Just like that I have mounted the GoPro Session to my 215 Hopper. Having a 3D printer on hand is really cool!
There was some delay between coming up with the design and printing it as I managed to break the printers glass build plate, my parts were sticking too well. Previously I was using the glue stick approach for 1st layer adhesion but I suspect I may have been a bit heavy handed with the application.
After replacing the glass I decided to try acetone/ABS slurry applied to the bed and have been very pleased with the results. Applying it is simply a matter of squirting a couple of milliliters of acetone on to the bed and then swirling a failed print (preferably of the same colour) around in it until it starts to become tacky. I’ve noticed a couple of advantages to this process. Firstly the surface quality of the bottom is as smooth and polished as the glass. Secondly, and more importantly, as the bed cools the part almost completely separates itself. Only the slightest twist or push is required to remove it. Further cleanup of the bed is not required unless you intend to change colour as the next application of acetone will simply absorb anything left behind.
The concern I had about prop. clearance to the straps on the side has proven not to be an issue.
I’ve not had a chance to fly with it in place yet so I’m not really sure what sort of result I am going to get with hard mounting and the 20° angle but when I do the footage will be here.
Just a quick update this evening. Shown here is is the first draft of a mount for the GoPro Hero4 Session on my 215 Hopper. I will need to check for propellor clearance on the side of the mount, there is clearance to the plastic but possibly not enough for a strap. If that does turn out to be problematic then I have an idea in mind for a plastic clip over the top.
Also on the cards are two other ideas for less typical mounting positions that should both provide an interesting perspective.
I have been a bit slow in getting this report together but I am pleased to say the as of late last sunday evening my second build of the 215 hopper design is airborne. For anyone keeping track that is just a 6 day turnaround from lost to flying again. There are a couple of things that worked in my favour to achieve this.
- Firstly the 3D printer from the workshop of SteelCity Electronics is currently on holiday to my workbench, as such I was able to start printing replacement parts immediately (more on this topic in the near future, I’ve got a big build cooking).
- The guys over at NextFPV actually had stock of every replacement part I needed. I expected I would need to order various bits from multiple sources internationally but discovering everything was stocked locally was a pleasant surprise. On top of that the service they offer is exceptionally fast. I am a very happy repeat customer of theirs. I the future I will gladly purchase through them whenever possible (hopefully there is an FPV setup in my future).
- When ordering parts for the first build I doubled (or more) quantities of the various bits of hardware so all of that was on hand. The PCBs were taken care of as DirtyPCB’s supply 10 per order by default.
With the rebuild I got the chance to address a few of the shortfalls of the first build. Changes included:
- Shortening ESC signal/ground wires so that there was less wire bulk in the body. it was a bit challenging squeezing everything in to the first build.
- Direct connection of motors to ESC’s. On the first build I needed to join the motor wires with the ESC wires as the motor wires were cut to short from the Spidex 220 build. In reality this is still not a very practical solution as the wires first need to be passed through the side plates. However for the sake of minimalism I stuck with it.
- Externally accessible USB. This is the most valuable change from the first build and is thanks mostly to the change in location of the port on the Naze32 Rev. 6 board. I also printed a unique side plate with an opening to suit.
- The upgrade to the Nae32 Rev. 6 also allowed for very tidy connections all around (no more soldering wires directly to the board). Seen in the photo below is the way I have installed the pin headers on the Naze32. 90° headers are used on all connections. The ESC outputs are installed in a fairly typical configuration. The RC input connections are under and towards the centre of the board and the extra features (Vbatt, buzz etc.) are directed back across the board.
- SmartPort telemetry. Rather than trying to fit a buzzer inside (which becomes surprisingly large in such a small space) I connected the SmartPort on the X4R-SB to a soft serial port on the Naze32. A buzzer will fit inside but for now I am relying on the telemetry to know the status of my battery. The photo below shows how tidy this setup is. Also note that for clearance to the ESC connections on the Naze32 I have to trim the top row of headers from the X4R-SB.
The only quirk with this build, and I find it strange given that it is newer hardware, is that LuxFloat can’t run reliably ( I had no problems with it on Rev. 5 hardware). It may in fact be the extra processing load from the soft serial so I will have to check into that further.
Next on my agenda for this copter is to get the GoPro mount(s) sorted. Then put together a more thorough look at the design and build process with a full bill of materials and files if you would like to build you own.
Given that the design of the 215 Hopper was inspired by the FliteTest VersaCopter I thought a family photo was a fitting way to end this post.
This shall be a brief post of 2 parts, the first sweet, the second bitter.
To start with I have completed the build and first flights. The photo below is just after completion and ready to take to the air. I very happy with how it has come together. Admittedly there are a couple of problems and mistakes to address but for a first iteration I am very happy with it.
Here the 215 Hopper can be seen airbourne on its first battery. Admittedly, the very first hover attempt was a complete failure as I had the motor connections mixed up side to side. After correcting that issue a successful hover test was completed indoors. I was immediately struck by how much more stable it was that its predecessor, the Spidex 220, and that right there was the main reason for starting this project. Success!
