After upgrade Mavic Mini firmware to 1.00.05, a new "function" of detecting motor speed error is introduced.
Like many other, my MM also have the error on test flight after FW upgrade.
I had replaced the rear right propeller and have another test flight. No error occurred since then.

Here is the motors speed from DAT file before and after replacing the rear right prop:



Both tests are made in a nearly no wind conditions.
The graph on the left is the data taken before replacing prop. The yellow line represent the rear right motor speed and the blue line represent the rear left motor speed.
As you can seen, the "before" rear right motor speed is higher than the front props as well as the rear left motor speed. And after the rear right prop is replaced, the motor speed is lowered and even lower than the rear left.
The motor speed will be higher when the related prop is not providing enough power. The data I got show that the new firmware is trying to check the props with problem. And the replacement can solve the problem at least in my case.

===============================================================================================
Edited on 1 May 2020
Have changed the rear left prop and made another test flight. The test is under a little wind condition. Now the blue line (rear left motor speed ) is also lowered:


I think it is a good idea to keep on tracking the motors speed regularly. To make sure the propellers condition is good for flight.

Are you able to replace the back left and repeat the test?

Some more experienced people have conducted some more detailed experiments on this: https://mavicpilots.com/threads/ ... escent-tests.86237/

The issue seems to be indeed low performing props (whether that is from incorrect storage in the box or smth else is not yet clear) that eventually get "flattened" mid-flight because of the motor's very high speed. When that happens, the rear motor maxes out. To compensate for this, the FC prioritizes maintaining the drone's pitch over keeping altitude (that's a weird one) and drops the rpm on the *front* engines, causing the Mini to lose altitude. The rear props need low RPM to "unflatten" but the FC doesn't allow that and keeps the rear motor maxed out. This perfectly explains why all Minis that hit the ground without crashing immediately bounce back and hover normally.

An indicator of this issue, before the uncommanded descent begins, is low forward speed (but normal speed if flying backwards).

I think the FC simply lands the AC after too many errors. In the past descents there wasn't any real pattern to the current flyability of the drone. And the props weren't maxed out during the descent.

Ice_2k Posted at 4-30 07:20
Some more experienced people have conducted some more detailed experiments on this: https://mavicpilots.com/threads/ ... escent-tests.86237/

The issue seems to be indeed low performing props (whether that is from incorrect storage in the box or smth else is not yet clear) that eventually get "flattened" mid-flight because of the motor's very high speed. When that happens, the rear motor maxes out. To compensate for this, the FC prioritizes maintaining the drone's pitch over keeping altitude (that's a weird one) and drops the rpm on the *front* engines, causing the Mini to lose altitude. The rear props need low RPM to "unflatten" but the FC doesn't allow that and keeps the rear motor maxed out. This perfectly explains why all Minis that hit the ground without crashing immediately bounce back and hover normally.

That was worth reading, thanks

Yes, thanks for the data.

Did you change both rear propellers ?? Did they have any deformations ???

Ice_2k Posted at 4-30 07:20
Some more experienced people have conducted some more detailed experiments on this: https://mavicpilots.com/threads/ ... escent-tests.86237/

The issue seems to be indeed low performing props (whether that is from incorrect storage in the box or smth else is not yet clear) that eventually get "flattened" mid-flight because of the motor's very high speed. When that happens, the rear motor maxes out. To compensate for this, the FC prioritizes maintaining the drone's pitch over keeping altitude (that's a weird one) and drops the rpm on the *front* engines, causing the Mini to lose altitude. The rear props need low RPM to "unflatten" but the FC doesn't allow that and keeps the rear motor maxed out. This perfectly explains why all Minis that hit the ground without crashing immediately bounce back and hover normally.

Awesome, great to see more people trying to debunk this.

Ice_2k Posted at 4-30 07:20
Some more experienced people have conducted some more detailed experiments on this: https://mavicpilots.com/threads/ ... escent-tests.86237/

The issue seems to be indeed low performing props (whether that is from incorrect storage in the box or smth else is not yet clear) that eventually get "flattened" mid-flight because of the motor's very high speed. When that happens, the rear motor maxes out. To compensate for this, the FC prioritizes maintaining the drone's pitch over keeping altitude (that's a weird one) and drops the rpm on the *front* engines, causing the Mini to lose altitude. The rear props need low RPM to "unflatten" but the FC doesn't allow that and keeps the rear motor maxed out. This perfectly explains why all Minis that hit the ground without crashing immediately bounce back and hover normally.

