Pages:
Actions
  • #1 by t0m541 on 04 Nov 2017
  • Been thinking about getting a quadrotor for a couple of years, recently decided to take the plunge and have bought a RTF, though it's not arrived yet.
    But looking at all the info and being a mechanical engineer by trade, I'm thinking of building my own from scratch and will be asking advice from the many folks with lots of knowhow on here.
    The idea I have is for a multirotor, with a gimbal camera mount with gps and capable of doing stuff like follow mode, height hold etc,
    Carbon fibre construction with decent flying time, not looking to build a racing one just very stable.
  • #2 by Icefever on 05 Nov 2017
  • and welcome, from one newbie to another.
    Your in the best place matey, I've just started building my own, and the guys on here are great. Ask a Q...and you'll get all the help you want, none of this RTFM sh*t like on some sites.

    From what you say I'm in the same class at the moment, flying for fun,  more on the photo side, but who knows I may build a racer just for the hell of it. ;)

    As an engineer by trade, you'll fit in nicely, I'll be asking you Q's... :whistling: you never know.
  • #3 by Two-Six on 07 Nov 2017
  • Hi Tom541  :welcome:
    Like IceFever says posting here is a good first step to multirotor success.  Which RTF did you buy?

    Very clever automatic multirotors with camera gimbals aren't my thing, I have a "250 racer" built for thrills and spills, so I am not a big expert in photography/video platforms but here are some general comments:

    Building your own one that will do "follow me" might be impossible or very hard to do.  GPS assisted flight, altitude hold, limiting the distance and height away from the take off location, position hold and return to home, loiter around a GPS location and "headless mode" are all quite do-able.   Way-point following I think is harder to implement.

    Look to carrying a small camera like a go-pro or a Runcam, if you want to carry a big DSLR or something you will need a beefy vehicle to lift it.  This can get pricey.  You will want at least a hex-a-copter (6 motors) for redundancy.  A quad-copter can't fly with a motor out.

    Implementing a useful on-screen-display (OSD) via an FPV set-up isn't too hard either.

    From the answers given to other people like you who are interested in the nice pictures/easy to fly aspect of multi-rotors probably the best advice is to buy a DJI product and forget about building your own. The phantom 3's are going at a very good price I think these days.  The phantom 4, the Mavic and the Inspire are very clever bits of kit but they do cost a pretty penny.

    They never go wrong or crash so not to worry about loosing them......  ;)

    Have a look at this thread, especially Orfordness's posting.  http://www.multi-rotor.co.uk/index.php?topic=22306.0

    Cheers!  :beer2:
  • #4 by shawdreamer on 07 Nov 2017
  • "follow me" was a ballache once to get right but its pretty much had the wrinkles smoothed out on the FC's that are capable of it.

    Even the silly affordable APM can pull it off pretty decent now and its more expensive but more capable counterpart the Pixhawk pretty much has it nailed down now, iirc its set up to receive commands from an outside source (such as a laptop or half decent windows enabled tablet) that send its own gps fix to the FC which then using its own gps fix as reference manoeuvres itself to the target gps fix.

    It basically works along the same lines as sending the fc a waypoint target and it flying there but instead the target sends the waypoint target every 2 seconds or so and the fc adjusts its heading to match everytime it receives the data.

    Even the lesser hightech Ardupilot (APM) could pull off waypoint missions quite well (I experimented with it a few times with my Big H-copter frame and it performed very well) and the follow me setup is basically just building on that just instead of going "here's a mission plan, off ya go" then waiting ten minutes for it to finish the plan and return its now more like "heres a mission plan off ya go.....<2seconds later> "sorry change of plan"........<another 2 seconds>........ "sorry again, change of plan"......<2seconds later>....... "I know Im probably pissin you off by now but change of plan..... etc etc (well maybe not that last bit but you get the gist)

    The pixhawk with its superior processing can pull it off even better.

    OP, If your looking to build a camera platform over a smaller more speed orientated build then Id suggest either of those two mentioned FC's as they are well established, their firmware and software has been well polished and is massively supported.

