Close to giving up drones

General Rants&Raves
1

Hi everyone,
after being involved with drones for 10 years now I am getting to the point where I’m close to giving up on drones and looking for a different hobby altogether.
The reason ?
So many crashes over the years compared to the actual flight hours which takes the fun out of it all.
…Been flying Octocopter, Hexacopter and more recently a large Quad. Mostly all self build except for two small ones I sometimes just fly for a bit of fun without all the GS setup and telemetry link.

I had some fly aways, one of which was lost for ten months whilst another was never seen again. I had crashes over the years caused by: Prop failure, ESC failure, software issues, multiple GPS related crashes, PM burnout due to poor quality manufacturing, loose solder joint on a battery connector, slow SD card causing FC crash which of course led to drone crash,…

The most recent crash just happened yesterday with a Quad based on a Tarot X4 frame and a 6S Battery / Motor setup. The first drone I had achieving around 20 minute flight time.
Finished the build in late May and by early June performed Auto-tuning during which it suddenly crashed due to burnout of a power module. Turns out was poor quality soldering - unable to handle anywhere near the rated amperage. After about $350.- in parts and a 6 week wait for various parts to arrive it was operational again.
Now I’ve only done about 4 flights since the repair and now once again a crash.
Quad Crash -August 2021

As can be seen not much left in one piece and the fact that I couldn’t find it until this morning and we had rain during the night didn’t help much either.
So essentially over $1000.- in damage for a drone that has had less than 4 hours total flight time!
:frowning:

I spent some time looking at the telemetry data ( .bin not retrievable as SD card on FD got smashed as well).
What I struggle with is the fact that everything started well and no problem shown in first minute of flight. Then there was a brief GPS glitch but recovered quickly and continued on its mission. Then there was a EKF variant issue apparently due to compass issues,…which is interesting given that my FC has got 2. Once again it recovered and continued on its mission. But then there was another GPS glitch at which point I tried to initiate a RTL but that was overridden due to the GPS glitch. If that wasn’t bad enough yet another EKF variant occurred, this time related to velocity sensor. Then once again I tried to manually take control but despite the fact that RC signal was received the FC ignored my inputs and proceeded with its own Land command.
Whilst this might be fine in an open field, it is not beneficial over a rainforest like in my case or over water.

So, my questions here:
How can it be that within seconds of each other a GPS glitch is encountered and then EKF variant issues, first related to compass and then to velocity, despite no prior issues or interference?
Been trying to find setting in regards to overriding Land command and having Alt-Hold instead when EKF failsafe is triggered. Does anyone know where to find it ?
Is there a way to have two GPS units attached and in that way redundancy build in ?

…So in a nutshell after spending over $6000.- over the years on various repairs and replacement items I’m just no longer happy proceedimg with it if I can’t achieve a more reliable setup.

Thanks

2 Likes
2

I hear ya. I’m planning on a rover build next for the reasons you detail. I haven’t had a crash in many years but my anxiety level while flying is much higher now and I go over everything about 10 times before I take off. Flights are always very short now then I analyze the logs. Feels like work sometimes not a hobby.

I added a 2nd gps to my pixhawk1. I’m hoping that helps with accuracy. There are several options in parameters to choose whether it chooses the best, the primary, or blends the 2.

Regarding the mag, might be better off disabling the internal. I was having problems until I did that, flies much better now with just the external (no toiletbowling or compass errors)

Good luck and sorry to hear about your crashes. I hope things go better for you

1 Like
3

Hi Derek,
yeah, considered something like a rover. Unfortunately I am on a rural property with some steep hills and as such would need something the size of a quad bike but with very low centre of gravity and perhaps tracks instead of tyres. But that would be also a significant investment with limited benefit for practical applications.