"To compensate for this, the FC prioritizes maintaining the drone's pitch over keeping altitude (that's a weird one) and drops the rpm on the *front* engines, causing the Mini to lose altitude."

To me it seems that maintaining pitch is more important than keeping altitude. What would happen if other way round? Drone turning over and crashing?

Eka Posted at 4-30 10:45
"To compensate for this, the FC prioritizes maintaining the drone's pitch over keeping altitude (that's a weird one) and drops the rpm on the *front* engines, causing the Mini to lose altitude."

To me it seems that maintaining pitch is more important than keeping altitude. What would happen if other way round? Drone turning over and crashing?

I mean the pitch required for keeping to move forward. It could just stop and stay level, it seems that would be better than falling down.

Ice_2k Posted at 4-30 10:49
I mean the pitch required for keeping to move forward. It could just stop and stay level, it seems that would be better than falling down.

Ah, yes. I was thinking about hovering only.

Ice_2k Posted at 4-30 10:49
I mean the pitch required for keeping to move forward. It could just stop and stay level, it seems that would be better than falling down.

They don't fall down. They descend. People keep erroneously thinking that these things are falling from the sky due to lack of lift with the rear prop maxed out. There have been videos of many of these descents all the way back to when this began. Basically when the drone was released. And these are controlled descents, but controlled by the FC and there is nothing you can do with the RC to stop it. I don't call them landings because they descend all the way to the ground (or water) without a pause right before landing, like what happens when you manually land.

Hi thanks for your testing,

how did you read the DAT file ? im not able to upload it anywhere or convert it..

greets

Deucalion Posted at 4-30 10:53
They don't fall down. They descend. People keep erroneously thinking that these things are falling from the sky due to lack of lift with the rear prop maxed out. There have been videos of many of these descents all the way back to when this began. Basically when the drone was released. And these are controlled descents, but controlled by the FC and there is nothing you can do with the RC to stop it. I don't call them landings because they descend all the way to the ground (or water) without a pause right before landing, like what happens when you manually land.

Perhaps he shouldn't of used "falling down" in that post, but his previous one quoting mavic pilots is spot on.

The FC prioritizes maintaining the drone's pitch over keeping altitude (that's a weird one) and drops the rpm on the *front* engines, causing the Mini to lose altitude.

This is what we see on video, it's not "falling per say" but it's losing altitude at a sustained rate.

ABeardedItalian Posted at 4-30 11:09
Perhaps he shouldn't of used "falling down" in that post, but his previous one quoting mavic pilots is spot on.

The FC prioritizes maintaining the drone's pitch over keeping altitude (that's a weird one) and drops the rpm on the *front* engines, causing the Mini to lose altitude.

None of the props are at max rpm, or even close to it, during the descent. Is that what you mean?

Deucalion Posted at 4-30 11:34
None of the props are at max rpm, or even close to it, during the descent. Is that what you mean?

Maybe I am wrong, but I get the impression that people think the drone is doing all it can do, but because of the bad prop, it can't generate enough lift on that corner, and it descends due to lack of lift. The instances don't look like this and if that were true, how did the prop heal itself, because they fly again. It seems that the FC goes into a different state and descends the AC to the ground. It is almost like walking on a sprained foot wrong and going to the ground in a controlled way to minimize the weight on that foot. Maybe limps to the ground is a better way to describe it.

Deucalion Posted at 4-30 10:53
They don't fall down. They descend. People keep erroneously thinking that these things are falling from the sky due to lack of lift with the rear prop maxed out. There have been videos of many of these descents all the way back to when this began. Basically when the drone was released. And these are controlled descents, but controlled by the FC and there is nothing you can do with the RC to stop it. I don't call them landings because they descend all the way to the ground (or water) without a pause right before landing, like what happens when you manually land.

Yes, sorry, I didn’t mean literally falling like a brick

Deucalion Posted at 4-30 11:50
Maybe I am wrong, but I get the impression that people think the drone is doing all it can do, but because of the bad prop, it can't generate enough lift on that corner, and it descends due to lack of lift. The instances don't look like this and if that were true, how did the prop heal itself, because they fly again. It seems that the FC goes into a different state and descends the AC to the ground. It is almost like walking on a sprained foot wrong and going to the ground in a controlled way to minimize the weight on that foot. Maybe limps to the ground is a better way to describe it.

