Let’s design a multirotor together.

[woofy 12]

@SanderG

That was my first thought and I looked at a screwed together shell, this is the original arm:

This has to be designed to be 3D printed, so the narrow end studs have additional support. We need 28 screws and studs which adds up to a fair bit of extra weight. According to Cura the cost of the glued design is about GBP9.78 using Faberdashery clear (uses 410g including brim and all support) . I think that's pretty good against a ready made shell such as this one which costs $119 and is also effectively a one piece design.

Also I think the design is likely to be fairly fluid. If it does get damaged there are likely to be additions or changes I want to make anyway, so I'd print an updated shell.

What do people think, glued or screwed?

@Conz

Yes, that's exactly what I want to do, have a waterproof quad that can land on water and dip a camera underwater. I don't want to fly in the rain. Ultimately I'd like to be able to move the quad around gently without using the flight prop's.

Biggest problem is that I don't own a quad and have never flown one, so I'm coming at this very raw indeed.

 

[woofy 12]

@Jameshs

The plan at the moment is to put the ESC's into the arms, hence the opening to the main body. Hopefully the design will not need high power ESC's that require heat sinking. I was planning to use 20 amp ESC's that have plastic heat shrink around them.

 

[woofy 12]

Just another thought, do we really have to pick one design over the other? If there is a desire for both of Erics ideas, why not do both. Maybe we can include foldable arms and wheels. What do you think?

@Sander

I like the idea to use Github to store the files. I've never used it and its about time I learned to do so.

 

[mrwong 5]

@Woofy, i like your design that you have made there. What program are you using? Can you get an estimated weight from that model?

I like the idea of a water proof quad. But again for your first quad build i dont think that is a good idea. Your going to be spending some money on decent electronics and testing around water just sounds like disaster waiting to happen. Motors can get wet but still have bare exposed connections and bearing life after getting wet is my concern. Depending on the power need to fly this sealed craft, you will need good cooling for your ESC's (exposed to some moving air) and your battery must also be stored water tight. Your design currently does not accommodate the power packs which will not be that small!

I love the enthusiasm! But i think we need to establish the components we want to used, flight time we want to achieve, dims and specs of all components and then build a frame to house all of it. Then make modular designs that can be water tight, fpv, or stunt.

let me know what you guys think!

 

[jameshs 42]

I am all for someone helping to define a kit of parts (of which I have no knowledge at all), which I think has to be the FIRST port of call - then from a common set of parts we can have a family of frame designs - and it would be great if things were then interchangeable with a few clever clicks - waterproofable with an add on case, wheels on the side (subframe) different leg designs if a gopro/gimbol is fitted - but if the 'core' component(s) are kept to one or two versions then we could start cracking

James

 

[woofy 12]

MrWong,

You are absolutely correct about the motors. I could not see a way to waterproof them, so I just thought I'd use cheapish motors. The motor wires will pass through a seal on the way to the ESC's.

I use Geomagic Design for the CAD. This morning I generated STL's to feed into Cura and got a shell weight of 410g, including brim and support.

 

[mrwong 5]

Those motors are actually not that bad! Check this all in one ESC that might be enough to power this model. The frame weight is not bad either! Let me see if i can find what those motors produce for lift with the 9x6 props. (Those motors look like they will accept the dji blades)

http://www.hobbyking.com/hobbyking/store/__42715__Q_Brain_4_x_25A_Brushless_Quadcopter_ESC_2_4S_3A_SBEC.html

 

[gr5 2,172]

So -- coincidentally I recently started getting much more serious about ordering parts and planning a quadcopter.

 

Here is a first draft parts list based on this quadcopter from Make magazine. My requirements are APM (easier to fly) and gopro compatible with brushless gimbals for the gopro. The total cost for the basic parts is $560 $395 including everything except the frame and some screws and the gopro (already have it) and probably some wiring (but not including shipping of all the parts). I think this list is complete. I am going to go over this list again and order these parts probably Thursday on my train ride to Makerfaire. Another requirement is that I build it myself so that when I crash it (I'm experienced which means I know I will be crashing it a lot) I can fix just the parts that I break.

 

I haven't decided if I'm going to do the simple plexiglass and wood frame or print one. I'm sure version 2 will be printed if version 1 isn't. The legs and prop guards will definitely be printed so probably the whole thing will be. Probably in Nylon which is much stronger than PLA and can withstand higher temps.

 

Any experts out there want to explain why something in my list is completely wrong? I've flown prebuilt indoor copters and also model airplanes but not quads before.

