Jump to content
Mr.RC-Cam

3D Printed DJI Inspired Quadcopter

Recommended Posts

I've been away from the heli stuff for a few years and these new electronics are amazing and a new learning experience for me.

Welcome aboard! You will definitely learn a lot if you build this DiY model.

More changes. The USB port on the Vector Flight Controller was buried behind the FPV vTX bracket (see post #19). This would have made firmware updates a difficult task. To fix this I used Autodesk 123D and created a mirror image of my vTx bracket. The new part was 3D printed and installed on the right side; Now the Vector's USB connector is accessible.

Here's the relocated FPV vTX bracket (swapped from left side to the right side). While I was installing it the ImmersionRC vTx was downgraded to a FatShark 250mW 5.8GHz.

May-09-2015 EDIT: The FPV vTx has moved to the Receiver Shelf so this bracket is no longer used. See post #34 for details.

post-2-0-76687100-1429841667_thumb.jpg

There's a lot of cables in this model!

post-2-0-69983000-1429841381_thumb.jpg

post-2-0-18574600-1429841426_thumb.jpg

To save time and expense I didn't do a custom PCB for this project. So the Arduino CPU connections are a bit messy looking. But it works!

post-2-0-68363200-1429841725_thumb.jpg

Here's a video of the retracts being tested:

Despite the great results recorded in the video, follow-up stress testing with the arms & motors attached eventually caused the gear motor to fail. So I'm back to the drawing board; I've been experimenting with using the mechanics from a high-torque R/C servo and pairing it with my H-Bridge driver board. This combination looks promising but retract time is slower. However it is faster than a standard servo because I can use a much higher motor voltage. I have more testing to do before I decide on a final solution.

Edited by Mr.RC-Cam
FPV vTx has moved.

Share this post


Link to post
Share on other sites

As mentioned the retract motor failed during stress tests. The gear motor was a marginal choice for the required heavy lifting, so back to the drawing board.

After some bench testing I returned to using a high torque R/C servo to handle the retracts. This is what the original design called for. But because the retract transformation time is excessively long (30 seconds!) some tricks were needed to speed things up. My solution is to use a much higher servo motor voltage and a different threaded rod. It is possible to achieve reasonable retract times (~10 secs) by doing this.

First the electronics in the servo were gutted out since they are not used. Because 360 degree rotation is needed the servo's internal end stop pin was removed too. BTW, it's a Vigor VS-2AMB metal gear servo. This photo shows that its internal PCB and feedback Pot are gone from the lower case; All that remains is the motor, which was outfitted with two .01uF RFI suppression caps and a 1A PTC fuse:

post-2-0-75976600-1430334884_thumb.jpg

To operate the motor the H-Bridge motor driver board from the old gear motor setup is used. There's not enough room inside the servo case for it, so it's placed on the outside using a small mounting bracket that was designed in Autodesk 123D. Here's a 3D rendering:

post-2-0-41191200-1430335051_thumb.jpg

Here's the motor control board installed in the mounting bracket (a dab of rubber adhesive holds it in place). It will be placed on the outside of the servo.

post-2-0-04545600-1430335713_thumb.jpg

Unfortunately these minor changes impacted the PDB cable placement. So it was re-arranged as shown in this photo:

post-2-0-88108600-1430336056_thumb.jpg

The original M5 threaded rod was changed to a 1/4-20 size because its coarse thread pitch will help speed up the retracts. But to use the 1/4-20 rod a couple plastic parts were revised to accept the larger diameter. Plus the frame strengthener changed too because of the new cable routing. Here's the revised parts that were 3D printed:

post-2-0-78912400-1430336508_thumb.jpg

The revised worm drive (V2) uses two captive nuts embedded in the plastic to ensure the worm block travels smoothly on the 1/4-20 threaded rod.

post-2-0-34590700-1430414816_thumb.jpg

post-2-0-67290900-1430414824_thumb.jpg

  • Like 1

Share this post


Link to post
Share on other sites

What is the reason you are using a servo and not a geared motor with a more suitable ratio?

Terry

Share this post


Link to post
Share on other sites

What is the reason you are using a servo and not a geared motor with a more suitable ratio?

Indeed a gear motor was originally installed. I spent a lot of time searching for a suitable gear motor that would fit in the available space, had low weight, and was low cost. I ordered several different versions to evaluate. The most promising one was installed and it failed during retract stress testing. There may be a gear motor out there that is perfect for this model, but I've run out of time trying to find it. I want to finish the build and fly, so I'm back to using servo mechanics for the retract lifting. Had I done that from the start I'd be flying now. :)

Building continues ....

