Pixhawk Vs Mikrokopter

[anaka]

Hi people
recently i'm considering to start a new build but depending by some feature that i need to implement and that are missing in the Mikrikopter world i have to look around to find a system that gives me what i need

at the moment the Pixhawk looks very promising but i never had one of these FC in my hands so here is my question:

Is the Pixhawk evolved enough to install it and use it quickly or is it still a development board? so a system that needs a lot of hours of struggling to make it work smooth and safe?

i need to build an heavy lift drone around 10 Kg at take-off

any answer and advice is very wellcome

 

[Av8Chuck]

You need to keep in mind that Pixhawk is both open source hardware and software so it's always in development.

Also, 3DR was the primary hardware manufacturer, once they left, or did what every it is they did, there really isn't a reliable hardware supplier. Although the clones will run PX4 or Ardupilot they proved to be unreliable.

Jordy, the cofounder of 3DR is getting up and going and starting to manufacture the original Pixhawk but that's a four year old controller. Everyone claims that there's always a new haardware controller just around the corner but none have really materialized.

Having said that it's really hard to beat Ardupilot. We use the X2 with Ardupilot for out heavy lifters and they have been incredibly reliable and accurate.

 

[Shaun Stanton]

I am not sure what features you are looking for. I can say as an adopter of MK FC's in the past and using Pixhawk FC's. I can say one FC that puts them all to shame is the A3. I had nothing but headaches with mikrokopter. Pixhawk 2 still uses open source Arupilot which works then becomes an annoyance with a firmware change. PH2 has not bee as impressive as I hoped for with a multirotor. Its great for fixed wing but still fall short IMO with multi's. I swapped to WKM after the version 3 6S MK board did not work as well as I wanted it to. Then I tried A2 which I thought was not really that good as it was advertised. When I went to the 3rd gen DJI FC, the A3, it has been everything I could have wanted in an FC. It is rock solid, very smooth control, holds altitude perfectly. It blows evey FC on the market including the Freefly Synapse on the Alta series IMO. Everything I could ***** about MK, WKM, APM, PH and A2. The A3 has been the answer. Coupled with LB2 the integration has been seamless on my rigs that use it on.

 

[Av8Chuck]

Depends on how you use them.

When it comes to autonomous missions Mission Planner is years ahead of DJI. I've been using the X2, which is a commercial replacement for the Pixhawk, for almost three years and have not had to upgrade firmware once. The problem with DJI is that none of their accessories work on other drones and you can't use none DJI accessories on DJI drones. Not only that you can't even use DJI accessories from one model to the other. You can't use Phantom P3 batteries in a P4 or Inspire batteries on an M600 etc...

We've been doing a lot of industrial inspection, mostly photogrammetry and recently a competitor showed up with an M600 flying a MX Ronin and GH4. It was blowing 25 knots gusting 30+ knots we flew ten missions before the M600 could get off the ground. Its not that it couldn't lift it, the gimbal would screw up, the controls for the camera wouldn't work or whatever. They didn't fly a single flight before we completed our mission, packed up and left.

We travel way too far, to locations in the middle of nowhere with no cell or internet and often there's a potential for a GPS denied environment. No way would I trust a DJI to accomplish this. Ardupilot has been great for us.

 

[Shaun Stanton]

I will give them that the mission planner for APM is pretty good, even back in the dark ages of APM 2.5's and 6's. will concede I have never tried the X2. I have never really liked the flight performance of the APM variants I have used in the past. I always felt like they flew dirty on my rigs even when tuned to the nats a$$. That's interesting that your competitor could not get an M600 off the ground. I have had no trouble in the wind with it. It has been rock solid for me. It sounds like your competitor is another story of "noob" with money and bought a machine that is over his head in operating. I can say the 600 has a few quirks like it is not very stable on landing with light payloads and the system sometimes does not recognize it has landed I think especially if you try to squeak it on the ground. It seems to want to keep the motors spooled to flight power for a few seconds before finally killing the motors. Of course this behavior I have noticed on all DJI's product lines after the WKM. I have one bird still using WKM which has never been a headache other than some flight characteristics sort of suck like it cant hold altitude very well, but it is manageable. My only other gripe with DJI is the lack of manual throttle in all modes. DJI does not seem to trust that certain pilots are skilled enough to fly without having permanent altitude hold on.

I suppose whatever works for the end user. I know I would not want to use anything other than an A3 on a set now, since its performance seems to have solved everything I would have bitched about with the other FC's through the years.​

 

[anaka]

Hi everybody and thakns for your answers

in my case i feel the limit with Mikrokopter when i can only use their ESCs

i ever had custom hardware, it would be nice to use KDE motors and ESCs together and with MK you can't

i'd like to have a tablet app that let's me setup all the parameters of the drone and this is impossible with MK as well

i'd like to have the alexmos board that works with the FC and have a follow me that woks great and drives drone and camera gimbal, similar to DJi X cameras

it looks like the MK guys don't want to implement these cool features in a near time

these are the main reasons that are pushing me to find an alternative

 

[Av8Chuck]

This is more of an issue of methodology than which is better. Both DJI and MK are vertically integrated, they don't want the end user deciding which components or accessories they can use with "their' drones. They certainly have a right to do business that way, but as a professional I need to know that the drone will work exactly the same way when I take it out of the case as it did when I put it in there.

I need to be able to decide how I will plan my missions and the type of payload that I want to integrate which requires an open and extensible architecture, which consumer drone manufacturers have decided for whatever reasons not to allow that.

There are other developers and manufacturers entering the commercial market, Intel/Ascending Technologies, MicroDrone, us, and others who understand how to empower professionals to transform the commercial drone industry.