 

Make Magazine article:

http://makezine.com/projects/make-37/the-handycopter-uav-2/

 

First draft parts list with prices and links of at least one place to buy them. Prices include qty so for example $72 buys you four motors. $10 buys you 4 pusher and 4 tractor props (so you have one spare set). I don't even have a LiPo battery charger so that's in there also.

 

• [*]APM 2.6 &gps&compass

$240 $70 see edit below http://store.3drobotics.com/products/apm-2-6-kit-1

[*]motors&mounts $72 3drobotics

[*]props (2 sets) $10 http://www.hobbyking.com/hobbyking/store/__34084__turnigy_carbon_mixed_slow_fly_prop_w_adapters_9047r_sf_4_pc_right_hand_rotation_.html

[*]brushless gimbals $46 http://www.altitudehobbies.com/ipower-2208-60-brushless-gimbal-motor

[*]gimbal controller $30 http://witespyquad.gostorego.com/brushless-gimbal-v3-0-controller-module-w-sensor-martinez.html

[*]transmitter/receiver $60 http://www.hobbyking.com/hobbyking/store/uh_viewitem.asp?idproduct=8992&aff=1195429 ALSO http://blog.oscarliang.net/turnigy-9x-review-9-channel-rc-transmitter/

[*]ECS $52 (found set of 4 for $37) - should research more: http://www.makershed.com/products/gemfan-30a-electronic-speed-controller

[*]battery $8 http://www.hobbyking.com/hobbyking/store/__8932__turnigy_2200mah_3s_20c_lipo_pack.html

[*]charger $23 (is this right kind?) http://www.hobbyking.com/hobbyking/store/__7028__Turnigy_Accucel_6_50W_6A_Balancer_Charger_w_Accessories.html

[*]power module $18.50 http://www.hobbyking.com/hobbyking/store/__43766__APM_Power_Module_with_XT60_Connectors_and_6_Pin_150mm_Cable.html?gclid=CjwKEAjwv9-gBRD5ofn2jd2N0UUSJACcdilsFn_hra2-AT-wOZihgZ1AAI7mB4cOvbWfIJvbYw96vRoCqG7w_wcB

 

edit:

By getting a few knock off things I dropped the price quite a bit. The key change was to get this from amazon (1 month shipping):

http://www.amazon.com/Andoer-ArduPilot-External-Compass-Controller/dp/B00NGDS9L0/ref=cm_cr_pr_product_top

Which is $90 and replaces $240+$18.50 above or a savings of $150

 

[gr5 2,172]

By the way - keep in mind that the APM 2.6 with GPS and compass and accelerometers is just amazing and way more than you need if you just want to have fun. This system is very hackable and can fly autonomously. That means you can plot out the path on your laptop and upload the waypoints to the APM and then just tell it to go and watch and not touch the transmitter. It should auto take off, fly the route, and land. During normal use, if you turn off the transmitter I believe it will just fly back to the launch point (not sure if that feature is available in the current round of open source code but if not it will be soon I'm sure). This functionality combined with the gopro gimbal mount is over half the total cost but it makes it much easier to fly - you can tell it to hover and put the transmitter down and forget about it while you record some video from that angle or whatever and you can hack hack hack. So many features! You don't have to be a programmer as someone else has already written most of the features you might want.

 

[jameshs 42]

George,

That sounds like a good way to go - I might build one by hand to begin to understand the whole thing and then look at the printed design options.

James

 

[coen 11]

Darn, can't believe I've only just stumbled upon this thread! One of the reasons I originally bought my Ultimaker is the possibility to create custom frames for drones! I'm a mechanical engineer by trade, and currently unemployed after a nice 6 months of traveling :smile:. As I have some time on my hands, and just yesterday got my ultimaker working again, the next project is to get a drone off the ground. I can help with CAD drawing (also solidworks here), in theory also with aerodynamics, and of course part selection etc.

For me important characteristics are:

- Waterproof / land on water (as the aquatic drones vid)

- Drone/autopilot functionality instead of toy/acrobatics

- Modularity of payload design:

- GoPro mount

- GoPro gimbal

- Dropping mechanism

- Other tooling ideas

- Long flight time (the option below might be necessary to do this)

- VTOL/flight transition is really high on my wishlist, but definitely no something to go into the design of the first quadcopter I'd think (something like this, but smaller: http://www.spacedaily.com/news/uav-04zzzl.html)

Is the enquiry still online? I might fill that out if that's useful.

With regard to 3D printing:

We might want to make the design parametric or modular. If it's gonna be a waterproof stowable design, we could just eliminate size as a design factor and make it in several different sizes. 3D printing would also allow for smart ways to design the couplings of the body and arms, instead of using screws or glue: let's brainstorm ideas of shape locking geometry, or just large scale threaded designs (will include drawings when Solidworks has been installed again from the NAS).

