Well, I should have taken the advice of, pretty much everybody, and built a simpler model as my first build. But I didn't... I jumped right in and built a DYS D800 X8, and a Tarot X6 at the same time. I took the D800 X8 out yesterday for her maiden flight.

I really took my time building this thing. I studied everything I could find. I used a Pixhawk, and I very carefully did all of the calibrations, and motor checks in Mission Planner. From my limited experience, it all looked good. Motor numbers, and rotations all look good. I was very careful about installing the props. I'm pretty sure I got all that right.

Here's what happened: I found a great place to maiden flight. Set up everything. radio on. Battery on craft connected. Tower as a GCS. Everything looks good. Good GPS. I armed the craft, and got very nervous, so I disarmed. Did that about 7 times. Finally got up the nerve to go for it. Throttled up, and she rose a bit and immediately did a very fast back filp to the port side and landed about 10 feet away. Some damage, but it can be repaired.

Now I'm trying to figure out what went wrong. I'm a beginner at all this, and reading flight logs. At first glance it looks as though the voltage drops out, and the current spikes as soon as I throttle up. After adding a few more lines to the chart, it looks as though the barometer is going crazy each time I armed/disarmed (when the props spin). The barometer really went crazy on the fateful throttle up.

Could one of you kind folks please take a look at my flight log to help me determine what went wrong and how to fix it? I'm a beginner who took on way too much. However, I am a quick learner. I just need some help. I have attached the log. Any help would be greatly appreciated!

Views: 950

Attachments:

Reply to This

Replies to This Discussion

My first quad was a f450 every intent for fly this machine I always recorded with my camera his Movements. For all people beginner should record his first flight of his quad. I recomend you read the coments of this post

http://diydrones.com/group/arducopterusergroup/forum/topics/apm-2-6... 

Best regards

fr

While it won't help for this crash it may help to prevent future ones. I do this on all new builds:

"On shortish grass/lawn, in an open area, in STABILIZE mode (NOT Loiter or PosHold): now carefully throttle up slowly so the multirotor is just starting to "get-light-on-the-skids", DO NOT throttle up so much that it takes off! If it's overpowered or low kv motors/big props this point is maybe at 25% of your throttle stick travel, otherwise normally a bit further up.

In other words apply just enough power for some lift to be created, but only just.

Now try to pitch forward gently, starting with 10% stick, you should hear and see the rear motors speed up and the fronts slow down, the craft should also tilt forward slightly.

Do the same for right roll - gently - right motors slow down, left motors speed up, craft tilts to the right slightly.

If this doesn't work exactly like this then something is wrong so don't try take off!"

Maybe you've already figured this out, but looking at your logs, you only applied a small amount of throttle, and motor 3 went almost to full power.

This would probably be because the flight controller is trying to level the copter by increasing motor 3 power, motor 3 is connected to the the wrong motor making angle error worse... so it applies yet more power and keeps doing this until it flips.

I'd be 99% sure this has happened either because the ESCs are connected to the wrong outputs or the flight controller orientation.

I've just finished a build using this very frame and a Pixhawk, and it was really confusing as the motor test function in mission planner uses letters which don't correlate with the numbers in the APM copter guides, and neither of these match up with the ESC connectors on the frame.

Use this diagram I made to match up the letters (IGNORE numbers) with the motor test function in mission planner, double check you've mounted the Pixhawk arrow forward and you should be good to go.

Attachments:

....and while I think of it, ignore the CW / CCW markings on the frame. They are for a DJI flight controller, Pixhawk is different, make sure the motors rotate the way marked in the diagram from my previous message. 

Hey Justin, Thank you for taking such a detailed look at my crash. I really appreciate it. Your timing is especially good, as I plan on testing her again on Friday.

I hear what you are saying about the motor numbers being set up for a DJI. I realized that before finishing the build last time. It was a major pain getting it all right... but I guess that I did not get it right. Strange, but I thought I had it all right after doing the Mission Planner motor testing. I'm new at this, so I must have made an error.

