Best FC for a new Hexacopter build.

P

Pauli 336

Member
Hi all,

New member here, hoping to soak up some advice and tips along the way.
I am about to start my first build in a few days time. All the parts have arrived bar a few connectors etc, and I should be good to go for the weekend.
I have spent over 2 months pouring over videos and webpages, making my own notes and studying closely every step I need to know.
I have decided to (hopefully) start with a reasonable setup, and want to include everything I can in the build.
To start off, I decided on a F550 Cabon Fibre Hexacopter. Mainly because I wanted to use this for FPC and Aerial video/photography via a GoPro Hero 3+ Black Edition.
So, Part I have assembled so far:

Carbon Fibre F550
6 x 920KV motors
6 x Mystery M30 SimonK ESC's
Ublox 7M GPS
Acro Nase32 Rv6 Controller
FrSky X9 Receiver
FrSky Taranis X9D Plus with case
Turnigy Multistar 4s Lipo's 5200MAH and 10000MAH.
Charger etc
Mass with Gimbal and 10a batt is 2.27kilo.

I ran the build carefully through Dronebuilder.io, and I am getting good (green) ratings for all measurements, with a hover time of 17 minutes, to mixed flight 10 minutes.

Most parts were purchased after extensive researching and also the need to get parts that were well proven and documented online for assistance. I am happy with the connection side of it, and the basics will work fine as I plan to use SBUS for convenience and 16 channels.
The problems appear when I start to add further hardware, namely GPS and a 3 axis Gimbal. This is due to the pins used and quantity of them required for a 6 motor configuration using SBUS.

Now, I am starting to see a few conflicts as I learn more and progress.
1. The Nasa 32 apparently only supports 2 axis gimbals, and I want a 3 axis.
Also, with the Nasa32 having 6 servo headers, i am going to be using all of them for the 6 motors so no room at the inn with this setup, even for a 2 axis gimbal.

2. I plan to use SBUS to gain all 16 channels. Pin 4 UART2o on the Nasa is for the Sbus pin from the Receiver. However, my GPS also needs Pin 3 and Pin 4 (on SBUS these are unavailable)to function and also allow connection via Cleanflight. I think I can move pins about but this will not allow GPS to work when connected to Cleanflight for diagnostics and checking.

Given that I want to use SBUS, GPS and a 3 Axis Gimbal, is there another FC out there that is SBUS compatible, will work with the FrSky X9 Receiver and will allow correct connection of the UBLOX 7M GPS and a 3 axis Gimbal to a 6 motor configuration?
 


T

Todd Pals

Member
The pixhawk will certainly be my suggestion if budget is no issue, however you don't NEED to have the flight controller control the gimbal. On my 3DR Y6 that has a Pixhawk it came set up where the flight controller does not control the gimbal, the gimbal software takes care of that.
 

S

Ssoap

Member
Have a look at the Vector from Eagle Tree. I have three and they work great. The OSD is very
nice.

Sudsy
 

OldGazer

OldGazer

Member
IMNSHO, Pixhawk is a much better choice, but a more pressing problem is I think you have chosen the wrong platform.

At 2.27 Kilos AUW (All Up Weight), the motors you have chosen and the prop sizes you are limited to by the frame size do not have enough power to give you a stable hover at 50% throttle which will have a highly negative impact on stability and flight time.

Form an aerial photography perspective, since you wish to use a 3 axis gimbal, you may want to consider adding retractable landing gear so you can pan 360 degrees and have a clear view. Of course you will incur an additional weight penalty that will tend to reduce flight times.

Putting this all together, I think you would see better flight performance with a 600mm+ hex with 740 to 750 Kv motors swinging 13" props on 4S.

As a case in point I have a Tarot 680 Pro with a Pixhawk, 740Kv motors, 1355 cf props, 2 5200mAh 10C LiPos in parallel, 2 axis gimbal, SJ4000 clone sport cam, M8N GPS, 915mHz telemetry radios, Tarot retracts, with an AUW of 3257 grams that gives me consistent 18 minute flight times.

I see you are using a Taranis. Excellent choice. I have an X9E. I use an X8R receiver. (FrSky does not make an X9.. ;-) ).

For telemetry, not only do I get full telemetry to my Ground Control Station (a laptop running Mission Planner), I also get full telemetry to my Taranis via a Craft and Theory interface.