A design which I really like for its modularity (but is not waterproof) is this one:

http://diydrones.com/profiles/blog/show?id=705844%3ABlogPost%3A1377245&page=1#comments

And of course this behemoth (which is more of a throw together, but has everything in it already, except that's its not waterproof):

http://diydrones.com/profiles/blogs/3d-printed-quad-2-0

Updated: more 3d printed drone designs:

http://diydrones.com/profiles/blogs/3d-printed-quadcopter-with-detachable-arms

Very nice integrated connectors in the detachable arms, not sure if these are the copyrighted designs that the designer refers to (it might be the lid).

 

[coen 11]

@SanderG

That was my first thought and I looked at a screwed together shell, this is the original arm:

 

Had a night of CAD-piloting as well... you could make a 3D printable (to be tested tomorrow) waterproof (with O-ring) arm connector like this:

I'll put this on the printer tomorrow just to see if threads like this are suitable, while printing it would probably be a good idea to compile the component list and their associated space claim together.

 

[closedcircuit 1]

@Coen: Great work! Thanks.

If I read your drawing correctly, it looks like you've used a classic assembly where the oring requires compression to work. I do a lot of technical diving and underwater photography and we generally avoid these types of orings if we can. There are two main reasons for that:

 

1. If the locking mechanism is not screwed tight, or if it becomes slightly unscrewed, the compression is lost and the the thing starts to leak. In addition, plastic threads are quite fragile and can easily get damaged by repetitive screwing/unscrewing or if you think that a leak will appear if you don't screw it really tight;
   
2. The oring can sometime "jump" out of its seat during assembly. If that happens, it's something that's very difficult to notice;
   

Instead, we prefer orings that sit in a groove located just above the thread. It requires a touch of lubricant (generally silicone gel), but it's much safer. I'm not sure if I'm making any sense and I've enclosed below a couple of examples. The locking mechanism here is a bayonet, but the oring principle remains the same:

Wide Angle Port - The oring is in blue on the far left of the picture

 

Adapter - The oring is the light grey bit above the bayonet locking mechanism

I hope that I haven't misread your drawing completely and that the above helps.

 

[coen 11]

@Coen: Great work! Thanks.

If I read your drawing correctly, it looks like you've used a classic assembly where the oring requires compression to work. I do a lot of technical diving and underwater photography and we generally avoid these types of orings if we can.

[...]

Instead, we prefer orings that sit in a groove located just above the thread. It requires a touch of lubricant (generally silicone gel), but it's much safer. I'm not sure if I'm making any sense and I've enclosed below a couple of examples. The locking mechanism here is a bayonet, but the oring principle remains the same:

 

Another diver here, thanks for the info, now I know why the O-rings on my diving light are the way they are :grin:.

The location of the O-ring should be easily adaptable... have you got any experience printing bayonets? The threaded one is on the printer right now, and already I see that the 45 degree overhang angle on the thread is not printing correctly :sad:. Well, the nice thing with having a 3D printer is that prototyping is really fast...

PS: quit curious if it's possible at all to 3D print an O-ring seal... due to the microchannels that will occur due to the 3D printing lines...

 

[closedcircuit 1]

I haven't tried printing a bayonet (yet)? I don't know if it would work in this instance as bayonet mechanisms often rely on water pressure to prevent accidental openings. I guess that it would require an additional safety lock, or something.

 

PS: quit curious if it's possible at all to 3D print an O-ring seal... due to the microchannels that will occur due to the 3D printing lines...

 

The oring needs to be both flexible (to increase the surface area connecting both parts) and strong...

@Sander: Would you be able to test UM's flexible PLA and see how an oring would come out? In particular, would it be smooth and strong enough to resist friction?

 

[jameshs 42]

Last time I used UM's flexible PLA it was flexible - but not rubber like.

But the UM can print ninjaflex which is rubber like and for small things like the O ring could probably be good - even better if an O ring is used in conjunction with a gasket.