Black snake skin mesh tubing was placed over the ESC wires to help with cable routing and improve the model's appearance. The H-Bridge Motor driver board & PDB pcb were installed on the outside of the servo with Velcro tape. Here are photos that show the tight fit for these two items:

post-2-0-18452100-1430413280_thumb.jpg

post-2-0-38135800-1430416365_thumb.jpg

Share this post


Link to post
Share on other sites

Finishing up the retractable arms requires installing the pushrods. The original design uses 5mm carbon fiber tubing with 3D printed ends that are installed with M3 screws at the pivot point. I prefer pushrods that can be adjusted, so I used 4mm ball links. Some minor design changes were incorporated, as follows.

A 5mm CF rod to metal pushrod adapter was created in Autodesk 123D. They are installed on some CF tubing just like the original parts. Here's the 3D rendering:

post-2-0-07468000-1430433044_thumb.jpg

The pushrod adapters are tapped for the 2.5mm metal ball link rods and then the adapters are epoxied to the 5mm CF tubing. Here's some photos of the adjustable pushrod after installation:

post-2-0-25784800-1430433149_thumb.jpg

post-2-0-01953500-1430433163_thumb.jpg

post-2-0-51247300-1430433175_thumb.jpg

Share this post


Link to post
Share on other sites

Finished the build late last night. Everything works; The navigation LED's look fantastic and the retract works great. AUW is 2.8kg.

This morning it had its maiden test flight. It flies! Unfortunately there are significant shake/wobble issues to resolve. Hover performance is docile but I did not try fast forward flight due to the shake problem.

Time for some beauty shots:

post-2-0-62513600-1430504311_thumb.jpg

post-2-0-35553700-1430504319_thumb.jpg

Share this post


Link to post
Share on other sites

I'm currently working on the shake problem. The flight controller was relocated and that did not help. Here's a video of the shake:

Update: The shake issue has been solved with changes to the Vector flight controller's PID settings. I'll post some photos of the final build details soon.

Edited by Mr.RC-Cam
Uploaded Test Flight Video #2

Share this post


Link to post
Share on other sites

Still celebrating that it flies great now. Here's some comments about the fix.

I mistakenly suspected the shake problem was due to the Vector flight controller's installation. Because of the limited space in the model it was tightly squeezed into the frame with the battery resting on top of it. This is not a recommended way to install a flight controller and I decided to move it to a larger area with better vibration isolation. This was more work than expected, but I like the final solution. Unfortunately it seems the effort was unnecessary because the shake was eventually solved with PID settings.

The Vector's new home is at the bottom of the frame. It originally could not fit there but that was solved after several hours using AutoDesk 123D. Here's a photo of the revised component placement:

post-2-0-42225100-1430932312_thumb.jpg

Here are 3D renderings of the parts that made it possible:

Revised Framework Tail: It is stronger and now has a shelf at the front to hold the 4S battery in place.

post-2-0-59225300-1431208526_thumb.jpg

The FPV vTx was moved to the Framework Receiver Shelf which required some changes to the plastic part.

post-2-0-34401500-1430933059_thumb.jpg

The Vector Flight controller is installed under the Receiver Shelf. The original design had two 30mm tall standoffs in this area that were in the way. So a replacement for them was created that provided a box-like space with more room for the Vector. Here's the part that does this:

post-2-0-64516300-1430933574_thumb.jpg

When installed it looks like this:

post-2-0-94515900-1430934357_thumb.jpg

Share this post


Link to post
Share on other sites

The model is flying great so this 3D printed Quadcopter project is officially complete. As promised, I'm sharing all the details in case there's something another builder may find useful.

Model Summary:

Brushless Motors: 4108-480KV Turnigy Multistar

ESC: ZTW Spider Series 40A OPTO (manually reflashed with BLHeli V13.1)

Battery: Compact 5800mAh 4s 40c Lipo

Props: 14x4.7 RJX Nylon+30% Fiber 1447 Blades

Flight Controller + OSD: Eagle Tree Vector with wire leads

FPV Camera: DPC-540AP 540TVL PAL Color CCD

FPV vTx: Fatshark 5.8GHz 250mW

RTF AUW: 2.8kg

Gimbal: NONE.

Flight Time: ~13.5min hover, ~12min mixed (90% battery mAH used).