As far as controlling the gimbal AND the landing gear, Pixhawk can do that. Not only does Pixhawk have outputs for up to 8 motors, Pixhawk also has 6 outputs for things like gimbals, servos, relays, and retracts.

You can read all about it here: http://ardupilot.org/copter/docs/introduction.html
 

P

Pauli 336

Member
Guys, thank you all for your replies, much appreciated. As I have all the above parts bought and assembled I have started with this build initially, but will focus first on getting the bird in the air before concentrating on the GPS and Gimbal. Old gazer, thats valuable points you have raised and I will certainly take on board. The nice thing is that with this interest, you don't need a second mortgage to swap out parts, so happy to tinker around until I get things right!
I went for well recommended parts, which is why I splashed out on the Taranis/X8 combo as they seem to be well recommended. Maybe you can help with some issues I am having, although I'll put these on another thread.

In the meantime, any other recommendations for FC for this build would be appreciated.
 


B

brunoau

Member
I certainly do not wish to put water on your fire, but before starting: do you have a pilot licence and a drone endorsement? If not, your maximum legally allowed MTOW (AUW) is 2000 grams. You better think about it.
 

P

Pauli 336

Member
Erm, I can read normal text thanks. :)

I'm in the UK, that rule does not apply here.

I can fly a UAV up to 20kg without a licence.

Drones are currently regulated by the UK's Civil Aviation Authority (CAA)and split up into three different weight categories: 20kg or less, 20kg to 150kg, and above 150kg. For the lightest tier, you don't need a certificate or permit to start flying your drone. The only exception is for "aerial work," which refers to any flights where you (the pilot) is paid.

If your drone weighs 20kg or more, you'll need a permit from the CAA first. However, most hobbyist consumer drones -- be they toys or serious enthusiast models -- are significantly lighter than this, so even if you're attaching heavy camera gear you're unlikely to hit this higher weight bracket. To give just a few examples:

DJI S1000+: 4.4kg

DJI Phantom 2 Vision+: 1.2kg

Parrot AR Drone 2.0: 0.4kg

Blade 350 QX2: 1kg
 
Last edited:

P

Pauli 336

Member
Ok, update so far.......Still not in the air!! :)
Ran into many first build issues and a huge amount of man hours studying up. I can say, I am 500% more tuned in to this now than I was just a few weeks ago, but it has taken a lot of study.
Firstly the build. No issues here. Ive a nice Carbon Hex frame with plenty of room for equipment. All nice and neat.
Eventually got the Nase32 connected to Cleanflight, calibrated the Mystery M30 ESCs and tested the motors. Immediately, one ESC smoked and died. Then another.
A third is causing concern due to stuttering and excessive heat, so I have removed this for safety reasons. The supplier is sending me replacements which should be here over the next day or so. I used the remaining 3 ESCs to check all 6 motors and they are all powerful and smooth right up to max revs via CF. I have had the Hex powered up quite a bit testing various stuff and the motors and three working ESCs are robust. In fact, this Hex would take off on two motors, so I am hopeful of a decent rate of lift when all is set up.

ESC issue aside, I had a heck of a time trying to get the receiver operating the various inputs. As it turned out, I was setting up SBUS on CF incorrectly. Once I got over that, I was able to get into the heart of the Taranis and setup Arm, Flight modes, and Failsafe. I have also preset switches for Retractable landing gear, buzzer, and Lights although these are not configured on CF yet as I am still to fit the parts.
Onto my experience with the Nasa32, and this has been a sad story. Fried the first one, no idea how but it may have been a misconnected wire.
Ordered a second, and fried it as well. I am awaiting a third. When they were working, I had control of Arm, rudder, throttle, ail etc and everything seems great via CF. Supporting the Hex with my hand on one side and powering up the 3 props I have fitted and it felt stable, even if I tried to knock it out, I could feel the stabilisation kicking in, so this bodes well.
When the ESCs and the replacement Naze32 arrive, hopefully that will be me quickly in the air. I will have hopefully 12 free channels (which I won't need all of) and can then get landing gear and lights attached.
I ordered and received a UBLOX NEO 7M GPS, and I could not get this to work at all.
As I connect via SBUS, I activated Soft Serial 2 on CF and used pins 7 and 8 for RX and TX, with power and ground from the servo rail. It all seemed straightforward. I Got a green (power?) light on the GPS, and a Green GPS light on the first page of CF. However, no blue light ever on the GPS unit, and it never picked up any Sats or got a lock, even after 1 hour outside. I reversed the RX and TX leads, but no difference. As I say, GPS is turned on in CF, and Softserial configured for GPS. I have tried several Baud rates too, but no response.
So I have given up on that for the moment.
Once I get at least airborne, I think I will invest in a Pixhawk, as it seems to allow for all the things I want to do. RTH is something I have thought about increasingly, and its becoming a priority for me. Even with GPS working on the Naze32, I don't think RTH can be relied on.
 

