Hello all you big-brained folks! I am attempting to determine the level of effort as well as the practicality of converting the flight controller on my M200 to a Pixhawk 2. I love the airframe and quality of the DJI products and the M200 fits my needs (almost) perfectly. However, going forward I do require more flexibility especially in the area of flight modes and functionality when communicating with an on-board computer in gps deprived environments.

We fly LiDAR equipped drones underground in hot/moist environments for the mining sector and so far the flights have been entirely pilot controlled, either line-of-sight or via a custom FPV camera. Currently there are issues with flight modes where the barometer readings fluctuate heavily and the thing can sure be a handful to fly in ATTI mode. We are looking to the near future where we will require connectivity to the flight controller for obstacle avoidance, then semi-autonomous flight and then full autonomy.

  We have quickly outgrown the M100 and simply do not want to waste any development cycles integrating a pixhawk onto it and would like to skip right past it and start modifying the M200 to suit our needs. Does anyone have any experience doing this? Theories, thoughts?  I am an experience drone pilot but my level of expertise falls a bit short when it comes to a project of this magnitude.

  My first step is to determine how to get the pixhawk to work with the ESC's and motors on the M200. Is this a pipe dream? Should I simply seek out a different platform and save the nightmarish conversion process? If so, any thoughts on a suitable platform that has a relatively small footprint and the ability to lift the same type of payload as the M200?

  Any and all comments, suggestions, instructions, referrals and theories are absolutely welcome! Thank you so much for taking the time to help a fella out.

Excuse my off topic post, but you're serious when you tell me you use sUAS underground with LiDAR for mining applications? This is possible? lol

My 2 cents say that you're building a custom platform, so get one with 3rd party reliable ESC's and motors that are known to work with pixhawk. From there look for the integrated solutions you're after, the M200's navigation and guidance are undoubtedly not built for your application so I'm a little skeptical about it working just the way you want it to with their base algorithms, but again I have no knowledge of how you operate underground and a lot of things I just won't understand as well as you.

Do you have any current systems in place that you know work well underground for navigation purposes?

tl;dr get a package with the basics required for flight already integrated, and cater the flight systems to your operational needs.

  
  I am entirely serious. We do it all the time. Check out our website safesightxp.com. (a bit outdated since we move SO damn fast in our dev cycles and product functionality and form factor)

  My experience with integrating FC's with propulsion systems is fairly limited so I certainly hear you that we may just want to go a non-dji route but I just love the sleekness and quality and durability of the M200. We are about to transfer our dual LiDAR hardware over to it and it will work just fine in the short term. Using another platform would shift our focus to waterproofing, increasing durability and user interface etc. I am riding a fence here because I just have not seen another platform that offers the same level of "pro's". I'm sure one exists but I just haven't seen it.

Jeff Shoults Posted at 2018-1-23 15:02
I am entirely serious. We do it all the time. Check out our website safesightxp.com. (a bit outdated since we move SO damn fast in our dev cycles and product functionality and form factor)

  My experience with integrating FC's with propulsion systems is fairly limited so I certainly hear you that we may just want to go a non-dji route but I just love the sleekness and quality and durability of the M200. We are about to transfer our dual LiDAR hardware over to it and it will work just fine in the short term. Using another platform would shift our focus to waterproofing, increasing durability and user interface etc. I am riding a fence here because I just have not seen another platform that offers the same level of "pro's". I'm sure one exists but I just haven't seen it.

Take a look at the Wind platforms. Expensive, but in your line of work... I think it's more than worth it.

http://electronicarc.com/catalog ... 85.html?language=en

I'm curious how you manage to align the point cloud from the LiDAR, as our team is having some issues doing that at the moment. But I understand some of your process is what makes you money, so if you don't want to discuss it don't worry about it. I'd be open to emailing and bouncing ideas off of eachother as professionals, feel free to PM me.

I looked around your website, I see you're using DJI's guidance. That's pretty easy to integrate into any A3 platform afaik, so the wind would work with it.

