Lost my Mini 4 Pro - log shows I was pulling climb but drone kept descending

Pavlito · 2026-01-10T04:59:02-0500

Lost my Mini 4 Pro in water yesterday. Did a full log analysis and found some things I don't understand.

I was flying low over water, around 3-4 meters. Video feed cut out. I pulled full climb immediately. Drone kept going down and hit the water about 1 second later.

What the logs show:

Throttle was maxed out (1684 = full climb) for 6 seconds before impact
Drone was pitched forward around -30 degrees
Speed was about 12 m/s
Battery 39%, signal 100%
P-GPS mode the whole time
No errors or warnings in the log
VPS was showing 12m then 2.8m (clearly not working over water)
Video died first, radio/telemetry kept working until impact

BossBob · 2026-01-10T06:54:48-0500

Hopefully @Meta4 will be along soon. He’s an expert on interpreting crash logs.

Meta4 · 2026-01-10T07:23:42-0500

Pavlito said:
Lost my Mini 4 Pro in water yesterday. Did a full log analysis and found some things I don't understand.

Strangest flight data I've ever seen - nothing makes sense.

I've had a look at your flight data and can't find anywhere that the throttle input has any relation to drone height.
Not just in the end of the flight, but right from startup.

Also .. The pitch, roll and yaw data doesn't show any crash impact.

There is something wrong with the flight data.
It's not possible to tell anything about your flight incident from it.

Kilrah · 2026-01-10T07:54:33-0500

Pavlito said:
Video died first, radio/telemetry kept working until impact

Looks like at the end while the RC was receiving telemetry from the aircraft the aircraft wasn't receiving the RC's commands, so anything past 920.3 seconds the aircraft would have been doing on its own without being able to receive your commands.

You took off with poor GPS reception and without a home point set, it was set 29.9 seconds into the filght. That in itself should normally not be a problem, but it seems it has been.

You were in a 60m height limit zone and once the home point was registered the GPS altitude was around 230ft, meaning about 70m. It then took that as reference and tried to fight any throttle up command as that was above the allowed limit even if you were actually way lower above ground.

Does seem to indicate a potential firmware bug to me, but it's by definition difficult to determine what value should be trusted.
Likely would have had no issue had you not taken off without a home point set as you should always ensure.

Send that to DJI anyway.

Also you set your RTH height to 100m, that might cause issues when you're in a 60m max height area but IMO that would be secondary.

Pavlito said:
No errors or warnings in the log

Plenty of them

FLDave · 2026-01-10T07:57:22-0500

First question: What did you visually see as the action of the drone during the flight, that is visually without the aid of your controller screen? That should be a big help.

Meta4 · 2026-01-10T08:55:09-0500

Kilrah said:
You were in a 60m height limit zone and once the home point was registered the GPS altitude was around 230ft, meaning about 70m. It then took that as reference and tried to fight any throttle up command as that was above the allowed limit even if you were actually way lower above ground.

Your drone sets its zero height when the drone powers up, not when GPS is acquired.
Height completely unrelated to GPS and homepoints are two-dimensional lat/long data with no height.

Kilrah said:
Likely would have had no issue had you not taken off without a home point set as you should always ensure.

Sorry but this isn't true.
The only issue that launching before acquiring GPS data causes is that the recorded home point is not the launch point.

Kilrah said:
Plenty of them

None of which were related to the loss of the drone.

TXaquaholics · 2026-01-10T09:12:26-0500

What were the weather conditions? Any chance icing was involved?

Kilrah · 2026-01-10T10:06:16-0500

Meta4 said:
I've had a look at your flight data and can't find anywhere that the throttle input has any relation to drone height.
Not just in the end of the flight, but right from startup.

Meta4 said:
Your drone sets its zero height when the drone powers up, not when GPS is acquired.
Height completely unrelated to GPS and homepoints are two-dimensional lat/long data with no height.

It should be, however as you can see at the start it climbs normally, it's only once home point got set that altitude did a jump and it started refusing to obey the throttle commands, when it did climb a bit it then went down again on its own, and that happens to be always trying to go back down to around 230ft altitude.

Dogbiscuit · 2026-01-10T11:25:30-0500

Meta4 said:
Strangest flight data I've ever seen - nothing makes sense.

I've had a look at your flight data and can't find anywhere that the throttle input has any relation to drone height.
Not just in the end of the flight, but right from startup.

Also .. The pitch, roll and yaw data doesn't show any crash impact.

There is something wrong with the flight data.
It's not possible to tell anything about your flight incident from it.

The RC was in Mode 3 and AC was responding to the commanded inputs. The AC was descending with throttle full down at end of log.

MS Coast · 2026-01-10T11:48:55-0500

The OP states that the incident took place yesterday, January 9, 2026. But the log file shows December 18, 2025 as the date of the flight. ??

slup · 2026-01-10T12:56:21-0500

Dogbiscuit said:
The RC was in Mode 3 and AC was responding to the commanded inputs. The AC was descending with throttle full down at end of log.

That's totally correct...

The throttle stick command correlates to heading speed seen in the chart below...

... and the elevator stick command correlates to height... & there it's easy to see that the OP constantly applied descent commands from approx. 911sec until the drone drowned (value 1024 is neutral stick, values lower than that means decent).

So this is a user error, nothing else.

Dale D · 2026-01-10T13:09:01-0500

Sorry about your loss. I am selling my Mini 4 Pro (bought a mini 5 Pro), but not sure about shipping to Serbia. Look for it in the classifieds if interested..

DISC · 2026-01-10T13:15:34-0500

slup said:
So this is a user error, nothing else.

So a quick summary for the noobs like me means mode 3 has caused the OP to send the wrong commands to the drone?
Isn’t this part of most pilots check when first launched - about 6’ up? Or is it capable of switching modes during flight??

slup · 2026-01-10T13:54:58-0500

DISC said:
...mode 3 has caused the OP to send the wrong commands to the drone?
...is it capable of switching modes during flight??

Maybe the pilot wasn't used to mode 3... & fell back to a mode 2 thinking there close to the water with no height to give him head room to catch himself. The drone might have been piloted by a friend that's usually using mode 2?

No, the stick modes doesn't switch by itself midflight...

WanderLost · 2026-01-10T17:17:35-0500

That is some impressive sleuthing. It never would have occurred to me that the issue was exacerbated by a switch to Mode 3 and the concomitant inability to control the drone in the intended fashion.