OldGazer

OldGazer

Member
I'll second the Pixhawk, but I've seen several posts on other forums about problems with the 7M GPS. I have had issues with a 6M. I have and LEA-6H on an APM quad and an M8N on a Pixhawk hex. Both work fine, but the M8N consistently gets more satellites that the 6H does.
 

P

Pauli 336

Member
Thanks. Yes, I think the Pixhawk may be the way to go with an M8 GPS. I am peeking through my fingers at Mission Planner, as I will now need to learn this, after just about getting used to Cleanflight, but the setup of the Pixhawk PX4 looks fairly easy. I like the option for 6 flight modes, although I think fitting a 6 position switch to the Taranis might be the best way to go with this. It seems GPS function is well integrated with the Pixhawk. What have your experiences been with RTL etc? I really want to have this as the thought of my Hexacopter flying off with a broken link or something fills me with dread.
 

P

Pauli 336

Member
Bit the bullet and ordered a Pixhawk PX4. I have enough spares to build a Quadcopter as well, and I will use the Naze32 on it.
So, I've been playing around with the Pixhawk since the weekend. I had another steep learning curve jumping to another stack so early in my experience and after just getting used to the Naze32 and Cleanflight., and I still have many unanswered questions.
I don't have a windows pc, although Ive ordered a Windows laptop just for the Hex. I have a Mac, and due to that had to download APM 2.0. From what I can see its like a light version of Mission Planner? Hence my decision to get a Windows laptop. Im fed up of reading solutions to problems, them being referred to Missionplanner, and I can't use it. So first question, arm I better off on APM or MP? I also have the option of loading Arducopter or Pix Firmware. Which is the best for this FC?
So, I did manage to get the Hex in the air after all the parts I needed were installed.
Took it into the garden and got it in the air, only to crash into the side of the house. I have had several minor crashes since, but hot glue is my friend! Only minor damage, have lost 2 props, and a broken GPS stalk.
More pressing as I test, is that stability is causing me concern. I can get to hover at about 37% -40% throttle. But on take off, the craft is lifting to the right and front. I have countered this by adjusting trim on the ELEV and AIL. Not ideal at the moment but it helps. However, I have noticed, on landing, disarming and rearming, I then have to readjust the trims, usually in the opposite direction again.
I have tried Autotrim once, but as a beginner, cannot get the Hex in the air quick enough for the Autotrim to work.
I have checked physically and the Hex is reasonably well balanced, the battery is low and centred and there is no gimbal attached yet.
The Pixhawk sits on a dampened perch. I can achieve reasonable results on the Stabalize mode, but I am disappointed overall with the stability. I thought and expected the Hex would sit relatively solid with only the throttle being used. I expected some drift, but not to have to constantly compensate with the Ail and Elev. Any advice?
 

P

Pauli 336

Member
Well, just to update from day one. After quite a few crashes and tumbles and a few strip downs, I have finally got the Hex flying the way I like it. I am quite pleased with the amount I have got done from the initial build.
Upgraded from Nase32 to a Pixhawk, installed an FPV camera and transmitter, upgraded the speaker added the 6 position switch to the Taranis for flight modes, upgraded to CF props. I have the C+T telemetry interface on order with Flightdeck (thanks to Oldgazer for that tip) and I have also ordered OSD kit and a telemetry TX/RX for Missionplanner. I have only tried FPV via goggles very quickly as I am still learning to fly, but looking forward to that. I have also ordered some backup bits to have for repairs, bottom board, arms etc.

