Li'l Bertie
lvl.2
Flight distance : 1480344 ft
United States
Offline
|
TL;DR: Powering the RC, in flight, via the microUSB may lead to problems.
(If that's where you stop reading, a reply will show it; there are many details, please curb your enthusiasm :-)
Lucky Me Lucky me, I'm about to leave for a visit to the AZ/UT region. I'll have six Mavic Air batteries; three are spanking new. But I'm afraid the RC power will poop out and put a stop to the fun at some point while I still have batteries ready to fly.
The microUSB port on the side of my RC is unused, because I connect my iPad Mini to the RC using the USB-A port between the grips (big shout-out to MavMount!). I believe that's common. :: The plot thickens :: I had heard that in this situation, you can plug an external power source into the microUSB port during flight; this should theoretically bypass the internal battery and power the RC. With a beefy external battery, I could fly all day without a care about the RC power!
So I gave it a try; simultaneously, I wanted to flight-test the three new MA batteries before hitting the road. And while the RC itself had a full charge, this setup actually seemed to work! The charge-indicator LEDs on the RC ran the charging sequence (1,2,3,4,repeat), but otherwise everything seemed normal. I kept the bird low, just buzzing around the yard and across my suburban street, maybe 120' (40m) distant.
But after a minute or two, the DJI GO 4 screen went dark except for a terrifying notice: Aircraft disconnected
!! YIKES !!
I do not like this. PANIC! It was only across the street, and I quickly realized that despite the black screen, the RC still worked! ("How can this be? What does 'disconnected' even mean, if the bird is. Not. Actually. Disconnected?!?") Flying back was easy-peasy, but I was queasy :-/ indeed. Without telemetry, flying blind is not likely to end well if the bird is essentially invisible, even if it's within LOS. Had this happened out over some arroyo or above a distant butte, I'd have had to rely 100% on RTH (stay on-topic, folks). My faith in failsafe systems is more fervent regarding the 'fail' than the 'safe'. [As it happens, the bird was actually under a tree; RTH would have made it—briefly—a wood chopper. I'm glad I took control before that kicked in!]
The Mavic Air occasionally has video lag, but if you trust the bird and pay attention, it's bearable. And I've had Strong Aircraft Interference warnings ("Go home now, Gomer!"). But a disconnect—the total loss of video and telemetry—is quite frightening, at least to a novice like myself. Even if it isn't really disconnected. (WTF?)
The battery in the M.A. wasn't exhausted, but I switched it out anyway (with the second of the new MA batteries). With the RC still plugged into that gassed-up 20,000maH power-pack, I reconnected and went for another flight. Again, I stayed low and just cruised around until the battery was 30%. There were no issues, so I began a shakedown cruise with the final new battery.
And it happened again: Aircraft disconnected Yikes again!! This time the bird was maybe 50' (16m) away, maybe 70° of elevation. I was done for the day & landed. Dalwhinney time.
After the Adrenaline
Today, I flew three similar flights with exactly the same setup, except no external power whatsoever. In particular, I flew two of the three new batteries, and one that has a few furlongs on it. Everything went swimmingly; no disconnects. YAY! [Except this nagging question: was the issue really external power?]
For those who will ask: all the firmware and software (bird, controller, DJI GO4, batteries, you name it) are current; in fact these were the first flights of any kind after the March 21 firmware update [Nag], and I'd applied it to all six batteries in one sitting. I had about 17 GPS satellites, and no compass warnings.
For the curious, this point-of-view in Google Maps shows the takeoff point (the driveway partially obscured by 2 green trash bins and a pickup) and the first point of disconnect across the street (±30" {0.6m} altitude, 5 feet from the open fence section where an American flag is visible, between the tree trunks there). There are no overhead powerlines, no obstructions, no utility boxes, a short distance, a neighborhood I've flown a lot with never an issue, and BTW greenery isn't greenery in March. When I replay the flight in DJI GO 4, it simply terminates at the point of disconnect (duh).
Baseless Theories and Idle Speculation
Maybe some confusion between the power-pack and the RC caused the disconnects. The power-pack is itself a smart device; maybe it's chattering data into that port and not just power, though I'd say doing so would be bad design. Cross-talk on that port would probably be 100% protocol errors, so there shouldn't be an issue, but shouldn't isn't couldn't.
Or perhaps there was a momentary disconnect-connect on that microUSB plug, sending the system down the rabbit hole. I was using a cable that I use all the time for my phone, but I noticed later that its microUSB plug doesn't softly click into place on the RC like it should (and like other cables do). But neither does it just fall out; once in, it tends to stay put. I didn't think to snug it in at any point, but I am a tad dubious that it was flickering power during these flights. I had the rig on a lanyard around my neck, and during this time I was bending and squatting and moving around with the power pack in my hip pocket; a plug so loose that it would flicker the connection should have just fallen out at some point. It didn't. Nor do the disconnects correlate with some significant movement on my part.
Or maybe it's just a bad idea, you know? Applying power to the RC whilst in flight might be stupid. As handy as it might be, it might also be just stupid. If doing this doesn't cut the battery out of the circuit entirely, then it's definitely stupid. I've been stupid before. But at this point I'm not going to risk disconnects! Nope. Whatever happened, it happened only on those three (mostly short duration) flights with external power to the RC; never before or (haha, three relatively short flights) since.
I can live within the power allotted to me by the RC; if the bird is grounded because the RC can't fly her, she's grounded until it can. So be it.
--Li'l Bertie
PS: Or maybe it was something else entirely. [Nag]
A whole separate thread, probably. Let's not hijack this topic; I'll start another if someone has insights I should know. Or PM me, if that's possible here.
For those who will suggest log files: I thought the flight records might hold a clue, so I used DJI Assistant 2 to pull them from the bird. On discovering they're binaries, I acquired DatCon 3.2.0 to extract the data within. While ExtractDJI appears to function, DatCon rejects its output files. OK, so I used DJI Assistant 2 to pull just individual flights, bypassing ExtractDJI. DatCon will accept these, and it spins like a ballerina while creating the CSV, TXT, KML, everything you could hope for. But when it's done, they're all empty files, zero bytes in size. Oh, great. Either DatCon has trouble with DAT files from the Mavic Air, or I missed some critical step. I'm guessing the former, since I work carefully, and DAT files seem to be pretty fluid in their structure. YMMV, mine sure does.
|
|