I would think that the issues of water should be addressed by making sure that the bits that need the most water resistance are heavily protected - and the rest is designed so that a landing on water is done so the drone sits with no seams in contact with the water. As these are not going to submerge the pressure should not be too great - and ironically there should be pressure relief to allow for heating/cooling cycles (I am a rower and we have small holes in our buoyancy chambers as boats have been known to explode in hot weather)

James

 

[coen 11]

Well, emptied my roll of white PLA and finished printing the coupling. Learned a little bit about designing threaded connections (mine screws the wrong way :shock:), but the connection it makes is really tight... really really tight because so far I haven't been able to pry it apart again.. hahaha so pictures of the thread will have to wait until I manage to get them unstuck... here's the assembly though:

 

[jameshs 42]

been there, done that

- need to get into my inventor thread tables and adjust them for clearance for printing, unless someone has already done it

 

[coen 11]

been there, done that

- need to get into my inventor thread tables and adjust them for clearance for printing, unless someone has already done it

 

Tips I recorded from this one:

bolt thread: counter clockwise

nut thread: clockwise

ensure that the drawing planes are perpendicular to the part, not to the thread (ie. solidworks lofting), no following guidecurves or start/end tangency. The sinking away of the thread is done by increasing or decreasing the radius of the helix in the last 1/4 turn.

With respect to tolerances someone else mentioned a 0.2mm clearance on this forum, which I also used and seems to hold up nicely. The way I interpreted is depicted in the picture below (both threads have a 5mm pitch):

 

This seems to give a nice screwable thread. I used the same clearance for the ring that slides around the pipe: pipe has a radius of 23, the inner radius of the ring is 23.1mm.

Anyway, I managed to pry apart the connected parts, so here's what they look like:

 

With regard to using this type of thread for a water proof arm connection I'm kinda doubtful. I think the design used on the drone I posted earlier (http://diydrones.com/profiles/blogs/3d-printed-quadcopter-with-detachable-arms)

is much more versatile:

 

you can just glue the arms shut (all they do is hold some wires, and use the external connectors to let the power through.... would it be too bold to assume the conductivity of water low enough to not cause a problem if those connectors might become submerged? Another thing is to make the mechanical connection to the frame in such a way that in case of a crash it will snap or break off without causing damage to the major parts, so perhaps include some sort of snap-on locking part that will break before the arm itself breaks (if that makes any sense).

 

So that would mean sizing woofy's design to the desired electronic components and making the arms detachable using this sort of mechanism. And we'll have to think of a way to mount a gymbal to that design, as ground clearance vs buoyancy in water might be a problem. I'll put up a google sheet shortly to start digging into component space claims for various size drones.

 

[coen 11]

Ok, google sheet is up:

https://docs.google.com/spreadsheets/d/1LF8jzXoFDleyP3YF3eb1JEQPQ8eOfPjmKctFca1AVGg/edit?usp=sharing

At the moment I'm slowly adding the information that was already in this thread... ow yeah and I've been so free to create three classes and baptize the drone too :wink: :

Ultiflyer mini

Ultiflyer (no google hits, except contamination from multiflyer :smile:)

Ultiflyer maxi

Coarse descriptions are on the sheet, but since I'm not experienced with UAV's, data might be way off, if any of the experienced flyers could comment:

Ultiflyer mini: toy/rc size, perhaps with small camera: <250mm rotor-rotor, short flight time

Ultiflyer: base model: autopilot, (gimballed?) GoPro, FPV option, +/- 350mm rotor-rotor? longer flight time

Ultiflyer maxi: pro model: autopilot, gimballed digital SLR, FPV option, >500mm rotor-rotor? Long flight time

Everything up for discussion of course.

 

[mrwong 5]

Thanks Coen for starting the spreadsheet. I will add parts for each copter that would be good options.

 

[woofy 12]

Thought you knew what a quadcopter can do?

http://www.youtube.com/watch?v=w2itwFJCgFQ!

 

[Blizz 53]

Also not entirely on topic but Cirque du Soleil designed

 

[SandervG 1,521]

things have been so busy it kept me from being very active on this thread, which I really want to succeed.

Great work Coen on the doc, I am going to read it now and contribute where I can!

Another suggestion was, when we are ready, start uploading files on github where most people are familiar with and everyone has acces to.

Again, great to see you guys are picking this up!

Also, I was thinking to call the project 'Daedalus' the father of Icarus who gave Icarus and himself wings to fly. Except he didn't go to high and survived. .. Or is that lame?

 

[jameshs 42]

Love the spreadsheet - i am trying to put together UK sources/Prices for some of the components

Love the name Daedalus - it may be that Ultimaker does not want yo use the word Ulti - but if they do ti could be the Ulti-Daedalus.

There could then be several versions that take their name from the myth. The base model perhaps being the Daedalus, The waterproof model being Icarus (or ulti-icarus)

Daedalus created the labarynth in which the Minatour was kept - part man part bull - so perhaps this could be the biggest, most powerful part quad part camera

What is really nice is that things of quality are called daidala which is a great description for a family of copters

but there is more as he 'invented (this is myth after all) "agalmata, statues of the gods which had open eyes and moveable limbs" - what better description of a quadcopter with folding arms and a camera.

So some good synergies there from Sander!

James