I’ve been using the Navio2 / Raspberry Pi combo in recent years and like how versatile it is and allows me to tinker around with additional features I prefer. But this FC also suffered in the crash.
Would have to investiage if and how a second GPS unit could be attached and what settings would have to be changed as you’ve mentioned.
…but at the moment certainly not in the mood or in the financial position to do anything more with drones.
Thanks for your open input. Good to hear that I’m not the only one with second thoughts.
(I guess if it was all perfect for everyone we wouldn’t have this conversation)

4

Maybe put it all in a box and come back in a little while? I did that a couple years ago and I’m now just getting back into it.

3 Likes
5

@Karl_Schoelpple Can hear your exhaustion. yeah, put it all aside for a while.
Why not get a small nice inexpensive fixed wing and fly around just for the pleasure of it. Unlike quads they don’t just fall like a brick from the sky if something fails.
Best
k

6

Hi Karl, @Karl_Schoelpple

I used every Navio from the very first one (no +, no nothing) to a Navio2 and an Emlid Edge. I really liked the idea of having the companion computer and the FC combined at first, but I always had very mixed results regarding the Navio GPS, the Edge’s GPS module is great. My copter would sit on the ground with a perfectly fine GPS fix (12 sats+). Take off and suddenly no flightmode beyond stab or althold would work. I landed and checked the GCS and sure enough, the GPS lost the fix. I left the copter on the ground, waited for some time, no fix. Power cycle, GPS fix, take off, fix gone.
I received several replacement antennas, a replacement Navio2 (nothing against Emlids customer service), but the GPS was always lacking.
A side by side test with a Pixhawk and a cheap M8N GPS indoors, led to the M8N GPS having a fix 2m from the nearest window after a few seconds, the Navio2 GPS never reported a single sat.
The versatility you mentioned is not really there, in my opinion. With any MCU based FC you get more connectivity (CAN, several UARTs…) than the RPi/Navio combo can offer. USB host is no real argument, because if I need USB, I simply choose whatever companion computer I like and use it.
I also do not like the idea of running the flightstack on a computer. I had several SD cards fail, because of power cycling. Having to shutdown the FC through ssh every time is a hassle. I can power cycle my Pixhawks, Matek Wings, etc all day without an issue. And after power up, they are up and running immediatly.
I could go on about this (no USB device mode, slow official update cycle, lacking wheel encoder support…).
I have six rovers now (and several copters, planes with and without ardupilot, but I am not flying at the moment). Rovers are great, they do not fall from the sky, you can build them really sturdy, they can carry a mass of payload and drive for hours. And the biggest plus, no one cares if you drive them around in the neighbourhood. No regulations and other bullcrap applies.

3 Likes
7

Dont know about anyone else, but I feel something might go wrong any moment with my plane while mapping missions. Although I am using high quality peripherals and cubes, and following guidelines with regards to powring them, I have a fear of flying that I cannot seem to part with despite having flown VLOS and BVLOS upto 4 to 5 km with range remaining as apparent from the radio messages.
Strangely, having access to fixed wing PPK capable drones, I seem to be more comfortable with DJI along with lots of GCPs.
I am unable to be sure with regards to my tuning and everytime I see the logs i begin to think something is wrong and am unable to find conclusive advice. Maybe I should get a mentor or paid log review from someone experienced.
Any advice/comments would be welcome.

8

I was cheated by a drone supplier from China in 2018 and thats when I vowed to make my own. I seem to have achieved the target but somehow it all seems empty.

9

Ahhhhhh…
Consider flying electric RC planes, wings, or even gliders and slope wings. Has some tech. As much or as little as you want. Has some build talent. Has some artistry if you desire. Best of all there is that glide path.

Completely understand where you are coming from.

Also how about a solar panel tracker?

Cheers … MowJack

10

Rovers are fun! During our shutdown last year I had a great time driving around with FPV. Put a 2W VTX on it and really covered some ground. The same fun build/tuning process but the worst that happens is it runs into something and strips a Spur gear. On mine it’s like a mechanical fuse :slight_smile: and they are cheap. Auto Missions are cool. All around a fun aspect of Ardupilot.

11

Thanks for all those replies.

Good to hear (at least from my current perspective) that I’m not the only one suffering from “drone fatigue” .