On this second go-around, I desoldered the esc, and ground wires from the lower board. I then extended each wire and sent the the proper numbered wires directly to Pixhawk (bypassing the pins on the frame). It all checks out in Mission Planner again.

I have rigged up a set of PVC pipe extenders for the legs, and a tether to the ground, making it impossible for it to flip. I'll check the thrust/roll/pitch/hover with that rig before ever setting her loose again. This has been an expensive learning opportunity. Like I said, I took on way too much for my first build.

I would like to see the motor information myself. Where do I look in the log?

Thank again!

Justin R said:

....and while I think of it, ignore the CW / CCW markings on the frame. They are for a DJI flight controller, Pixhawk is different, make sure the motors rotate the way marked in the diagram from my previous message. 

I plotted RCIN.C3 (throttle) vs RCOU.C1 through  RCOU.C8 (motors)

(have a look at attached image, red line is throttle, others are motor outputs)

Good luck with your test.

Justin

Attachments:

Hey Justin,

I finally figured out that the rcou... parameters were actually the motors. Thanks for the graph. Does the voltage/current graph concern you at all? Seems strange that the voltage drops, and current spikes as soon as I throttle up.

My test today went a bit better. I rigged up a system where I extended the legs horizontally using PVC pipes. I also used a dog leash lawn screw, and some p-cord to keep the copter from lifting more than a foot or so. Impossible for her to flip. OK, this time she lifted level! Problem is, as soon as I throttled up, she went in to battery failsafe/RTL. At that point I lost all throttle control, as, I assume, she was trying to land. She just hug there tugging on the tether. Does it make any sense to you that she might have gotten confused about landing due to the prop wash? I could not throttle down at all. I finally got her to land and disarm by switching the radio off. I have uploaded today's log. Could I bother you to take a look?

Attachments:

Yes, the battery failsafe triggered and then it switched to RTL and so actually tried to gain altitude to fly to the home location. 

It looks like your motors are wired up OK now though which is a good start.

Your battery voltage does sag a LOT as you say. What are you using? Also, what are you using for the current / voltage sensor?

Thanks for taking a look Justin. I really appreciate it. Always amazing when total strangers are willing to help out. I suck at building and flying, but at least I have interacted with some really cool people.

I'm using this battery: https://hobbyking.com/en_us/multistar-high-capacity-6s-10000mah-mul...

What battery are you using? I'll order them tonight.

Thanks!

I had a few kind people help me out when I first got started so feel obliged to give something back!

I'm using the the 12000mAh version of the Multistar - the 10,000mAh is totally fine though, although I'd suggest running two inparallell as you won't get great flight times with only one.

I think we need to focus purely on the power system. Your log shows that with all 8 motors at full throttle, you're only pulling 56 amps so your current sensor has not been calibrated properly.

Are you using the standard 90A sensor that comes with the Pixhawk?

If so, you need to replace it STRAIGHT AWAY (don't fly it!) as it's really not designed for big rigs like this and they can burn out cutting all power to the copter

I'm using a Mauch current sensor (http://ardupilot.org/copter/docs/common-mauch-power-modules.html) and would suggest you buy his 200A sensor and his 5.35V 2-6s BEC as a minimum. They're great and the developer (Christian Mauch) responds to e-mails and even gives you calibration values you can just type into mission planner so you know you'll be getting really accurate readings.

I've done this slightly differently and used 2 x 100A sensors, the "Sensor Hub X2" and 4-14S Hybrid BEC as I am using two batteries in parallel.... if you can spend the money (and you really shoudl!) this is a very worthwhile upgrade.

You'll need to be handy at soldering though as you'll need to wire these directly to the main PDB and make a really strong connection to handle the current.

OK, good to hear you are using the Multistar 10C batteries.

I'll order the suggested parts tonight. Did you order from Mauch, or one of the US resellers? Craft and Theory has a Pixhawk/2 lipo kit.

Thanks!

I'm actually in the UK and ordered from AltiGator but the 2 LiPo kit from Craft and Theory looks like everything you'll need.

Reply to Discussion

RSS

© 2020   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service