Did you read the mavicpilots thread and the attached logs? It answers all of that. The prop gets flattened because of its previous issue plus the very high rpm. It is kept flattened by that top rpm. Once the drone hits the ground and bounces back, the motor slows down allowing the prop to “de-flatten”

Ice_2k Posted at 4-30 12:15
Did you read the mavicpilots thread and the attached logs? It answers all of that. The prop gets flattened because of its previous issue plus the very high rpm. It is kept flattened by that top rpm. Once the drone hits the ground and bounces back, the motor slows down allowing the prop to “de-flatten”

Regarding the thread you refer to, we know the case has nothing to do with it, because we all use the cases, and this would be happening at least to most of us.

He said he deliberately flattened them, and we know that doesn't apply to at least most of the uncommanded descents, because most of the users were adamant that they didn't do such a thing.

So, I guess it comes down to, do the marginal props "flatten" out in flight due to the RPMs and become ineffective till the craft lands and the motors stop? Plausible I guess, but I need to review the past incidents if that was what happened. It does make sense but I do not recall props spinning at max RPMs in those incidents, but I will verify. If true, then I guess the issue with the bad props is that they are too thin/flexible.

The thing is, if this requires pre-flattening the props as he did, then how does that help us? I don't think the users to whom the descents were happening to were pre-flattening their props. I would feel better if this just required marginally bad props.

Is there a solution to the problem?

Wait until somebody like that old **rt called hallmark007 kicks in and tell everybody it wasn't the RR prop defective, it was all the other 3 pairs running too quickly.

At the end of this still I expect the same old questions again and again AND AGAIN as if we didn't exist, the problem was never investigated, verified, reported and solved several times. -Oh look! The Fly app tells me my right rear propellers aren't up to the task and suggests that I replace them! Must be defective, let me downgrade to version 0.0.1 alpha.-

I am absolutely creepily worried that people like this are flying our skies...

Drone668 Posted at 4-30 13:26
Is there a solution to the problem?

More rigid props I think.  It would be worth a try

m80116 Posted at 4-30 15:13
Wait until somebody like that old **rt called hallmark007 kicks in and tell everybody it wasn't the RR prop defective, it was all the other 3 pairs running too quickly.

At the end of this still I expect the same old questions again and again AND AGAIN as if we didn't exist, the problem was never investigated, verified, reported and solved several times. -Oh look! The Fly app tells me my right rear propellers aren't up to the task and suggests that I replace them! Must be defective, let me downgrade to version 0.0.1 alpha.-

I just had an idea about '' motor error ''. Just install a small parachute on top and when the famous message appears, you trigger the parachute.
It will reduce the load on the rear motor.

JGFly Posted at 4-30 15:35
More rigid props I think.  It would be worth a try

Thank you.

m80116 Posted at 4-30 15:13
Wait until somebody like that old **rt called hallmark007 kicks in and tell everybody it wasn't the RR prop defective, it was all the other 3 pairs running too quickly.

At the end of this still I expect the same old questions again and again AND AGAIN as if we didn't exist, the problem was never investigated, verified, reported and solved several times. -Oh look! The Fly app tells me my right rear propellers aren't up to the task and suggests that I replace them! Must be defective, let me downgrade to version 0.0.1 alpha.-

what are you trying to say?

Since everyone else is doing it, I had to do it as well. I tested my original props, which have never given me a speed error, and there was some variation in RPMs. I swapped them all out with a new set, and 3 of the motors were closer, but now the right front was high. So I replaced it with another new pair, and now they are closer together. What I am seeing is that there is quite a bit of margin of error for the props, which explains why uncammanded descents were so rare. Even my original props with the largest variation never came close to maxing out in RPMs. And that was with me flying it straight up at max speed sport mode in wind 400 feet. While I did swap out the originals for this test, had it not been for that, I wouldn't have ever swapped them out unless I got a warning from the FW, which is probably the best advice. Even the new props can have variation, but the drone has a pretty good margin.

Here are the originals being pushed in wind...



Then I took them inside for a hover without wind...



Then I replaced all of them with new...



And finally, because the front right was still high, I replaced it with new...

Mini User Posted at 4-30 11:06
Hi thanks for your testing,

how did you read the DAT file ? im not able to upload it anywhere or convert it..

