Any custom PX4 code for Mantis Q?

[j76]

Hey everyone, sort of like the title says...is or has anyone worked on any custom PX4 code (outside of Yuneec)?
I ran across this repo on GitHub: kyChuGit/YuneecDroneCode
But doesn't look like there has been any activity on it in quite some time.

The reason I ask is that there looks to be quite a few unused ports available on the mainboard (serial ports from the looks of them). There are a number of pads labeled tx\rx0,1,2,3,5,6,8 and wdm\wdp,etc. Also a debug header which is unpopulated. Pics attached.
I'm mainly interested because my mantis Q flies like crap and support in the US is non-existent anymore so I would just as soon use it as a test bed since I have two mainboards for it.

 

[Vasir]

j76,

The hardware is pretty slow and thermal throttling is limiting it badly. Lets say that you have full access to the hardware, what are you planning to do with it? I have designed many RTOS related pet projects, this one however I decided not to mess with. Now if we had headroom on the hardware side then I would agree with you. ( well legally that would be an issue (?). IMHO
For a starter drone I would say that its a good deal, if you want to play with drone os and related functionalities then there are many open source drone project you can play with.
My 2 cents.

 

[j76]

j76,

The hardware is pretty slow and thermal throttling is limiting it badly. Lets say that you have full access to the hardware, what are you planning to do with it? I have designed many RTOS related pet projects, this one however I decided not to mess with. Now if we had headroom on the hardware side then I would agree with you. ( well legally that would be an issue (?). IMHO
For a starter drone I would say that its a good deal, if you want to play with drone os and related functionalities then there are many open source drone project you can play with.
My 2 cents.

Seems you're pretty quick to stiffle any curiosity or creativity aren't you? As I said...my Mantis Q flies like crap. It's not interference as I live in the middle of nowhere and only get 1 bar of cell service from 1 provider on a good day.
The drone won't hold a stable stationary position no matter what surface it's over, or what height - unless it's in IPS mode...which then it will hold a stationary position but then it's height and speed limited. Also when in IPS mode it will randomly just switch to manual mode and take off in a random direction and not respond to controls.
The last time it did that and crashed the casing broke - so now ventilation is less of an issue.

Last time I checked, most drones on the market were using an F4 processor which is what's under the lid on the mainboard. Lower-Middle of the road if you ask me now that there is an H7.

So what would I do if I had access to the hardware you ask? I would do what anyone in the hobby would do in their spare time with their kid. Add various other sensors\controls to it and experiment with how they work.

Why do this with the Mantis and not "plenty of other open source projects"? Because I already have the mantis and wouldn't need to buy new controllers\esc's\etc.

Also, I'd hate to just toss a bunch of perfectly good drone parts on eBay when I can re-home them on a new custom frame eliminating heat constraints and have some fun and my son and I learning while doing it

 

[Vasir]

Not really, I disagree. I am all for creative juices flowing, that is what makes this "hobby" fun.

As I mentioned in my prior post, this thing is suffering from thermal issues (in my case). I understand that your issue might be different and that could be another nail on the coffin. Here in California where right now its around 105 @ times the image stabilizer is having hard time to give me stable images. My kid who usually plays with this drone tells me that when turning (180) visual artifacts appear (odd swings side to side, or plain old pixilation) when ever it gets hot (weather).

Now you are right and some people do tinker with such hardware and it might be great in places where not too many people are around. I think you are very lucky living in the "middle of nowhere", but that is why I mentioned the legal aspects of such modifications. You can do all the tinkering and post them here, other will get inspired and modify their drones using your ideas and bad things could happen.
I am an engineer and not a lawyer so please take that into consideration however I do not believe tinkering a closed system is a good idea publicly.

Q is a great drone and I enjoy it a lot, I hope that you can find a solution that works for you and makes you happy.
Have you tried to get the longs out and see what the board/sensors is/are telling you?
In my case I know that my drone was an early sample and had issues with GPS unit. Logs show at times drop/rise of > 2000 feet which means that either the GPS unit had issues and/or something happened internally that the cpu had issues understanding its position.

I think DoomMeister had a post that explains how to get the logs from the Q and how to parse and understand them.

