Printed parts for FPV builds

Where building stops and buying ends

Every other part of an FPV quad is bought: the frame, the flight controller, the motors, the ESCs. All of it is someone else's design, and the most a builder does is choose between options. Printed parts work differently. The builder draws them, adjusts them, and reprints them until they fit. A camera mount two millimetres too shallow for a lens housing is not a return, it is a five-minute edit and a short reprint. This is where a builder stops being a customer and starts making the small decisions the industry does not make for them, which is why this page covers which filament to reach for, which settings matter, and what should never go on an airframe.

Why TPU dominates on an airframe

A crash puts a sudden, sharp load through whatever hits the ground first, and on most builds that is a printed part: an antenna past the frame edge, a camera mount ahead of the canopy, a battery pad under the belly. That part's job is not to be strong the way a structural beam is strong. It has to absorb the load and give it back slowly instead of passing it into the screw, the standoff, or the flight controller stack underneath it.

A rigid plastic resists the load until it cannot, then cracks, because it has almost nowhere to put the energy except into failure. TPU, thermoplastic polyurethane, is flexible: instead of resisting the impact, it deforms, the energy goes into that deformation rather than a fracture, and the part springs back. That is the whole argument for printing accessories in TPU: a part that deforms and returns beats one that shatters and usually takes the camera, the antenna or the battery with it.

The shore hardness question

TPU is sold across a range of durometers on the Shore A hardness scale, and the number matters as much as the material name. 95A is the common general-purpose choice for FPV parts: soft enough to flex and absorb a hit, stiff enough that a camera mount still holds its angle and a threaded boss still grips a screw. Go softer and the part isolates vibration and absorbs a harder hit more readily, which is why some motor soft mounts use it. The trade-off is that softer TPU strings more during printing, holds tolerance less precisely, and sags under its own weight, making fine features, a threaded insert boss or a thin camera cage wall, harder to print cleanly. Most builders default to 95A and only go softer for a part whose entire job is absorbing impact.

What actually gets printed

The same short list recurs on almost every build: small, specific to one frame or camera, and cheap enough in filament that a failed print costs a few grams, not money.

• Camera mountHolds the FPV or HD camera at a fixed tilt; usually TPU so a knock does not crack it
• Antenna mountKeeps the VTX antenna clear of the props; almost always TPU, since it is often first to clip a gate
• VTX holderA cradle or strap point for the video transmitter, sometimes rigid where it sits inside the stack
• Motor soft mountA grommet between motor and arm that damps vibration before it reaches the flight controller's gyro
• Battery padA textured or strapped surface on the belly that stops the pack sliding and takes the first hit in a belly-down crash
• Arm guardA sacrificial cover over an exposed arm tip, printed to be replaced rather than to protect the carbon forever
• GoPro / action-camera mountA dedicated cradle for an action camera carried alongside the FPV camera, shaped to its own mount

That specificity is why these are printed rather than bought: no manufacturer can stock a mount for every frame and camera combination, but a builder with a spool of TPU can produce the one they need. The motors and ESCs guide covers what a soft mount is damping.

Print settings that survive a crash, not settings that look good

Walls over infill

A part that resists an impact gets its strength mostly from its outer walls, the solid perimeter loops the slicer prints around the shape, not from the infill pattern inside it. Infill supports compression and fills volume; it does little against a sudden bending or shear load, which is what a crash applies. The practical consequence: a crash-durable part is built with more perimeter walls than a default slicer profile and a comparatively modest amount of infill. Pushing infill higher gives back little strength for the added weight; adding another wall usually gives back more. Neither figure is fixed, since it depends on the printer, the material and the part, which is why this page states the principle rather than a number.

Orientation decides whether a hit cracks the part

FDM printing builds a part in layers, and the bond between two layers is weaker than the plastic within a single layer, true of every FDM material to some degree, TPU included. A part printed so its layer lines run perpendicular to an impact's direction asks the weakest bond in the part to resist that load directly, and that is how a printed mount delaminates: not a clean crack through the plastic, but a split along a layer line. Orienting the part so the layer lines run parallel to the expected load, or so the load compresses the layers together rather than pulling them apart, is the difference between a mount that survives a crash and one that comes apart along a seam. That means thinking about which direction the part will be hit from before it goes on the print bed, not after.

Where PLA and PETG still make sense

PLA is rigid and dimensionally precise, which is also why it is wrong for anything exposed on an airframe: it does not deform under impact, it cracks, and a canopy or antenna mount printed in PLA turns a survivable knock into a broken part. PLA also has low heat tolerance, low enough that a quad left on a dashboard or in a closed car in summer can warp a part. None of that makes it useless: it suits a battery strap guide that never gets hit, a printing jig, a stand, a case for spare props, anything on the bench rather than the airframe.

PETG sits between PLA and TPU: more layer adhesion and impact tolerance than PLA, but still a rigid plastic that cracks under a sharp enough load rather than flexing. It tolerates heat better than PLA and suits a rigid VTX holder or a flight controller mount inside the stack that rarely takes a direct hit, but it is not a substitute for TPU on anything likely to be first to the ground. The flight controller guide covers what that stack is built from.

Weight discipline: printed parts are where builds quietly gain grams

Every part above is light, which is why weight discipline slips here. A camera mount, an antenna mount, a VTX holder, two motor soft mounts, a battery pad and an arm guard do not look like much on their own, and they are usually printed thicker than needed, because a thicker wall is faster to design and rarely fails on the bench. Add them up on a small quad and the printed parts alone can account for a noticeable share of the all-up weight, on an aircraft class where every gram changes thrust-to-weight and flight time. The fix is not thinner walls on the parts that take impact. It is being honest about which parts need TPU's toughness and which, an arm guard on a frame that rarely tips over, can be printed lighter or left off.

This domain's frames were open source, and the files are still findable

Before this site existed, mode-2.com belonged to Mode 2 FPV, a US-based maker of carbon FPV frames active from around 2013 to 2020. It released its frame designs, not just the accessories, under the GNU General Public License version 3, with source published on GitHub and Thingiverse. That is a large part of why this page exists: a real share of the links pointing at this domain come from Thingiverse, from builders who found a frame file years ago and are still citing it.

The Thingiverse profile behind those designs, handle mode2fpv, is still online, with no connection to this reference site: thingiverse.com/mode2fpv/designs is where to find the actual design files. The GitHub side is a genuine gap: the archived pages linked out to GitHub for each frame, but the repository URLs were never captured and no account has been located since. If a GitHub search turns up nothing, that is the honest state of it, not a broken link here.

What an open licence gets a builder is concrete: a GPL v3 release means the source files were published, not just finished parts. In practice: you can print the accessories for a frame you own without asking anyone; you can modify a file, thin an arm, widen a mounting pattern, and print your own version; and where a frame is discontinued, as most of Mode 2 FPV's now are, the file to reproduce a broken or lost part still exists, under a licence written to outlive the company that used it.

None of this means Mode 2 FPV's frames are documented in full here. Verified specifications for the frames that survive in the Wayback Machine live in the open-source frames archive. For choosing a frame for a new build rather than tracking down an old one, the frames guide covers geometry and material, and the wider components section is where to start.