I have not yet mounted it but in the future I will fly a GoPro session with this platform looking something like this:
Now for the bitter part. On just the 5th battery I lost the 215 Hopper. I am not going to try and make excuses, I got a little too far away, that’s not very far for something this small and fast, orientation became difficult. Before I knew it and before I could regain my composure the 215 Hopper was but a small speck high in the sky and a long way away. The signal began failing so I reduced the throttle to try and bring it down gently, more out of concern for where it might land than to protect the copter itself. Evently the signal was lost. I did a few laps of the neighbourhood but my chances of finding it were slim. Most disappointing was that it likely came down over a residential area. I always like to think that I am responsible with where and how I fly and am aware of my surroundings but today I let the community down. I can only hope that it hasn’t caused any damage or injury.
If you have found the 215 Hopper (this is the only one in the world) in the Fairy Meadow/Wollongong area then please drop me a line via the contact page.
It has been a while since my last update. Primarily due to the holiday season followed closely by an international getaway which spanned several weeks (some photos from my adventure will appear here in the near future). Since my return I have started to put together the 215 Hopper. It is not quite complete yet, there are still some wiring details to complete/tidy up however it is close as can be seen below. After the maiden flight I will put together a full write up on the parts, build, what I’ve learnt and how it has worked out.
This progress also means that the Spidex 220 has officially been decommissioned.
Thanks to my friend over at Steelcity Electronics I now have all the required 3D printed pieces on hand to build my 215 Hopper Quadcopter. Unfortunately I was a bit distant from the printing process as our schedules simply didn’t overlap at this time of year. I had hoped to have a closer look at it and learn a bit more throughout the process. That will Have to wait for next time. There is some cleanup and finishing work required as part of the build process but that was by design to ensure a nice fit.
Also shown in the photo above is the printed circuit board as received from DirtyPCBs. They are my go to for prototype PCB’s at the moment as their prices and options really can’t be beaten so check out what they have to offer. I will also let you know that the 10cm x 10cm size they state as the limit for the higher price point is not in fact a fixed limit. This board (fabricated as a connected pair as shown) had overall dimensions of approximately 15.5cm x 8.5cm which is just over 130cm². Thats an extra 30% more area that the quoted size of 10cm x 10cm and they were made without a question or hold up.
This was the first time I produced connected boards. DirtyPCBs is particularly suited to this as they will happily cut internal profiles/slots at no extra cost. The snap tabs drilled with small holes approach worked very well and the edges clean up easily and quickly with a small file. I have found a couple of small errors on the board but no show stoppers thankfully.
Next up will be a full bill of materials and build log. If everything works out then I will also supply the files for 3D printing should anyone wish to duplicate what I have created. That however (and unfortunately) will have to wait several weeks.
Through the cumulative effect of multiple crashes and initially over tightening I managed to snap one of the G4 plates that make up the motor mount tube clamps on the Flite Test VersaCopter. Unfortunately at the time I ordered the crash kit for the VersaCopter it didn’t twig that these plates were sold separately. As such I was faced with either making a very small order with Flite Test for replacements or coming up with my own solution.
I took this as an opportunity to test my tube clamp motor mounting method and the folks over at 3dprint-au.com both of which are critical elements of the MultiChase project. I knocked up a very quick model of a direct replacement part for the Flite Test designed tube clamps and sent it off for printing. 3Dprint-AU use an SLS printer which I hope will provide better dimensional accuracy and a more homogeneous structure than a FDM style printer.
The results are not as spot on as I expected but I generally found the parts are oversize rather than undersized. For example the bottom edge of the clamp, as shown above on the right, is raised rather than rounded as designed. A quick touch up with sandpaper (applied only to the motor side) brought the overall measured height down to the designed number. Also interesting to note here is that the ‘black’ material offered is actually just the white material dyed black on the outside. I’m not sure of the specifics of this process, if it is done by the print head or if it is done as a post process but it is something to bear in mind.
To ensure a nice clean holes for the mounting bolts I actually printed the hole under size and drilled them out as appropriate, they are shown here as printed.
The installation process was simpler than the flight test motor mounts as there are less pieces to hold together. I also think it looks cleaner than the original pieces.
Strength wise I have no doubt that this well be less susceptible to the same style of failure as the original design. Everything is nice and snug with no visible deflection when tightening. That does however raise some concerns about the next failure. My mounts fit much more tightly around the tube and seem much less inclined to rotate as the originals would in a crash. This could make the frame itself or the arm tubes more vulnerable. Despite the much bulkier appearance these mounts only add 1.7g per corner. Each printed half weighs 4g for a total of 8g per corner whereas the Flit Test design weighs in at 6.3g per corner.
Now it’s time to get this rig back in the air, it has been several weeks since I
broke flew it last.