< Previous Page
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Next Page >
November-December 2022 - The Resurrection

Wanting to take my mind off of this new bout of allocation anxiousness, I devoted my time to my poor drowned drone, the one back from early October on one of the sailing shoots for the "Staying Active" video. As you may recall, we left off at the point where we had semi-miraculously recovered the drone from the Ottawa River, and I had placed it in a cooler of distilled water for safe-keeping. But now, it was time to do something with it. Either trash it, or just rip it apart for some basic parts (like the props and maybe some structural components), or, ... or try and see if I could actually revive it. I had been doing some reading about electronics restoration after water immersion, and it seemed like there was a small chance that it could work. The most important fact in my favor at this point was that the drone had been sitting in freshwater, with a relatively low concentration of dissolved salts and other minerals.
From Distilled to Alcohol
My research had revealed that the first step was to attempt to fully displace all water before attempting repair. The folklore remedy for this was putting the item in a bag or tupperware filled with rice, but I learned that this was mostly an old-wive's tale. A better recommendation seemed to be an intermediate dunking step but in 100% isopropyl alcohol. This would completely displace existing water (hopefully along with any dissolved substances) while avoiding any drying out. And then, after a period of time, I could take the drone out of the alcohol and air-dry it without any water (or minimal water) present.

I felt the alcohol immersion would be more effective if I first dissassembled the major pieces of the drone into a set of sub-parts. This way the alcohol would have an easier time getting right into all of the little bits and pieces. After an evening of YouTube watching and following, I had carefully had the drone into nine or ten major pieces.
Initial Dissassemble
Overall (at least mechanically), everything seemed generally ok. The prop motors turned smoothly - save for one that seemed to have a little grit in it; the camera and gimbal assembly seemed to articulate well, with no stiction. The fan that cooled the main block of electronics, curiously, was the only location where I found real physical damage. The plastic casing/shroud around the fan had cracked. Weird. Maybe the sudden deceleration from spinning at max rpm to zero when water entered had done this? In any case, that seemed pretty straightforward to fix.
Initial Dissassemble
The main electronic board and a few of the structural metal supports did seem to have a bit of residue on them - probably the beginnings of salt deposits and mineralization. I put them aside and continued dissassembly.

After getting all the components apart, it was straight to 100% isopropyl alcohol immersion, where I left the pieces soak for a good week or so.

I retrieved the week long-submerged drone components and laid them out on the table before me. They quickly dried off, filling the room with the characteristic sharp antiseptic smell of isopropyl alcohol. I then examined them more closely. Of all of the components, it seemed that a long, gray, somewhat sarcophagus-shaped slab, that was the main electronics board. It was the item that seemed the most "dirty" - with numerous chalky white deposits on it. I also realized that what I was looking at was an exterior casing around other, more interior components. That had to come completely apart if I was to truly clean everything.
Mainboard, Air 2S drone
I carefully pried open the slab to reveal various little walled-off compartments, each covered with their own little panel of perforated metal. I presumed these were there to minimize electromagnetic interference. In turn, I carefully pried open each of the panels. This revealed all of the major integrated circuits that controlled the drone. All of them had a dollop of blue heat-conducting compound. I'd have to remember to re-apply that compound when I eventually put everything back together.
Mainboard, Air 2S drone
Water - and deposits from water - had clearly made their way into all of the nooks and corners. Many of the tiny little surface mount components had a bit of whitish powdery looking stuff near or on them. The leads of the large surface-mounted chips had gluey, whitish stuff in between the leads. The concern, of course, was that these substances would conduct electricity and possibly cause various random short-circuits.
Cleaning Materials
Basically my plan was to try and clean off all of this crusty buildup, without damaging the components themselves. To that end, I had purchased a few cans of a respected organic/alcohol-based solvent, a few cans of compressed air to blast and quick-dry, and various brushes, both hair and foam-based. I figured that I should be able to clean away most everything, given enough focus and time. Hopefully I wouldn't sever any leads in the process.

I started off fairly gently, not wanting to damage what I imagined to be very sensitive components. And although some of the initial deposits came off, there was always a bit that remained, and my cleaning actions gradually became more vigorous. It was a long cycle of spray, scrub, spray, scrub, rinse ... and then I'd come back the next night and do some more. And even though the drone is a small device, at the rate at which I was managing to clean the boards, it was slow going. Towards the end, when I was still encountering stubborn deposits, I took to applying a mixture of baking soda and water and liberally applying it as a wet paste and then letting the board soak, before rinsing off with distilled water and then the chemical cleaner. I even resorted to using my dental waterpik to shoot stronger jets of cleaning water after baking soda soaks. Finally, after about a week or two of this, I got all of the electronics to what I felt was a decent level of cleanliness. The question was ... had I broken anything in the process? I mean, there were some *really* tiny little components on those boards.
Motherboard Cleaning
Soda Soak
Cleaned Mainboard
Cleaning Before/After
Re-assembly
Ready for first power-up
With boths sides of the main motherboard and the daughterboard cleaned to my liking, I then spent a bit of time on other components. As I said before, all but one of the prop motors ran freely, and I did a cursory chemical rinse and dry of them. For the forth one, I had to spend many cycles of spray and clean, also at one point spraying cleaner while spinning the assembly with compressed air on an attached propeller. Eventually I got all of the crud out of it. I had taken an exploratory look at the gimbal and camera assembly, and surprisingly, it seemed pretty clean, so there was not much to do there. I carefully re-glued the cracked main fan casing with superglue, and it now seemed sturdy enough and the fan still spun smoothly. And so, after waiting for everything to dry, I carefully re-assembled the core components - but did not put all fasteners or covers back on. And then for the first time in two months, I put a fresh battery into the drone and standard DJI two-press push to turn it on.



