17 replies to this topic

[John888]

Hello, I've been in this hobby for just over a month, with about 2 weeks flying time out of that, and no real RC experience prior.

Pretty hooked as I've already built 3 different DIY quads in that time.

Here's another idea, perhaps build log over the next little while.

Seeing I have a perfectly functioning standard quad, the priority is concentrating on learning how to fly FPV, with a bit of tinkering in the background.

I know there's a guy called DEX who makes some really sexy looking fibreglass frames. Would love to buy one of those but believe that they're hard to get, not totally waterproof and perhaps a bit small to fit much FPV gear?

Anyway, here goes...

The capsule/case for my normal quad started off waterproof, a tupperware container.



It's easy to just waterproof part of the quad, can just seal the holes I made for ESC wires, the TX antenna and USB cable - with silicon sealant, and the CC board and RX would be in their own waterproof capsule. Will probably do that soon.

The other simple solution would be to run pool noodles down the arms of the quad, which would provide more buoyancy than the weight of whole unit, making it float if crashed into water.

In this scenario, all electrics would be lost, bar the CC board and RX. Possibly main FPV electrics could be preserved, if that stuff was in it's own tupperware container too.

But the challenge is to build an almost totally waterproof multi rotor.

Here's what I've done so far...

1. tested making a motor housing at the end of existing arms. 50mm PVC piping (from Bunnings) was used. This allows a little airspace around the motor, as heat might be an issue in an enclosure that's too small. Can always move to larger diameter piping, or most other motors have smaller bells than the DT750 used for reference.

This was waterproof, I left immersed overnight (pic below). Buoyancy was also added by the airspace in the 50mm housing.

The wires should be ok as long as sealed at each end and shielding is intact. Have made drop cams with dive buddies that we use in the ocean with housed security cameras at the end of 200m of Cat 5 cable, and haven't had a cable prob yet.



The photo might not show it properly, the housing is immersed and so are the shielded parts of protruding wires.

But... using PVC tubing allows it to look a bit neater and resist abrasion, with wires threaded through the middle.
Plus possibly adding a little buoyancy, due to the airspace inside.




2. Sealing. The next question was how to seal the motor housings. The bottom and arm slots are easy, can use epoxy or silicon. The motor will be hard to access once sealed, an idea is to probably make a complete spare arm, and swap it out if there's a problem.

Using silicon means a flexibe sealant, so it can be cut away, then resealed once any work is done to the motor.



I used a 50mm hole saw to make top and bottom caps (pictured below), drilling into the side of a plastic (Ikea-style) storage box.



So this round bit, sealed with silicon glue will be the bottom of the motor housing. Epoxy is something to think about as an alternative, then restricting access from only the top.

By chance, the hole saw makes a hole in the middle, which is great to poke the motor shaft through at the top.

The primary weakness of this design will be the prop shaft, as a millimetre or two needs to be allowed around it to spin. With the battery mounted low, it should float upright, so the main danger will be flipping the multi rotor in the water.

Going to try a number of different ways of sealing the prop shaft, using vaseline is one idea.

Another is trying a flexible boot (pic below) that they use in model boats. Not very confident that this will handle a fast rotating prop shaft. Perhaps this plus vaseline?

Suggestions would be greatly appreciated!







The arms could then be epoxied into holes drilled into the main body of the unit. Each motor housing will be sealed separately, so if there's a leak in one, it shouldn't reach the main body or other parts of the unit. A bit like chambers are individually sealed in boats, hopefully containing a potential problem to only one area.

3. Body - Tupperware again here.

The container needs to be oversized, not only to fit everything in but because of heat.

The ESC's and FPV electrics will all be housed in the main body, meaning that extra airspace will be needed to deal with heat.
I'm using Turnigy Plush 30A ESC's which don't seem to be too overworked, keeping head down in general.

If heat becomes a problem for the ESC's or other electrics, I have some small computer heatsinks from Goodluckbuy to deal with the issue.

Being clear helps you see if water has seeped in.



I picked this particular model of container for a few main reasons;
a. It has 4 clips, instead of 2 that the rectangular ones have (giving more impact resistance and security).
b. It's deep enough for multiple layers of electrics and a small frame to hold the arms.
c. And wide enough to (just) take the battery across its width. A 3000MaH battery fits flat.
d. the volume is 1.5 litres, giving approx 1.5 kg of buoyancy, more than the weight of its contents.
e. there seems to be enough spare airspace for cooling (I hope).

Once loaded up, it might be slightly too small, if a problem with space or heat, I might have to try and find the same thing in a 2 litre size.

So the CC board, RX, ESC's, battery and FPV gear will all be sealed in this enclosure.




Guessing that making it waterproof will result in about a 1.8kg AUW (say 2kg max), so thinking it will be a quad with large props, or maybe a hexa with 8 inch props.
Perhaps a hexa might provide the added advantage of a little redundancy? Not sure yet.