    On a budget the APM 2.6-2.8 will give you pretty much everything your after at a very cheap price (think Ive seen em going for £22 lately)

    but If at all possible opt for the Pixhawk, though their capabilities seem the same on first approach, as you progress in the hobby you'll inevitably want more from your FC and the pixhawk is as said alot more capable, just a little more pricey (around the £70 mark cheapest iirc)
  • #5 by hoverfly on 07 Nov 2017
  • As above , I'm still running a folding 500 frame with  a A.P.M.  f/c  as  backup camera ship.   In comparison with the  D.J.I.   A/C  it isn't  sophisticated  but it does the job.  The Pixhawk is a vast improvement  apart from the DF13 connectors which will self destruct  if you look at them to hard.
    For a lightweight camera the Runcam takes a lot of beating , the stills of the video are quite good.  :smiley:
  • #6 by t0m541 on 07 Nov 2017
  • In answer to a couple of questions.
    The RTF I bought is the Visuo XS809HW, according to what I read it's a Mavic clone in looks, but I only paid £35 with 3 extra batteries and free delivery, so not breaking the bank and has decent reviews.
    As for my build, I was thinking about the action cam, it seems robust and has good reviews, I also want FPV on the copter too, so possibly 2 cameras ?, from what I understand.
    As for a FC I was looking at the DJi Naza M lite or the Shark X6 with gps.
    Motorised gimbal.
    I am reckoning on around 2.5kg as a ballpark figure on copter weight and running 6 motors, that should give me lots of lift but probably at the cost of battery and flight time, though I'm sure you guys will be on hand to help me make an informed decision on motor/propellor combination.
    Batteries, 8000mah S3.
    I'm not going to try to build it in a week, so to speak, some changes and configuration will alter as the build progresses.
    The frame is going to be carbon, 3mm for the deck and 8mm box section for the motor supports, that should give me the rigidity for semi decent images and cut down on vibration and weight.
  • #7 by Two-Six on 09 Nov 2017
  • Build it in a week :frantic:
    So are you going to build the frame yourself? 

    Just a thought, you want to use 8mm box section to support your motors (and everything else that adds up to 2.5Kg)?  Really? 

    I don't think that is going to be very rigid or strong.  Could be wrong about that but it doesn't seem very substantial to me.  It might be tricky to get rigid and durable arm-to-frame mounting as well, are you going to make aluminum bits to hold it all together?.  It will be interesting to see how you are going to do this.  How are the motor-to-arm mounts going to work?

    It would be best to get a separate FPV camera that uses 5.8 Ghz video transmitter obviously with a 25mW power output (cough).  No higher (ahem).

    That Visuo XS809HW looks OK for the money.  How are you getting on with it?
  • #8 by t0m541 on 09 Nov 2017
  • The Visuo hasn't arrived yet, I thought it would have done by now but it took 5 days from dispatched to en-route according to the tracking info, and it's been en-route for 2 days, so not exactly Fed-Ex, but hey-ho, free shipping and all that.

    So you reckon 8mm carbon fiber box section might be a bit undersize for the weight ?
    The reason I was going for that size is that I have a round hollow fibre-glass tent pole that's 8mm diameter, though slightly thicker wall section, and 500mm long, I hung a 2.5kg weight off the end of it and it had no bend at all.
    So I was thinking that even doubling up the copter weight, divided by the number of supports, is only 834g at a lot shorter length.
    The web site selling the carbon has a data sheet for the box section, and if I'm reading it correct and doing the right conversion, it has a shear strength of 356kg, though someone with more experience of carbon fibre might tell me different.
    I can bond the alloy motor mounts to the carbon, likewise bond the arms to the main body.
    Making an "A" frame spar from the bottom plate of the main body, I'm planning on having 2 decks 125mm wide x 225mm long 3mm carbon spaced 50mm apart, that should be ample space for fitting all the electronics onto/into and under.
    210mm arms should allow for decent size propellers.
  • #9 by Two-Six on 09 Nov 2017
  • Well, I admire your enthusiasm to have a go at building a frame yourself.  It sounds like it might work.  How are you going to bond things on? If the motor is bonded to the arms will it be easy enough to unbond it if it doesn't work?  How do you mean an A frame spar?

