New build. I can say there was not a battery failure, no voltage drop, no current drop, no vibration errors, no EKF errors, no GPS errors, rangefinder looks good and there was no change in flight mode as well.
What happened: I armed via the zero throttle/full right rudder after having armed via the kill switch which is mounted to the drone. There were no errors on the HUD. Battery was in mid range of capacity. Arming spun the motor at about 18% throttle, rotation of props was correct and thoroughly tested beforehand. I raised the throttle to about 40% and got no change which is what I expected as I had not hit the “dead band” where the autopilot takes control. This was a test.
I then raised it to about 50% and the drone pops up to an altitude of about 2 feet via a throttle burst of about 24% that settled down to 15%. I let it sit for a second or two to see if it destabilized (new build). It held nicely so I raised the altitude to about 5-6 feet and the autopilot held it in place very nicely. I gave a forward command via the transmitter and it moved about 3-4 feet forward as expected. It was VERY stable. I walked a few feet around it to get a different viewpoint of the drone while it was stable (I was looking for potential mechanical failures). After about 5 seconds of very good stable flight at an altitude of 5 feet (maybe 6 feet) it began an un-commanded drop in altitude, un-commanded by me that is. It dropped fast enough that it made a slight bounce when it hit the gravel road and then bounced again and finally it tipped over.
I recall giving it a “up” command briefly but it was not adequate to save the bird. When I realized it was going in for a hard landing I immediately went to throttle zero but it was too late. This all happened in a matter of seconds.
If you watch the tlog carefully you will see a steady 15-16% throttle and then a dip to 12 then 9% throttle then it goes back up to 15%. I can’t explain this and it was what caused the drop in altitude. Maybe if I had it higher I could of saved it but I don’t know.
Could the hover throttle be too low? 30% has always been a magic number but never one with an explaination.
It looks like your throttle input commanded the decent. Was there any chance you have an RC failsafe set that would go to minimum throttle or something like that?
If I am reading the graph right, the red line is MY throttle input. The sharp decline is me pushing the throttle to zero (missed complete zero but got it a split second later) but only AFTER the drop in altitude. To see this run the .tlog in the .25 speed. The area of the red line that is flat is when I started to walk around to visually inspect the flight but if you divide that flat red part into say, 3 pieces the dip in throttle takes place in that last third. My hand was off of the throttle.
No failsafe is on that I am aware of, I’ll take another look when I get back to the bird.
I’m willing to say that since I had sweatpants on (cold outside) it is very possible that I accidently snagged the throttle on them and NEVER felt it! And me thinking I went to throttle zero could have been that last little distance seen in red that occurs after the big drop in throttle command. Yeah, that’s entirely possible!
Especially if you add a little brain fog in the mix, like I said, this all happened in seconds.
I tried to do the “en mass” ESC calibration several times via the Motor Calibration option under setup but I never did get it to work. I usually, and did in this case, calibrate them one by one via the original method.
I am unaware of how to confirm an ESC calibration except to just turn them on without them having the autopilot in the loop, which I did. Since the prop spun up at just 3% throttle (or thereabouts) I considered the calibration successful (plus I got the beeps during the process) the ESC’s are default firmware with no modifications and are the Hobbywing XRotor Pro 50A with the DEO option OFF. Here is a link to them. If I am wrong about these being the ones installed I’ll correct this post later on.
I’ve been able to repeat this four more times. Each time it presents as a dip in throttle of about 6% followed by an immediate loss in altitude and a restoration of hover throttle.
The only thing I can say is that I am below the 30% rule for hover BUT I have a second drone that hovers just fine at 20% so I can’t figure this one out.
There is some other funky stuff going on like GPS data not writing, a crazy barometer graph the compass being “unchecked” and the orientation box for it not working yet a manual input via the Full Parameter Tree worked to restore it as well as a check against a known good drone’s parameter file shows no unexpected differences. NONE.
I’m going to factory reset this drone and bring it back up the old fashioned way (one thing at a time and with a BIG HAMMER).
The tlogs did not do so well due to poor antenna management on the GCS side but I might have some to show if you need them, TYVM!!
OH, if it helps: All up weight is 6.1kg, I’ve done this while up too high for ground effect and very low that probably did have some ground effect. Not sure what else to note just yet.
EDIT: I did think of something. This bird has a gear box to raise and lower the landing gear that has always been present and active. I can get them to go up and down by flicking a switch on the Tx but I’ve never seen them work while in auto mode, perhaps because I’ve never gone high enough, or perhaps there is a fault in the manufacturer supplied gear or control box for it??
Ok, lets have a look. As you can see below every flight the throttle commands the decent. The interesting thing is the throttle does not go to minimum throttle. Each time it goes specifically to 1155.
I still think your radio is doing some sort of fail-safe where and the throttle is set to 1155. The problem is the other channels look like they are still working. I only see this specific throttle output when you are crashing. Is there some sort of programmed feature on your radio that might be doing this?
I scanned through all options and found no mixing activated, I DID FIND one throttle Fail Safe that was active and should not have been! It was made to set throttle at 20% position when B-F/S (battery failsafe?) went ACT (Active?). It is now off, I’ll spool her up here shortly.
There is also an ATL that is ON under the Parameter page 1 section, I’ll turn that booger off too.
Present in the T10J but not present in my T12K is a section called PARAMETERS. Page 1 at the bottom has something called ATL which was ON, I turned it OFF. (see below for definition of ATL)
***Next is there is a FAIL SAFE section. Under page 1, number 3: THR the F/S was on and set to 20% POSI(tion) and B-F/S (Battery Fail Safe) as ACT. I turned that to NOR(mal).
Between those two adjustments the issue of throttle dipping is gone.
TYVM!!
FYI:
Adjustable travel limit (ATL): makes the channel 3 TRIM LEVER (THROTTLE TRIM) effective only at low throttle,disabling the trim at high throttle. This prevents pushrod jamming due to idling trim changes. This function defaults to ON.If you are not using channel 3 for throttle, you may want trim operation the same as on all other channels. To do so, set ATL to OFF. If you need the ATL to be effective at the top of the stick instead of the bottom, reverse the THR-REV setting. Note that this affects all models in the radio, not just the model you are currently editing. THR-REV, see p. 92.
FailSafe (loss of clean signal and low receiver battery) submenu (F/S): sets responses in case of loss of signal or low Rx battery. FailSafe (F/S): instructs a 2.4G receiver what to do in the event radio interference is received.
The F/S ) function moves each servo to a predetermined position.
SUMMARY:
The location I was doing my initial testing in was between steel/aluminum buildings with about a 30 foot wide gap between them. All places there have WiFi going 100% of the time. I’m willing to bet that there is a LOT of interference going on and that is was kicked the radio to put the throttle position to the 20% mark as it certainly was NOT a low battery. There are no low batteries anywhere in the system and they are monitored as well. Since I can’t measure interference I can’t say if it was the Fail Safe setting OR the Adjustable Travel Limit (ATL) setting that was messing with everything but I turned those off and so far the problem is gone.