I have to say, contrary to a post above, I am achieving very stable and easy hover at 50% throttle with fast and quick manoeuvrability. I haven't accurately timed flights yet, as I have been testing flight modes, landing and taking off several times in succession etc. But today, I kept it up for a good while and landed well before the lipo alarm went off, and I am sure I was flying maybe 17 or so minutes on a 4s 8000mah. The 3s 5200mah are giving me about 9-10 minutes I suspect, and the big 4s 10000mah lump should give me about 18-20 minutes. I wasn't thrashing it about today, but it seemed decent enough. I am still considering a motor change though.
That was the first really stable and enjoyable flight I have had. In learning, testing and tweaking, I couldn't really relax to enjoy flying. The Hex was not well balanced, it was twitchy in the sky and definitely hard to control. I have still to run Autotune due to the wind, but I did discover last night that one of the esc's was causing a motor to run slower than the others (it could be nearly heard on testing via the motor test on MP). I swapped out the ESC, checked over everything, recalibrate the ESC's, ran calibration on the Pixhawk, and got the GPS running well with 8 sats picked up. First take off after and it was solid and stable. The esc issue was likely having a large bearing on stability. I will run autotune when I get a calm day, but I am much more impressed now with how it is functioning. I hope now I can enjoy more flying and less problem solving and rebuilding after crashes. Having 'forgiving ' flight modes like Loiter now programmed in to my 6 pot switch has helped enormously.
So far I have broken 2 arms, legs (multiple times, and hot glues back on), snapped 2 GPS stalks, broke the cover on a Fat Shark antenna, broke at least 8 props, snapped the wire to one of the X8R's antennas, fried several Nase32's and 2 esc's early on due to learning the right way to build!, and finally trashing a new 5200mah 4s Lipo in just 2 days due to underestimating the importance of following the tried and tested handling procedures for charging and discharging. I had two large crashes, and I am still not fully sure what happened. Both involved complete loss of power at height. I 'think' I may have brought the throttle down and rested it at the bottom setting on both occasions after 'panicking', disarming the esc's, but I am not sure. I have checked everything else and thats the only thing I can think of, so hopefully it won't repeat itself. I am going to put quite a few hours on it at 10-20 feet, and within a short distance until I am sure all is ok.
But, its a steep learning curve, and I expected to make mistakes. At least its relatively inexpensive to get new parts in most cases, and I knew it was going to be impossible to get things right without many crashes. I can get on and enjoy things a bit more now before upgrading again! One thing that has stood out is the robust and reliable way the Pixhawk performs. It has taken a bashing, but has performed flawlessly. Missionplanner is excellent ( compared to my experience with the Nase32 and Cleanflight), but the Pixhawk fills me with confidence.
Definitely going to remain in the larger sized Video/ photography use UAV area, as photography is a passion. I doubt that I will be interested in the 250 class UAV's for racing etc. I like power and size and I am looking forward to capturing some iconic Irish landmarks over the coming months. Having said that, I have enough spares to build a 450 quadcopter, so might just do that for the heck of it.
 

C

censrd

Member
Wow... thats quite the list of items that have required replacing. Takes me back to my first hexa build, the main difference being that my research meant i followed the tried and true methods of battery managment. In fact i still have a healthy 10,000mah pack that was in my initial purchase of the 550 flame wheel some 5ish years ago. Not trying to brag or anything but making the point of really listening to those that have come before you and the advice they give, as failure to do so can become a very expensive endevour. The other thing that i wanted to say after reading your post is that it sounds like you've just dived straight into the deep end. Sorry if I'm wrong here.

Whilst i commend the willingness to join the drone community the one thing i suggest to anyone that asks me where to start is
1. Use a sim, there are a heap of realistic drone sims on the market now that make practicing easy as ****.
2. Then once you get the hang of that, the absolute best plan of progression is to get hold of something like a Blade nano QX. They are practically indestructable, i have no hesitation in lending mates with 0 flight experience the controls to this little quad, and the best part is they fly very comparably to bigger drones, I beleive better than any other small-mid size quad. But these still have auto leveling. To learn how to fly properly get a Blade 180. No auto level just having to learn the skills required to fly properly. That way wether your other quads have gps or not you can be confident in your ability to fly with or without the FC's brain being in absolute control. If GPS fails or some other issue pops up its essential to be able to fly full manual. I have met too many people that soley rely on GPS auto that its downright scary .