Yes, one can be a bit anxious after spending all that time and money on a drone especially when build from scratch rather than buying a kit or ready made unit which comes with some certainties in regards of expected success.

As to fixed wing aircraft: Probably crashed at least 4 over the years as well. Not due to technical failure but rather due to bad combination of steep countryside, lots of trees, no flat and open area to take off and land from and doesn’t help being an inexperienced pilot when it comes to fixed wing aircraft. - Hence was doing much better with drones,…at least in regards to my piloting abilities.

I think right now will be going for Derek’s advice and put everything in a box and let some dust settle on it for a while.

12

Sounds like a good plan. One last word…VTOL :grinning:

13

Don’t go there…

…That’s how it all started:

Last year I’ve ordered a Freeman 2100 because I wanted to explore the VTOL technology now that there has been more support added in ardupilot / copter. But the actual aircraft never arrived and had to do all sort of things to get my money back after almost 6 months of back and forth and multiple failed delivery attempts. Yes, I know there is a slight issue with a certain Virus doing the rounds but companies are all to happy to sell you something and then can’t deliver. Strangely enough received the electrical components and decided to keep it for a future second attempt. So, since I had those parts sitting in a box I’ve decided to get a frame and make a quad out of it for the time being.

Well, …the rest you can see in the photo above.

14

There will be a time in the future when you’re again excited about it. Maintain momentum by switching to another project that is fun.

15

Have you considered getting into miniquads? Most of them are designed to withstand hard crashes.

16

Hi Isaac
yes, do have two mini quads.
Good fun indoors but rarely bother taking them outdoors as we often get a slight breeze and then those small quads getting easily pushed around. Given the proximity to trees it doesn’t take much to run into trouble if a sudden breeze pushes them towards trees. Also run into trouble in the past by trying to fly above the trees only to discover getting quickly out of range and hence loosing control.
…but here and there good fun without having to setup GS, telemtry link and everything else.
Must say been impressed that for some of those small drones you can even get spare parts.
The large quad that crashed was very powerful (total of 6kW motor power) and hence didn’t care about a breeze. In fact after flying Hexacopter and Octocopter in the past this was the most stable in loiter mode. …even with 20km/h wind gusts.

17

A 4 or 5 inch quad on 4S power has plenty of power to deal with wind, even a lot of wind. I have been experimenting with fast auto missions running ~100kph. And as @tenten8401 noted these can be tough little quads.

1 Like
18

I hear you. Lost a self build in June "Radio Failsafe - Disarming" and haven’t mustered the enthusiasm to start over, yet.

19

Hi Dave,

that’s interesting to hear. In the past I’ve had plenty of GPS related issues (glitches, drop-outs, processing delays,…) on various drones with different configurations during missions just doing around the 25km/h range. So to to hear that you manage that at speeds of up to 100km/h is rather impressive.

Especially looking at the maths of it: If flying at 100km/h that equals about 27.8 m/sec. Most GPS these days run at 5Hz despite claiming that they can do up to 10Hz (ideal case). The older ones only did 2Hz or just 1Hz. More than 10 years ago it was a struggle finding those doing reliably at least 1Hz. So, back to the calculation,…with a GPS doing 5Hz you would cover (27.8 divided by 5) =5.56m. That doesn’t allow for much of a safety margin at that speed because if you had for some reason a GPS glitch lasting just 2 seconds the drone would have already covered over 55m in distance without really knowing its actual position.

…As mentioned, impressive.

Anyway, appreciate the input but all my past drones and batteries been 3S and the latest one was 6S. I don’t think I would currently be inclined to start all over with 4S now. Besides, for the first time managed to get past the 20 min. flight time mark on that latest drone build and was considering changes to get closer to the 30min. mark. (…until that crash happened.)

20

There have been several improvements in the latest releases giving stability to high speed in assisted flight modes.

Here is one trial mission I did on a 5" last week. ~70° pitch ~30 m/s. It’s about all this quad will do. I have a 4" that’s a bit faster but still doing some fine tuning.

Speed mission
Speed mission3382×1032 181 KB