    I hate to dampen your enthusiasm with my rampant cynicism but why not buy a real hex frame like this one:

    https://www.unmannedtechshop.co.uk/s550-hexacopter-frame/

    There might well be better sexier frames available, hopefully somebody else will post some links for you.

    Its not expensive and unless you have lots of CF laying about, it isn't cheap stuff, its about £12 a meter isn't it, not including postage.  That could quickly add up to the cost of a really nice Hex frame.
  • #10 by Two-Six on 09 Nov 2017
  • How about this kind of thing?
    https://tinyurl.com/ybx4kt7x
  • #11 by t0m541 on 09 Nov 2017
  • The second link is very reasonably priced, £47 + free postage and around the size I'm thinking too.
    The CF for my build is around the £70 mark.
    The only downside with that frame is it's limited frame space for the electronics, not a lot of mid-deck space, it looks like there is around 25mm between the layers, though this would be enough to squeeze a FC and ESC's into.
    The centre section is roughly 230mm going by their dimensions.

    Got me thinking now... :hmm:
  • #12 by Two-Six on 09 Nov 2017
  • "Got me thinking now... :hmm: "

    Good :-)

    I am not saying those frames are the best or anything.  I am sure there are better options available.  Hopefully somebody who knows about these things will link a few more for you to have a look at.
  • #13 by t0m541 on 10 Nov 2017
  • The other concern regarding the frame in the link is it's weight capacity, the spars aren't CF reading the details, again this leads to concerns with upgrading and changing configurations later on.
    I think I'll go with plan "A", buy commercial grade CF and self build and over-engineer the thing  :o
    With regards to motors and props, from the reading I have done, the lower the Kv the higher torque and lift capacity using bigger slower props, the higher the Kv the lower torque but smaller faster props, (Chinook vs Little Bird).
    Is that right, and if so then assuming 2.5kg copter and using the 2:1 ration for lift and sustained mid power hover, what motors/propeller combinations would be a good choice, I realise it's going to be pricey and there will be six sharing the load.
    If I get the motors decided on then work the link back to the FC leaving some spare capacity so thing aren't struggling.
  • #14 by Two-Six on 10 Nov 2017
  • If you set about building your own frame then you can have what-ever prop size you want.  However there will be "THE BEST" size of prop, in terms of efficiency running on a 3 cell battery.  I guess the size of your props should be dictated by how easy it is to get good but affordable props with the most sensible fixing method as close to the most efficient size you can find.  Nylock nuts being the preferred fasteners for props.  So when you have decided on which prop to use then pick the motor.

    Another thing, there will be a limit to the current your battery can supply, your battery will most likely be 30c continuous-40c burst (in theory) errm that's a maximum of 5 amps per motor.  That's quite low and it will be more like 3.5 or 4 amps per motor.

    Sorry but I can't be more help on this area.

    Perhaps try this: https://www.ecalc.ch/xcoptercalc.php

    You might be better off with a higher cell count on your battery.  I dunno  :shrug:
  • #15 by t0m541 on 10 Nov 2017
  • Ahhh, the battery.
    This is going to be the heaviest component, and as rightly pointed out the cell count determines the overall performance.
    The more cells the more current available but the voltage increases, (3.7/7.4/11.1/14.8) etc.
    Theoretically it should be possible to run as many cells as you want, (weight permitting), but regulate down to the voltage needed for the FC/ESC/motor combo, building the regulating circuit into the power distribution board.
    IE:- 2x8000mah batteries in parallel at 14.8v, giving 16000mah, regulated down to, say 7.4v, the output of the batteries would then be x1.5 equating to 24000mah for the same weight. There would be some capacity loss due to heat through the regulator hence not X2 or 32000mah.
    These figures aren't exact by any means but I'm just putting the idea out there for critique.
    As for the props, the research I've done seems to point to either CF of wooden props being the best as far as balance and minimal vibration, 15inch props are possibles for the build.
  • #16 by Two-Six on 11 Nov 2017
  • The thing about battery capacity is that there is an optimal size because as your capacity increases so does the weight and the power needed to lift the batteries.  So you could have massive batteries and still only have a 20 minute flight time.  You need to be realistic about flight time.   