 

[j76]

Not really, I disagree. I am all for creative juices flowing, that is what makes this "hobby" fun.

As I mentioned in my prior post, this thing is suffering from thermal issues (in my case). I understand that your issue might be different and that could be another nail on the coffin. Here in California where right now its around 105 @ times the image stabilizer is having hard time to give me stable images. My kid who usually plays with this drone tells me that when turning (180) visual artifacts appear (odd swings side to side, or plain old pixilation) when ever it gets hot (weather).

Now you are right and some people do tinker with such hardware and it might be great in places where not too many people are around. I think you are very lucky living in the "middle of nowhere", but that is why I mentioned the legal aspects of such modifications. You can do all the tinkering and post them here, other will get inspired and modify their drones using your ideas and bad things could happen.
I am an engineer and not a lawyer so please take that into consideration however I do not believe tinkering a closed system is a good idea publicly.

Q is a great drone and I enjoy it a lot, I hope that you can find a solution that works for you and makes you happy.
Have you tried to get the longs out and see what the board/sensors is/are telling you?
In my case I know that my drone was an early sample and had issues with GPS unit. Logs show at times drop/rise of > 2000 feet which means that either the GPS unit had issues and/or something happened internally that the cpu had issues understanding its position.

I think DoomMeister had a post that explains how to get the logs from the Q and how to parse and understand them.

My apologies then sir. I misinterpreted your post.

I'm in the Midwest (Missouri) so temps right now are fairly mild when I fly (or try to) fly the mantis. Mornings are normally in the 60's, afternoons may peak in the mid 80s until some point in July they may rise to the 90's.
I've never actually experienced any of the issues you mentioned with the camera. To be honest, the camera may be the best functioning part of my mantis.

I've pulled the logs but haven't looked through them yet. I was just looking into options because honestly before I scrap it I was just looking for something else to do with it. If I can't reload the stock firmware (in case it or parameters got corrupt), then I figured since it is based on PX4 then maybe I could load PX4 on it and go from there. My understanding though is that Yuneec doesn't make the mantis specific configs available - hence my asking about any custom code. Much like there is for the h500 or q500 series (whichever is linked in the way to fix drones project).
I'm not looking to do anything to the platform and post it for fame, I'm just looking for something to do with it. Learn, have fun, experiment, etc.

 

[Vasir]

Understood, now you get me into tinkering also. Many people here have issues with their wifi passwords.

 

[j76]

Not sure I've ever had an issue with the wifi password either....just the flight dynamics and random changing of flight modes.

I do have a correction though...I said it used an F4...however I took some more pics this morning and found it was an stm32f765. So it should have plenty of processing power as long as it's not thermal throttling. Which it could be as it is under a shield with no heatsinking whatsoever (not even thermal paste like other chips).

Also the wifi chip on the drone is an RTL8812AU which supports 2xMIMO but Yuneec has only implemented a single u.fl connector on the board. Another SKY85728-11 front end module and supporting caps\resistors would have to be added in order for the second u.fl connector to be enabled via hardware. The pads already exist...it probably would have cost them $5-6 to implement since the FEM is less than $3.75 online for singles. Perhaps I would have more than 1 bar for signal strength when the drone is 25ft away in an open field had they filled in those pads. And maybe more people wouldn't be starving for more signal strength and modding the stock system because they can only get 1\3-1\2 the advertised range.

Pic of the f7. Had to get at a steep angle due to the shielding.


Pic of the wireless chip area. Single FEM and pads for the second (traces are even there back to wifi chip).

 

[Vasir]

Interesting .... Now you are getting me into trouble ..... asking the boss for debug gear budget and I am sure she will not be happy.
The wifi password issue is: People forget the password and also lost the sticker on the drone.
Password could be cracked or pulled from the board.

By the way, I did make the antenna mod as few people here have suggested. That has helped with the range a bit but where I live wifi is very dirty. So it might help you since you have less issues with wifi.

 

[j76]

Haha... I'm very tempted to order the extra parts for that second antenna and see if that makes a difference. I'm not sure if it would need to be enabled via software\driver or not though.
I had a HolyStone hs700d where I was able to get root access and change the filesystem. Perhaps if enabling the second antenna required an update to the driver something similar would be possible if flashing the latest PX4 build wasn't a possibility.