  The exact process is the department of of our Software Engineer but we use a custom SLAM algorithm. It is bloody tricky tuning SLAM parameters to produce the trajectory but we have some pretty smart cookies on our team so we are there. We are very much believers in collaboration and unlike a lot of other companies we are fairly open about our processes and technology. Businesses are built on relationships and a rising tide raises all ships. We are confident that there is plenty of business out there for all of us and good relationship skills with clients will bring quality companies to the top. We will all eventually have the same tech in our hands because this all moves so damn fast so why be secretive and weird about it right?

  The Wind series looks fantastic and our dev work to get a pixhawk into an M200 might pave the way toward using that platform in the future for sure!

Jeff Shoults Posted at 2018-1-24 07:27
The exact process is the department of of our Software Engineer but we use a custom SLAM algorithm. It is bloody tricky tuning SLAM parameters to produce the trajectory but we have some pretty smart cookies on our team so we are there. We are very much believers in collaboration and unlike a lot of other companies we are fairly open about our processes and technology. Businesses are built on relationships and a rising tide raises all ships. We are confident that there is plenty of business out there for all of us and good relationship skills with clients will bring quality companies to the top. We will all eventually have the same tech in our hands because this all moves so damn fast so why be secretive and weird about it right?

  The Wind series looks fantastic and our dev work to get a pixhawk into an M200 might pave the way toward using that platform in the future for sure!

Interesting, we're looking into localization but that can be... difficult.. to say the least. I respect your view on business, I believe we share the understanding that collaboration can solve a lot of problems. My background is that companies are very careful with their handling of their process so I do my best to generally respect that, but we aren't working with ITAR so much of it is unnecessary.  If you're interested in collaborating with our engineers who are working with LiDAR let me know and I'll get you our robotics engineers email

The commercial UAS data collection market is more than large enough for everyone and their brother to join in, provided they satisfy their niche accordingly.

What will you do with the battery? Is it possible to connect that to the Pixhawk 2?

You have a very interesting set up. However, the controller for the Matrice 200 is not going to be compatible with the Pixhawk 2. At the very least, the firmware would not be compatible. You would want a controller that is more third party in this situation.

DJI-Mark Posted at 2018-2-16 13:45
You have a very interesting set up. However, the controller for the Matrice 200 is not going to be compatible with the Pixhawk 2. At the very least, the firmware would not be compatible. You would want a controller that is more third party in this situation.

Now I'm kinda curious.

Are you saying that the Pixhawk 2 won't work with the ESC and batteries of the Matrice 200,
or do you mean that the Flight controller wont work with the Pixhawk 2?

JanPieter Posted at 2018-2-19 00:50
Now I'm kinda curious.

Are you saying that the Pixhawk 2 won't work with the ESC and batteries of the Matrice 200,

To make sure I am answering your question correctly, can you send me a personal information on what your exact set up is looking to be? Please provide the model controller you are referring to.

Here's where I am at now: It appears that the pressure range on the barometer is simply not wide enough to operate at the depths we are running. Last week I tried to start the M200 deep underground and got a barometer error that prevented operation of the drone. Based on a quick barometer lesson it appears that the mems sensor barometer has a fairly specific range of values and higher air pressure underground exceeds the limits of the sensor, causing an error and it is unable to fly. This sucks bigtime!

This leads me back to my original issue of installing a Pixhawk into the M200. Does anyone know if the ESCs will even talk to the pixhawk? Or am I completely screwed and have to abandon the M200 for a pixhawk platform?

Any thoughts on how it would react if I were to "trick" the barometer? For example: if I were to simply enclose the whole baro sensor (in a ball of hot glue for example) would zero change in the pressure sensor cause erratic flight? My worry is that zero change in data from the sensor would make it think that my throttle input is having no effect and it would attempt to climb until it hit the ceiling, regardless of my stick input. I love the M200 and until recently we had not been this deep underground so the problem had not shown itself. It would be a shame to have to switch platforms.