 

[Shaun Stanton]

Hi everybody and thakns for your answers

in my case i feel the limit with Mikrokopter when i can only use their ESCs

i ever had custom hardware, it would be nice to use KDE motors and ESCs together and with MK you can't

i'd like to have a tablet app that let's me setup all the parameters of the drone and this is impossible with MK as well

i'd like to have the alexmos board that works with the FC and have a follow me that woks great and drives drone and camera gimbal, similar to DJi X cameras

it looks like the MK guys don't want to implement these cool features in a near time

these are the main reasons that are pushing me to find an alternative

You can implement other ESC's with MK if you get a I2C to PWM converter. The problem I have with it, is that it is not a perfect solution. It takes more time to tweak it and make it stable. Even when you that, you now have a huge single point of failure on your ESC's from a $40 dollar dongle. I know some people who flew with it for year without issue, I tried it and it was successful, but I did not feel it was worth the risk. I ended up dumping MK for a WKM on that rig. Funny thing is that there were a lot of aspects of MK that I liked over the DJI, like manual throttle and more gain parameters to tweak. IMO I think MK is not going to last that much longer. I think their downfall is that they approached this with a hobbyist mindset, like having the BL PDB requiring the end user to solder all of the motor terminals. I think they went overboard trying to over engineer the FC early on with what sounded like a bunch of impressive features that never really worked very well, while also having not so awesome flight performance. You can see thousands of threads on the Freefly forum with Cinestar owners and frustrations trying to solve stupid little quirks of the MK, that even the MK gurus could not figure out sometimes. I think they cant implement new stuff with their systems because they are so far behind the curve now in all of this.

Integrating the camera and gimbal with the FC is a big challenge that only DJI seems to have the closest ability at the moment with the exception of Freefly's Alta series. Even there stuff is not seamless unless you are flying a RED. Since I have been doing this, for about 7 years,I have tried so many different systems that promised to do that type of integration but would never really work 100% some stuff worked whole other stuff failed. The only camera company other than DJI that designed a camera to integrate with drones was Black Magic, unfortunately their camera is only 1080 which was good 3 years ago, but now everyone wants 4K so that made the BM drone cam only good for small budget independent projects. But with the cost of an Inspire 1 Pro now, its not even worth it.

I imagine if you have the time and coding chops, sure you can probably integrate a bunch of systems to work in the opensource realm. But if you need things now, which is me. I just need a system that is reliable and can get shot. The producer at the end of the day does not care about how cool the tech is. He just wants the shot in the best format possible. Sometimes that's a RED or Alexa Mini. Most of the time an X5R or X5S is what is needed. It will be interesting to see what Intel does, but for now I am staying with the closed systems.

 

[anaka]

Yes i know about the i2c to PWM but i don't trust this kind of solution especially when i'm flying...i want to minize risks

i follow by enough time the other developers than MK and have to say that the only one that got the legacy of MK and made the drones ready to be driven by everyone without any particular skill level is Dji...this means also that a lot of potentially dangerous people joined this field and a lot of regulation arrived as well....but this is another chapter...

i work with BlackMagic cameras, i need to have enough color depht for compositing, so i can't use other cameras with more compressed video data, i need to test the X cameras from Dji but what i read on paper doesn't convince me to change my mind with custom drones and cameras configurations...i like to feel the freedom manage all the aspects i want, if something happens on the SET in front of the clients i need to be able to fix the things quickly

one more thing i'm trying to improve is the camera control, to train a camera operator is not that easy so i need a more intuitive way to control the camera, i spent some good money on that but also in this case i still can't get the desired results....i bought the ACR Link that is similar to FreeFly Mimic but there are some limitations that make this solution not complete for drones...especially if you have to fly more distant than 30 meters!!!! awful...

actually i have Mikrokopter Alexmos and ACR and all these guys work well but not completely, i like the performances of mikrokopter fot the flight, i never had a problem or a bad surprise but i feel the ESC limit, all the rest is great for me
i need to develop a better camera remote solution and ACR with Alexmos is not production ready IMO
i dislike the APPLE style of Dji in general
so i was attracted by pixhawk that looks the closes one to what i need
that's my thinking in the last months

The goal is to lift up a RED in a safe way and with a good control for mid range distances...150 meters from the pilot is enough for me now

For me to spend some time to tune up the elements is ok but this doesn't mean i like frustration...

I'd like to use proximity sensors as well and they are available for pixhawk but i don't know how easy they are to implement

so is Pixhawk the good candidate?

 

[anaka]

Av8Chuck your model looks really great!!

 

[Shaun Stanton]

View attachment 28802

You need to keep in mind that Pixhawk is both open source hardware and software so it's always in development.

Also, 3DR was the primary hardware manufacturer, once they left, or did what every it is they did, there really isn't a reliable hardware supplier. Although the clones will run PX4 or Ardupilot they proved to be unreliable.

Jordy, the cofounder of 3DR is getting up and going and starting to manufacture the original Pixhawk but that's a four year old controller. Everyone claims that there's always a new haardware controller just around the corner but none have really materialized.

Having said that it's really hard to beat Ardupilot. We use the X2 with Ardupilot for out heavy lifters and they have been incredibly reliable and accurate.

What is that model you are using? Is that something you had fabbed yourself? The motor config looks real similar to the Shotover I see the shafts are a little different. Just curious.

 

[Av8Chuck]

Yes we manufacture our own drones. Our booms are removable and that motor configuration allows us to us it on an amphibious drone, a VTOL and fixed wing.