Share this post


Link to post
Share on other sites

This model was built using the plans sold by rchobbysuk. The plans are available for about $7 USD from their online store:

http://www.rchobbysuk.co.uk/blogs/dji-inspired-3d-printable

However, several of their STL 3D Printer files were modified (personal improvements). Plus I created many custom parts. Here's my STL files that are installed in the finished model (please review ALL the previous posts for more information on them):

Custom Canopy Related Files:

post-2-0-53863300-1431394578.jpgExtra room for FPV camera. Allows up to 37mm square FPV cameras. canopy_nose_with_fpv_holeV3.stl

post-2-0-84110900-1431394994.jpgNose extension protects FPV lens. canopy_nose_lens_protector_mag_V1.stl

post-2-0-01360900-1431395371.jpg. Adapter for 30mm size FPV cameras. canopy_nose_cam_30mm_adpV3.stl

Share this post


Link to post
Share on other sites

Additional Canopy Related Files:

post-2-0-33544100-1431395666.jpgDust cover for FPV lens, requires canopy_nose_lens_protector_mag_V1. Held in place with magnets. canopy_nose_dust_cover_mag_V1.stl

post-2-0-64795200-1431395690.jpgEagleTree Vector GPS stand. canopy_nose_gps_puck_2pc.stl

post-2-0-18354000-1431395715.jpgUse this to hide hole if Vector GPS stand is not installed. canopy_nose_gps_hole_coverV3.stl

Share this post


Link to post
Share on other sites

Custom Motor Arm / Boom STL Files:

post-2-0-24834500-1431396194.jpgStronger boom support part, easier installation. Use several small drops of epoxy to hold on CF tube (prevent slip rotation). boom_support_linkage_V2_2pc.stl

post-2-0-78711800-1431396204.jpgMatching stop ring. Use one drop of epoxy to hold on CF tube.boom_stop_ring_V2_2pc.stl

post-2-0-71719100-1431396218.jpg5mm CF tube adapter that is used with 4mm ball links. Print 4 pieces. pushrod_adapter.stl

Share this post


Link to post
Share on other sites

Customized Arm Lifting Related STL files:

post-2-0-49192200-1431396445.jpgUse hex nut and T-nut with 1/4-20 threaded rod. wormdrive_end_stop_holderV2.stl

post-2-0-54304000-1431396456.jpgLarger size hole and hex nut outline for 1/4-20 threaded rod. Uses hex nut on top and bottom sides to lock in place. servo_capV2.stl

post-2-0-74601400-1431396572.jpgHole for 5mm bearing. File the end of the 1/4-20 threaded rod diameter to fit bearing. framework_main_body_with_bearing.stl

  • Like 1

Share this post


Link to post
Share on other sites

Custom Framework Parts For Vector Flight Controller Installations:

framework_receiver_vector_shelf_V4_part1_250.jpgThe FPV transmitter and LED driver are installed here. framework_receiver_vector_shelf_V4_part1.stl

framework_receiver_vector_shelf_V4_part2_250.jpgEagleTree's Vector flight controller/OSD is installed in here. framework_receiver_vector_shelf_V4_part2.stl

 

 

framework_receiver_vector_shelf_V4_part3_250.jpgSMA Clamp for FatShark 5.8GHz 250mW vTx. framework_receiver_vector_shelf_V4_part3.stl

 

receiver_case_D4R-2_250.jpgOpen face case for Frsky D4R-II R/C receiver. receiver_case_D4R-2.stl

framework_tail_V2_250.jpgFramework tail V2.  framework_tail_V2.stl

Edited by Mr.RC-Cam
Added missing framework tail file.

Share this post


Link to post
Share on other sites

The big round nylon wheel that came with the servo fits snuggly in the 3D printed servo cap. Four self tap screws hold the wheel to the servo cap.

  • Like 1

Share this post


Link to post
Share on other sites

I am almost done with my build. I want to thank you for your time and effort you put in to your modifications. I based my frame on your design changes since i am also using a vector and immersion vtx only difference I am using a 20 degree arm. How is your inspire holding up with esc below the motors? I am about to wire mine up but waiting on some electric parts.

Share this post


Link to post
Share on other sites

Absolutely no problems from the ESC's being under the motors. I highly recommend this mod.

I also recommend that you duplicate the bullet connector wiring scheme I used. I learned just how important the connectors can be; I had a minor crash a couple weeks ago and had to remove the arms to replace broken T-Joints. Without the bullet connectors it would have been a miserable repair. BTW, get the Castle Creations machined bullets and avoid the cheap China made banana bullets that are sold everywhere.

http://www.rc-cam.com/forum/index.php?/topic/4022-3d-printed-dji-inspired-quadcopter/?p=28087

Share this post


Link to post
Share on other sites

Hello Mr.RC-Cam,

  I just wanted to thank you for your work. I just got the plans from the same guys and then I found your post and they are very helpful. I want to use the apm board for my build and Taranis Plus. Is there any thing that you think may work better then what you used as motors ect. ?  

Share this post


Link to post
Share on other sites

I am very happy with my ESC, motor, and prop choices. Be sure to check out the massive discussion on rc-groups; There are a lot of mods available for this build.

  • Like 1

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...