    As for cell count, I think that as the voltage increases the current needed for a given power requirement drops and this is more efficient.   

    The DJI products actually do a stunning job of stretching flight times to about 20 minutes.  They can do this because they are super efficient, amazingly well optimised.  I find it amazing that they can do that, all of my RC aircraft can only manage 6 or 8 minutes . 

    Do you really want to fly for more than 10-15 minutes anyway?  I get bored and loose focus after about that long..Perhaps that just me.

    With props, they all need balancing, I still haven't come across a balanced one.

    Regulators are easy you have a big choice of them, no need to build one.

    I do feel like you are re-inventing the wheel a bit with this project.  Sorry to say ;-)
  • #17 by t0m541 on 11 Nov 2017
  • Re inventing the wheel...it's why I'm here, to find out and get advice as to what's already out there that I can "plug and play" and what I'll have to build.
    Ideally, with the good advice and help from here, I'll build the frame and the electronics and hardware will be what folk with the flight and build knowledge point me toward off the internet shopping shelf.
    When you say the propellers need balancing, I'm assuming across the blade and not motor to motor.
    In your experience, which have been closest on balance and needed the least amount of truing up as in material and manufacturer ?
    For a hexacopter looking at 5kg lift what would be the best motor/prop combo, the choice is quite confusing for the newb.
  • #18 by Two-Six on 11 Nov 2017
  • You will need a prop balancer and you balance the actual prop.  Magnetic ones are the best.  You can't really balance a motor but they should be balanced already.  Cheaper motors might be more out of balance than posher ones. It will be easier to balance a plastic prop as you sand off material on the back side of the blade to make it balance.    It would not be a good idea to take sand paper to a laminated prop, you should use tape instead.

    The choice of motor and prop is beyond me, sorry I can't help with that one.  :shrug:

    What do you want to do with your drone anyway?
  • #19 by Icefever on 12 Nov 2017
  • As a newbie myself and halfway through my first build  I've been following this thread,  and knowing Two-Six the info that he's posted will be 100%.
    Having said that, if you still want to run with plan A this is down to you,  but as I  said I'm a newbie and I wouldn't want to spend a load of money to find I'd built a drone that had problems that I'd made through my inexperience.
    Just my thoughts matey on it,  hope it all goes well and will follow you're build..
  • #20 by Icefever on 12 Nov 2017
  • Don't know what happened  :thumbdown:  it came up as a double post.  ::)
  • #21 by Two-Six on 12 Nov 2017
  • Thanks for you vote of confidence in me IceFever!  I do try my best not to imply I know about stuff I have no clue about and try to stick with subjects I feel I can actually add value too with my two credits worth.
  • #22 by t0m541 on 12 Nov 2017
  • Use of drone...
    Well apart from this small neutron star that needs placing to open a portal to a parallel dimension..
    It's going to be used to explore things like, inaccessible anomalies in high cliffs, which is why I want a stable, fairly well weighted copter with a good bit of overhead power that will stand buffeting from changing wind and thermals.
    This is why I want to have the build space to include the avionics for it to almost self fly should my reaction on the controls not be quick enough.
    And when not exploring, but just using it for flying and taking "selfies" again I want it to be able to suffer the changeable conditions that we have here in the UK without chucking itself into a tree.
    Today for example we have 40-50 wind gusts, yet the weather is lovely and sunny.
  • #23 by t0m541 on 12 Nov 2017
  • After lots of reading and comparing, I've chosen motors, props and esc, also the cf for the frame.
    Motors Tarot 4006/620kv
    Props Tarot 680 pro 1355 cf
    ESC Hobbywing platinum 30a pro.

    From what I read, this combination should give 1250g lift each with around 15min hover time with a hex of similar weight to what I'm looking at on a 14.8 5000mah 4s battery.
Pages:
Actions