4. Buoyancy - The airspaces in housed areas should provide greater buoyancy than components, making it float.

Going to try and position weight (the battery) so that it naturally floats upright. As mentioned, flipping in the water will be the main danger.
Will probably add some flotation at the top of the body, so it wants to right itself, naturally.

An added precaution will be pool noodles along the arms, probably resulting in more than 5 kg's of floatation for a multi rotor that weighs about 1.8kg. Buffer if an area gets flooded.




A GoPro in its housing will be stuck to the outside of the main body. A FPV camera could also be made splash/water resistant on top.

That's about it for now. Everything needs to be thoroughly tested with the right weight but without electronics inside, which will take a while...

Edited by John888, 31 January 2012 - 06:47 AM.

[Mat Wellington]

Be careful with the poolnoodle close to the props, it kills lift and yaw authority.

Cool project

I am toying with building a quad from the large sistema lettuce container, about 200 mm diameter and 150 or so deep, house everything including lipo.

Use 4 plastic glands for the arms to come out, wires in the arms, silicone at one end to stop water. Wouldn't worry about waterproofing the motors, water won't kill them, just dry and re-lube the bearings, cut down on alot of weight

[Reddog]

Very cool. One thing that may make the project simpler is you do not need to water proof brushless motors, they can run in water and it does not hurt them.

Wrote the above and then saw Mat's comment above, sorry.

[Drifter]

another thing to consider is how hot will the motors get without the air cooling

[John888]

Thanks Mat and Reddog, I think trying to waterproof the motors was half the challenge!

Isn't Sistema from New Zealand Mat? Every time I go to the supermarket I buy more! Will take some photos next time.

Drifter, it sounds Mat and Reddog have done away with the need to house the motors.

Was planning on bench and water testing a fair bit before trying outside.

eg. running the motors in their enclosure for say 30 mins and seeing how hot they got.

The same with heat of ESC's inside the main body.

And of course dunk testing.

Mat, know what I'd really love you to try.

Maybe make a breakthrough re. props.

As a newbie, it seems as though a lack of really large counter rotating props is holding back longer flight times.

What about if you designed a model that had motors at an angle, so it wouldn't need counter rotating props, then use 18", 19" or 20" model plane props?

Maybe double the time/increase the range for FPV-ers and AP people.

Could help set a new benchmark for amateur search and rescue type builders too. Particularly useful when the Revo becomes available.

Would ones like these work?

http://www.hobbyking...?idProduct=8086

http://www.hobbyking...idProduct=11372

http://www.hobbyking...idProduct=11403

With a motor like this?

http://www.hobbyking...pter_Motor.html    (was in stock the other day)

http://www.rctimer.c...75&productname=

[Dado]

Awesome dude, just awesome!!

im also very intrested in makeing a system for a multirotor rain flying gear, altho motor can stay open thats the section of my biggest intrest.
btw you could make a thick bell that is mergred with rotating bell in upper part of motor, and it gets just a slightly bigger radius before it meets 2nd bell fixed to the arm bellow so it overlaps it. (probably will require precise ballancing)

no watter comeing in but enough air flow.
just a thought ofc.

Edited by Dadorcp, 31 January 2012 - 10:17 AM.

[XXL-Wing]

Making ESCs waterproof is not a big issue. I put electronics coating finish on the ESCs which makes them at least proof against rain, never tried with immersion but should also work.

[Adverse Effects]

Master Airscrew propeller 16x10 3 blade

Master Airscrew propeller 16x10 (Pusher)3 blade

[John888]

Adverse Effects, on 31 January 2012 - 11:28 PM, said:

Master Airscrew propeller 16x10 3 blade

Master Airscrew propeller 16x10 (Pusher)3 blade

Thanks!

Just ordered a set, out of curiosity.

[Mat Wellington]

the model boat guys wrap ESC's in gladwrap, still transfers heat, but keeps water at bay


for fully enclosed, you could get a plate of aluminium, cut a hole in the box , glue the plate to the outside and remove the shrink wrap, add a bit of heat transfer paste, and stick the esc's to the plate on the inside - effective heat transfer !

similar idea had the same plate, and a small 12 vdc cooling fan on the inside circulating the air, to help keep everything cool

yup, i have heaps of sistema boxes too, all sizes, some we even put food in !

people already put a 5-7 deg angle on the motors that should have pusher props when using normal props, it works well aparently

i would not bother with any waterproofing of motors, the centre of the prop is more than enough to stop rain blasting the bearings, and the water in the air will help cool the motors, just dry out well afterwards and lube the bearings straight away.

with the big props you can start running into issues with the mass being to great to effectively control, not sure what the absolute maximum is, have heard 18"

if you have massive props on a lightweight multi, your power required to hover drops , but becomes super sensitive to and small gusts and control starts to become difficult

i have found, a heavily loaded system, with an efficiency of 170-190 watts to hover 1 kg, you get a very stable system, as the air you are hovering on is in well defined, fast moving colums  ..... at the other end, something like the lotus RC T580, using 120 watts or so per kg to hover is badly affected by any air disturbance

[Adverse Effects]

John888, on 01 February 2012 - 08:11 AM, said:

Thanks!

