we can tolerate, before we start to get the occasional Motor Speed Error warning.

Anyways, good analysis, glad we looked into it, definitely learned something new about the MM, thanks to the OP for starting this thread, and to all who contributed to it.

One last test, I did replace my Rear Left props because the motor was averaging over 10k RPMs in a hover, and wanted to see with my own eyes the effects of a new prop. Here are the old & new numbers:

# Original Props
Right Front: 9730 RPMs
Left Front: 9345 RPMs
Right Back: 9843 RPMs
Left Back: 10292 RPMs

# New Left Back props:
Right Front: 9582 RPMs
Left Front: 9386 RPMs
Right Back: 9960 RPMs
Left Back: 9431 RPMs

Pretty significant drop, going from 10292 to 9431! Looks like the Right Back would be the next one up for replacement, but I'll wait. Also, I said it before but it's worth repeating: my MM never threw any Motor Speed warning, nor experienced any Uncommanded Descents, it's been absolutely flawless. I replaced the prop purely for the love of science. Note also that I store it in a $10 eBay case, which I modified from the very first day so that the rear props would not be under any pressure.

120ccpm Posted at 5-14 16:31
No idea why you get motor speeds in different columns... maybe a different version of CsvView?

Did you replace also the Left Back prop, for your second flight? From the numbers, it doesn't seem you actually needed it... you were at 9850 RPMs, and went to 9920... very similar.

I might have selected different option settings in CsvView. (See image)

I only replaced the props once.  Starting left-rear, then adding right-rear, then left-front.  The right-front are the only remaining original blades.

I did not put any of the old props back on.  I figured that, with the loctite on the screws, they're meant to go on just once and stay on.  I didn't want to risk taking any of the screws out again.

Agreed.  The props likely do not need to be replaced unless an actual Motor Speed warning is triggered.  Even then, if the props are only marginally bad they may continue to fly well and never trigger the same warning again.  But if you get repeated warnings, it's certainly worth paying attention to.  These graphs do illustrate that there is a measurable difference in the motor speed output when using fresh props compared to old props.


We still don't know what threshold actually triggers the Motor Speed warning.  Is it just when the motor exceeds some finite number like 1600 rpm?  Or is it when some threshold difference is exceeded between the actual response of the motor versus what thrust the controller is commanding and expecting of the motor?