And lo and behold ... it actually powered up (I guess I was having a hard time believing that a week on a murky riverbottom and a month of vigorous spraying and soaking and scrubbing hadn't fatally damaged it in some small but critical way). But nope, it started up. It did the little DJI beepy sound, and the fan started, and the lights came on as expected and the gimbal did a range-of-motion self-test. Wow.

From: Luke Ward 12/6/2022, 10:53:53 PM
Wow. Is that running its diagnostic boot up procedure?!?!?

From: Andrew Lavigne 12/6/2022, 10:54:09 PM
that's what all dji drones do when they start up.
so at some base level, the thing is "sane"
which in itself is pretty remarkable

From: Luke Ward 12/6/2022, 10:54:52 PM
Well ... that's impressive to say the least

From: Andrew Lavigne 12/6/2022, 10:55:10 PM
I was half expected some weird error beeps and a flashing red light and nothing else. or just dead.

From: Luke Ward 12/6/2022, 10:55:24 PM
Just dead I was expecting

From: Andrew Lavigne 12/6/2022, 10:57:13 PM
next step is to see if it will pair with the remote controller


I did just that during my next round of basic sanity experiments. And again, happily, the remote control paired with the drone. GPS lock and telemetry were established. I could even get the props to spin. But, alas, no video feed. Just a message that said "incompatible image data". Unsurprisingly, not everything was perfect.
First Flight!
The next day, I took the drone outside just to see if in fact all major systems (Inertial Measurement System, Propeller motors, GPS lock, obstacle avoidance sensors, general stability and control), if they were all working. Without that basis of functionality, there would be no point working on fixing the video feed.

And well ... success here, too! The drone started, flew, and hovered completely normally. Controls responded normally. The drone flew perfectly well. Excellent progress.

But the video was still dark. And a mobile camera platform without a working video feed is pretty useless. I tried quite a few things over the next few days. No luck. Researched on the forums. No luck. Some talk about having to match the gimbal's software with the main body's software. But these were linked components before, so couldn't be that. To this point, I had only taken out the camera module itself, did a little bit of eletronics cleaning, and had a close look at the lens (which visually appeared fine - no fogging or water spots or anything). But doing that one more time, still nothing. No video.
A few days passed. I thought, maybe I should take the actual gimbal mechanism itself apart (this is the part that articulates in three different axes and keeps the camera steady). The bundle of control and video cables also ran through the joints of these axes. Maybe something was pinched or broken or corroded. So, I fully dissassembled the articulating parts of the gimbal (which I had to this point avoided because it was tricky). I saw nothing obvious that looked like breakages. There was a bit of gunky stuff on the cable bundle, so I did a bit of cleaning, then re-assembled everything (which was again, kind of tricky). Then finally attached the connector for the bundle into the right spot on the mainboard. And made sure it was firmly in place.

And then I powered on the drone again. I didn't expect anything different, since I had not really noticed anything amiss. But then ... a view of my desktop. on the display! Whoa! Video!

From: Andrew Lavigne 12/14/2022, 9:49:43 PM
Holy crap

From: Luke Ward 12/14/2022, 9:50:11 PM
It's a sign!!!!

From: Luke Ward 12/14/2022, 9:51:23 PM
you actually have a feed?!?!?!

From: Andrew Lavigne 12/15/2022, 12:05:03 AM
Yep

From: Luke Ward 12/15/2022, 12:05:38 AM
Well do tell. More details please.

From: Andrew Lavigne 12/15/2022, 12:39:46 AM
I can't say for sure, but I took the gimbal apart again and noticed some areas of deposits on the electronics on the outside of the sensor itself. Cleaned that, did some tests, nothing. dissassembled fully to examine the main wiring harness to the camera and didn't really see anything that looked like breakages.
re-assembled everything again the next day (today) ... and .... I have a video feed now
one possibility was that I had attached something backwards? or ... the cleaning helped. I can't be sure
A video feed!
So yep, I got the video feed working. Luke and I speculated that perhaps I had attached the main video cable backwards? Really, I couldn't think of much else. But no matter, I had pretty much a functioning drone now. A nice win. Hopefully portents of things to come?
< Previous Page
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Next Page >
Send feedback or leave comments (note: comments in message board below are separate from those in above message board)
send feedback | homemade custom message board
(Message Board failed to initialize. )
Web Page & Design Copyright 2001- 2025 by Andrew Lavigne. (Privacy Policy)