In the meantime I think I'm going to see about the thermals of that f7 CPU. See if I can find a temp sensor and use an Arduino nano to do some logging.

 

[j76]

Haha... I'm very tempted to order the extra parts for that second antenna and see if that makes a difference. I'm not sure if it would need to be enabled via software\driver or not though.
I had a HolyStone hs700d where I was able to get root access and change the filesystem. Perhaps if enabling the second antenna required an update to the driver something similar would be possible if flashing the latest PX4 build wasn't a possibility.

In the meantime I think I'm going to see about the thermals of that f7 CPU. See if I can find a temp sensor and use an Arduino nano to do some logging.

I have the parts to add on the second antenna to the drone board. I just need to find the time to do it since it's a delicate operation. The caps and resistors are about the size of grains of sand.

Strange how Yuneec has docs submitted to the FCC with dual antennas though.... Makes me wonder if the dual antennas were an addition to the original or were the duals the original and then cut back for cost reasons?

YUNMQA Mantis Q Teardown Internal Photos 1 Yuneec Technology

 

[Vasir]

I have the parts to add on the second antenna to the drone board. I just need to find the time to do it since it's a delicate operation. The caps and resistors are about the size of grains of sand.

Strange how Yuneec has docs submitted to the FCC with dual antennas though.... Makes me wonder if the dual antennas were an addition to the original or were the duals the original and then cut back for cost reasons?

YUNMQA Mantis Q Teardown Internal Photos 1 Yuneec Technology

Very interesting ..... I think that would be their stock design and then mod it for different areas. Also that the antennas are part of the arms which would be a better design than sandwiching a single one in a dense area. What where they thinking?

 

[flashcoder]

Interesting .... Now you are getting me into trouble ..... asking the boss for debug gear budget and I am sure she will not be happy.
The wifi password issue is: People forget the password and also lost the sticker on the drone.
Password could be cracked or pulled from the board.

By the way, I did make the antenna mod as few people here have suggested. That has helped with the range a bit but where I live wifi is very dirty. So it might help you since you have less issues with wifi.

Sorry to hijack the thread, but I am one of those people with a lost WiFi password. The QR code sticker is gone. I know its been awhile but I would appreciate if you could help me to retrieve the WiFi password from the drone. Already took it apart a bit and pulled the internal SDCard, nothing in logs, there isnt any config related to WiFi neither..

P.S.
I would also be interested in getting PX4 on it if there have been any developments in this area. I know you said its closed platform but this is an old drone. I mean its old. Almost nobody have them anymore.

 

[DoomMeister]

Sorry to hijack the thread, but I am one of those people with a lost WiFi password. The QR code sticker is gone. I know its been awhile but I would appreciate if you could help me to retrieve the WiFi password from the drone. Already took it apart a bit and pulled the internal SDCard, nothing in logs, there isnt any config related to WiFi neither..

P.S.
I would also be interested in getting PX4 on it if there have been any developments in this area. I know you said its closed platform but this is an old drone. I mean its old. Almost nobody have them anymore.

Call Vertigo Drones and ask for support. They have the program to generate the password. You will have to open the body and get the MAC address from the label on the main/WiFi board. @joe@vertigodrones may be able to help here.

 

[flashcoder]

I have the parts to add on the second antenna to the drone board. I just need to find the time to do it since it's a delicate operation. The caps and resistors are about the size of grains of sand.

Strange how Yuneec has docs submitted to the FCC with dual antennas though.... Makes me wonder if the dual antennas were an addition to the original or were the duals the original and then cut back for cost reasons?

YUNMQA Mantis Q Teardown Internal Photos 1 Yuneec Technology

Have you tried to figure out the thermals and add the 2nd antenna?
Is there anyone else interested in such developments?
I believe thermals problem can be solved by adding very slim radiators where it is needed. They have done absolutely nothing to cool the chips, definitely there is alot of room for improvement. There are a couple pathetic thermal pads between chips and RF shield cans that are too thin to even fill the gap let alone provide any cooling effect, obviously noone at Yuneec cared about thermals.