Here is a photo of our current build...

Jeff Shoults Posted at 2018-5-28 07:29
Here's where I am at now: It appears that the pressure range on the barometer is simply not wide enough to operate at the depths we are running. Last week I tried to start the M200 deep underground and got a barometer error that prevented operation of the drone. Based on a quick barometer lesson it appears that the mems sensor barometer has a fairly specific range of values and higher air pressure underground exceeds the limits of the sensor, causing an error and it is unable to fly. This sucks bigtime!

This leads me back to my original issue of installing a Pixhawk into the M200. Does anyone know if the ESCs will even talk to the pixhawk? Or am I completely screwed and have to abandon the M200 for a pixhawk platform?

Don't know about ESC, but if they work with pixhawk, I think you wont be able to use DJI TB50/TB55.

Jeff Shoults Posted at 2018-5-28 07:29
Here's where I am at now: It appears that the pressure range on the barometer is simply not wide enough to operate at the depths we are running. Last week I tried to start the M200 deep underground and got a barometer error that prevented operation of the drone. Based on a quick barometer lesson it appears that the mems sensor barometer has a fairly specific range of values and higher air pressure underground exceeds the limits of the sensor, causing an error and it is unable to fly. This sucks bigtime!

This leads me back to my original issue of installing a Pixhawk into the M200. Does anyone know if the ESCs will even talk to the pixhawk? Or am I completely screwed and have to abandon the M200 for a pixhawk platform?

Looks like an awesome application Jeff Shoults!

Unfortunately I cannot give any advice regarding the M200 system but I do have experience implanting a Pixhawk 2.1 into a DJI M100.

The M100 used a N3 autopilot. I assume the M200 uses a similar N3 or A3 autopilot.
It should be easy enough to access the M200 motor inputs either from the servo connectors as they enter the PDB, or from the trace/wires that go into the autopilot. What I did with the M100 was to snip those connections and wire them to my Pixhawk2 (as shown here).


I have wanted to do with with a M200 / Wind 4 for a while but haven't had the money to fork over for it.


You can find a lot of pictures and some info here on the M200 platform.

I do not have this platform personally so I cannot speak to it specifically but this is what I think will work.


Take off the top of the M200 and get to the flight controller. You will be competely removing this flight control system once you determine the motor inputs.

The first way of determining this would be to simply look at what comes out of each arm. The ESCs should be in the arms/under the motors so you should see a power and ground wire (reading ~24v on a multimeter) going into the arms as well as three control wires (signal, 5v, and gnd) from the ESC as well as two wires for powering the LEDs.

Once you are able to determine the ESC signal wires we are in business. You will want to leave the power connections unchanged at this point and just cut and splice the ESC wires to servo connectors and test them with a servo tester for correct motor spin. Next you can continue and remove the rest of the DJI autopilot/GPS stuff.


For the power connection it looks like the M200 uses a basic wiring harness without any fancy PDB. This is good because it means we can connect a Pixhawk current sensor inline with the main power harness. You can try using the supplied Pixhawk current/voltage sensor or get an upgraded Mauch system for higher current requirements (recommended). Put that in line with the battery power supply and leave the rest of the intelligent battery stuff alone. Hopefully the batteries wont mind that they have no telemetry and will just work like the M100 battery did.  


You are basically done then. You can run your telemetry radio and RC system however you want. The pixhawk should be able to spin the motors and sense the battery voltage etc.

Just make sure the motor order and spin direction are correct. The M100 worked out fine with the Pixhawk2.


Let me know if you have any questions! I would really like to hear how this conversion goes and if you have any lessons learned. I want to do this conversion to benefit from the dual battery system (battery swap without powering down) as well as for the IP rating of the aircraft.

LOLS

I'm thinking about trying this as well.  Has anyone successfully integrated a Pixhawk FCU with the M200 ESCs?