You can use CsvView to view / convert the DAT file :
https://datfile.net/CsvView/downloads.html

"Like many other, my MM also have the error on test flight after FW upgrade."

What was the exact error message? Most others are getting "left rear" message. But you replaced the right rear. Was your error message about the right rear?

I also just tested my backup mini with the banged up props, indoors hovering only (no wind or maneuvering). It shows just about a range of 1000 rpm from the lowest speed to the highest. Should I assume that is acceptable? I have seen no motor error messages so far after the FW update.

The actual trigger is whether a motor goes above a certain RPM. My drone didn't trigger anything (and has never) and it was at 15,500 at its very max, full throttle up sport mode, and that was the fastest prop. The other props were well below that. Anyways, I think 16,500 rpm is max. I saw that somewhere.

Anyways, yes, 1000 difference should be fine. My example above had a 3000 differential, and that still was fine.

Deucalion Posted at 4-30 12:38
Regarding the thread you refer to, we know the case has nothing to do with it, because we all use the cases, and this would be happening at least to most of us.

He said he deliberately flattened them, and we know that doesn't apply to at least most of the uncommanded descents, because most of the users were adamant that they didn't do such a thing.

With all due respect, you're not paying attention This only happens if you store the drone *incorrectly* in the case, not just if you have the case. Meaning if the rear props are pointing towards the rear of the drone and getting caught between the drone's belly and the case. And this is exactly how @sar104 on mavicpilots flattened them, by deliberately storing them like this. If you store them normally (the two props pointing towards the other's motor, there is no problem with the case).

GaryDoug Posted at 4-30 18:56
"Like many other, my MM also have the error on test flight after FW upgrade."

What was the exact error message? Most others are getting "left rear" message. But you replaced the right rear. Was your error message about the right rear?

Yes my motor speed error is right rear. the right rear motor beep after landing

Drone668 Posted at 4-30 13:26
Is there a solution to the problem?

Hi @Drone668, hope your'e fine !

I answered you '' More rigid props I think. It would be worth a try ''

But,  Don't do that. You could lose the warranty. As already explained by some other members, always choose the original parts.

Cheers !

JGFly Posted at 5-1 03:47
Hi @Drone668, hope your'e fine !

I answered you '' More rigid props I think. It would be worth a try ''

Ah ok thanks!

Ice_2k Posted at 5-1 00:05
With all due respect, you're not paying attention  This only happens if you store the drone *incorrectly* in the case, not just if you have the case. Meaning if the rear props are pointing towards the rear of the drone and getting caught between the drone's belly and the case. And this is exactly how @sar104 on mavicpilots flattened them, by deliberately storing them like this. If you store them normally (the two props pointing towards the other's motor, there is no problem with the case).

Yes, I agree that you can store the drone in the case stupidly and mess up the props. I have mentioned that in the past. But that is not the cause of the bad props we have seen. The users were not storing the drone stupidly in their cases. They said so. That is all I am saying. I am still on the fence on whether a symptom of a bad prop (as in the bad props we have seen) is this centrifugal flattening. I need to do more research.

Any new cases of "unwanted descents" since new firmware?

Nope.

fans1cafe718 Posted at 5-1 05:59
Any new cases of "unwanted descents" since new firmware?

Nope.

Did you perform tests after the new firmware?

Drone668 Posted at 5-1 06:04
Did you perform tests after the new firmware?

I actually agree with fan. I know it hasn't been long enough yet, but I believe this FW will eliminate most of the uncommanded descents, assuming people replace their props if they get the error.

That is good to hear.

Drone668 Posted at 5-1 06:04
Did you perform tests after the new firmware?

...never had this prob...but, it's been a concern from reading other's unfortunate dismays...

Deucalion Posted at 5-1 05:46
Yes, I agree that you can store the drone in the case stupidly and mess up the props. I have mentioned that in the past. But that is not the cause of the bad props we have seen. The users were not storing the drone stupidly in their cases. They said so. That is all I am saying. I am still on the fence on whether a symptom of a bad prop (as in the bad props we have seen) is this centrifugal flattening. I need to do more research.

I always stored my props as they should be in the flymore case, and then received the motor speed error on the left rear motor.
I changed the props and flew it again with no errors.
At the end of the day I decided to buy a case where the props are stored inline with the drone, still have to hold the drone upright to fit the props as they should be.

https://www.ebay.co.uk/itm/Porta ... 2057872.m2749.l2649