Someone else mentioned further up that the 550 is going to be on the small side for what you want to do, whilst i can attest that a well specced 550 can indeed do a decent job of flying a gimballed gopro around, after recently stepping up at a tarot 680 pro, with 1355 cf props it doesn't feel like its flying on the edge of its limits like the 550 was with the same gear on it. I have been using a NAZA v2 but am super keen to swap over to a pixhawk when the funding allows for it. Last point is especially until you are used to the drone you better off using the plastic props over the CF, I found on my 550 the cf props made for a very twitchy and harder to fly drone. Not to mention the cost of replacement after a crash. Good Luck you you :)
 

P

Pauli 336

Member
Thanks for the comments censrd. You are right, I did lose quite a few parts along the way. Not expensive though. The motor and escs were faulty and replaced by the dealer. At least one of the Nase32 Fc's was a dud as well. I probably spent about £80 replacing the other bits and pieces.
Far from rushing in to things, I never turned a screw until more than 2 months or studying, reading up and digesting drone builds. I then decided on the spec I wanted, and addressed a few weaknesses early on. The crashes were a result of learning to fly, and I would be surprised if many people didn't have the same experiences. I have 3 flight simulators, and used them all with my Taranis prior to taking to the sky. My initial flights were very much by the pants because I had no GPS for the first few weeks, so it was all done in Acro and Stabilize mode.
I also have a Hubsan which I used for training. Some minor crashes were due to instability during tuning, some due to hardware issues, but most of the damage came from 2 complete power cuts at about 15-20mtrs up. There are numerous reports of similar experiences online, and I am concerned about this. One may have been as a result of me disarming mid flight, although I'm not 100% sure about this. The second time I was changing between flight modes and power cut totally. My latest theory is that it was connected with the throttle PWM value for failsafe, so I have disabled failsafe until I get my head around it. I have suffered several flips on landing after perfect 15 minute flights in Loiter or RTL mode. This is a widely reported issue on APM but was supposed to have been sorted by v3.2x. I am experiencing it on v3.3.x
Basically, the FC in Loiter and RTL mode is meant to sense the landing and cut the motors. It is not always doing that. Subsequently, the craft will not disarm, stats to speed up, and starts to roll, normally flipping. Even after flipped it continues for up to 5 seconds, battering the blades etc to bits. This is a known issue. The fix seems to be when landing switch to stabilise mode.
I am mainly flying under 20mtr height and well within view in the meantime, as I am still fine tuning things.

Its interesting what OldGazer said above re my motor selection on a 550 frame for a gimbal and GoPro setup. It got me thinking and I checked out as much as I could online. Certainly 2212 920kv motors seem to be widespread in 550 hex and 450 quad frames, and these are the staple sizes for GoPro and Gimbal setups.
I ran my current 550 spec through Ecalc, and I was getting no warnings, and a thrust ratio of 4.3:1 which is more than acceptable. I am also getting flight times around 15- 20mins, and very stable hover at 50% throttle. I can't use bigger blades than 10" so the motor almost picks itself for this size, although I could go sub 800kv, but end up with a TR of 2.5:1 which I feel is too low. On Acro, I could throw this biggish Hex around like a 250 racer.
I did take on board some of the suggestions above. I felt the Nase32 in Rev5 and Rev 6 guise is not robust. Easily damaged and I expect would not last long. I changed to the Pixhawk PX4 and it was a great move. Missionplanner is excellent, and the PX is robust and flexible. I have GPS, TX FPV and OSD plugged in to it and it just works as it should. There are plenty of inputs for servos, landing gear Gimbal etc, and huge expandability.

Interesting what you said about Carbon blades. I have plenty of plastic blades and they are fine if not terribly strong. I thought I would try CF blades, and they have held up better to a few knocks. I have noticed a subtle difference in flight characteristics. Its almost like a sharpening up of everything. Almost like driving a car with soft tyres against driving a car with pumped up tyres. There is a very distinct noise from these blades, but really the cost is not prohibitive. I will switch back and forth until I find my happy medium but at the moment I am more comfortable with the CF blades.

On batteries, I have a range of Turnigy 4s and 3s blocks. 5200mah, 8000mah and 10000 mah. I want to try a few different sizes again to see what gives the best experience.
I am careful hadling these batteries. Apart from the obvious cost, almost £200 for the above, I am cautious for safety reasons. I have a legit balanced charger and always balance charge.
What happened with the one I lost was simply an oversight. I had it connected to my Hex on the kitchen table for several hours as I was working on it and tuning via the laptop and MP. I didn't realise that the power had depleted to a very low level and I could not revive oner of the cells, even with the usual tricks. It began swelling then, and I disposed of it. All the joys of learning a new activity!

I agree totally about the step up to the next size, and I have been looking at the Tarot 680. I will wait a few months and possible start that build in the spring.
 
Last edited:

You must log in or register to reply here.
Top