Just ordered a set, out of curiosity.

did you use my links to buy them?

i didnt get any points from them ;-(

owell

[John888]

Adverse Effects, on 01 February 2012 - 11:46 AM, said:

did you use my links to buy them?

i didnt get any points from them ;-(

owell

Thought I did???

[elmuchacho]

On my waterplane I use non water based vaseline, I've read in a post here someone mentioned CorrosionX and the Video on youtube kind of sold it for me.

For the brushless motor nothing is required has a brushless motor can run under water without any extra protection it might help and avoiding you building that small cage around the motor.

[BigpatAus]

Hi John

Interesting build.  How much extra weight does the tupperware add?  Also did you know that Tupperware has a lifetime guarantee?  Though I wonder how you'd go explaining you were using it for a Quadcopter when it broke

Happy flying
P

[John888]

Ok, time for an update...

I've been using my standard Tri (have 3 now, a light one for endurance testing, standard one and a 3rd made out of stormwater and plumbing pipe - just swap a single CC board between them) with a tupperware container underneath to house the battery and ESC's.

The CC board and RX are already in a small tupperware container on top.

One of the main questions was re. heat build up in an enclosed space.

In the sealed tupperware container the Turnigy Plush 30A ESC's got hot without modifications.

I cut slits in the side of the heatshrink, which reduced heat a bit.

Then added some small computer heatsinks to one side of the ESC's (thermal glueing to the existing flat heatsink). This seemed to help a bit but still faced the limitation of operating within an enclosed space.

So went a step further and cut slits in the side of the tupperware container and attached some thin aluminium bar - which had contact (through the slits) with outside air. I then thermal glued the ESC to the aluminium bar. The bar got hot and had contact with outside air, which seemed to be enough heat transfer for 10 min flights.

One mistake though, was to use a cheap bathroom silicon sealant to attach the alu strip (DIY heatsink) to the slit in the tupperware container as it was too weak to stick properly.

Have got some marine grade Sikaflex now, it's much stronger.

Here's some pics




Above is the small computer heatsink on one side, with the ESC thermal glued to the long DIY heatsink strip I made with outside contact a window cut in the container on the other side.

The photo below shows the DIY heatsink better. Pity the cheap (black) bathroom silicon sealant came unstuck. Marine grade Sikaflex is much tougher, will reattach tomorrow.



Patric (BigpatAus), think the tupperware container is about 100g, other extra stuff is maybe another 100g all up?

Yes, I might find it hard to get a lifetime guarantee after holes for the wires and heatsinks have been cut .

This is sort of an interim step, along the way to another design, a Tri using 90mm stormwater pipe as the main housing.

From the Multi Rotor section in Bunnings.



The pipe is relatively light and has enough room but end cap sections are fairly heavy.

I made a prototype that has an AUW of around 1940g (inc housed GoPro) and flew it for the first time yesterday.

The first 5 mins of testing went fine. After videoing for a few mins disaster struck, with an unexpected flip.

What had happened is that I weakened the rear 15mm PVC arm by cutting a slit in it to hide wires.

It cracked just under this slit, meaning that not having full cylindrical strength made it fail while hovering under load.

Will replace the rear arm with a new one minus the cuts, and reinforce again around the motor and servo mount.

Either that, or back to 12mm Tasmanian oak arms.

Here's a video.



Specs are DT750 motors
GWS 13 x 6.5 props
3000 mAh 4S battery
Turnigy Plush 30A ESC's
AUW inc housed GoPro - around 1940g

Will shift to Rc-timer 5010-14 360KV motors and GWS 15 x 7.5 props after testing's finished on my endurance Tri.

Have waterproofed the servo by coating it in Sikaflex.



Will move to Plasti Dip for future servos when my eBay order arrives.

[Breezemont]

...

Edited by Breezemont, 01 April 2012 - 02:52 AM.

[Breezemont]

...

Edited by Breezemont, 01 April 2012 - 02:52 AM.

[John888]

Ecalc says 21 amps max (6A @ hover). Will test soon.

Hover is around 50% throttle. My kind of flying is slow, so don't really get near max throttle much.

The 30A ESC's don't even get warm when left in the air, on the arms of the Tri, just when enclosed.

The heatsink seems to work when enclosed, temperature is warm but not hot. This is just an interim solution along the way, until I use 360KV motors and 15" props.

HK doesn't say much about the DT750 and 4S, mine (6 on 2 different tri's and 4 previously on a quad) seem to run better on 4S.