Blog

Stripping GoPros To The Bone For Model Rocketry | Hackaday

The small size of action cameras has made them a great solution for getting high-quality experimental footage where other cameras don’t fit. GoPros are [Joe Barnard]’s camera of choice for his increasingly advanced rockets, but even the smallest models don’t quite fit where he needs them. They also overheat quickly, so in the video after the break, he demonstrates how he strips and customizes them to fit his required form factor.

[Joe] starts out with a GoPro HERO10 Bones, which is a minimalist version intended for FPV drones. He likes the quality of the 4K 120 FPS video and the fact that he can update the settings by simply holding up a QR code in front of the camera. The case appears to be ultrasonically welded, so careful work with a Dremel is required to get it open. The reveals the control board with an aluminum heat sink plate, and the sensor module on a short ribbon cable. For minimal drag[Joe] wants just the lens to poke out through the side of the rocket, so he uses slightly longer aftermarket ribbon cables to make this easier. air cooled heat sink

The camera’s original cooling design, optimized for drone airflow, meant the device would overheat within 5 minutes when stationary. To increase the run time without the need for an external heat sink, [Joe] opts to increase the thermal mass by adding thick aluminum to the existing cooling plate with a large amount of thermal paste. In an attempt to increase heat transfer from the PCB, he also covers the entire PCB with a thick layer of thermal paste. Many of the video’s commenters pointed out that this may hurt more than it helps because the thermal paste is really intended to be used as a thin layer to increase the contact surface to a heat sink. It’s possible that [Joe] might get better results with just a form-fitting thermal block and minimal thermal paste.

[Joe] is permanently epoxying three of these modified cameras into his latest rocket, which is intended to fly at Mach 3, and touch space. This may look like a waste of three relatively expensive cameras, but it’s just a drop in the bucket of a very expensive rocket build.

We’ve seen GoPros get (ab)used in plenty of creative ways, including getting shot from a giant slingshot, and reaching the edge of space on a rocket and a balloon.

I assume the limiting factor is the weight of the camera. The best tradeoff in terms of weight vs. passive cooling capacity is probably a relatively thin and wide heatsink. But on the other hand, adding a tiny fan could be a good tradeoff if audio quality is not important.

That’s more like what I was thinking. Start with a finned heatsink, mill it down to size, then ad a small fan. The trick is to REMOVE the heat, not “contain” it as he seems to be doing. (I learned a lot burning holes in carpet playing with those TEC modules)

The problem is that there’s really no way to get the heat out of the body. Yeah, you’ve got a fan, but it’s circulating air inside of a 5″ x 8″ tube, it’s not going to make much of a difference. You can’t add many significant holes to the tube because the motor that this rocket is flying on is known for shredding rockets. The more holes you add, the weaker the airframe gets, the more likely the rocket is to shred. Among other things. This simple solution is sufficient in this use case.

The issue is this rocket is going to be so fast he’s going to cook a piece of meat under the nosecone. So there’s no place to remove the heat to thanks to the supersonic heating. It seems like that’s the case for most of the rockets in this class, hence why other people are taking this literal approach of just dumping heat into a sink.

if it is just sinking the heat rather than dissipating it, probably water? Huge heat capacity and if it overheats it will just show itself the exit. And is less dense than aluminum to boot.

I recently watched this in its entirety and while it definitely fits the “hack” definition, the creator is somewhere basically just a NileRed-quality ‘enthusiast’ who either is playing dumb for the camera or actually has no idea what they are doing at all. Despite the video script, demaking a commercial camera for its raw components versus just buying them is stupid and wasteful and with so many DIYPCB from scratch tutorials out there, I left the video disappointed and cheated that I’d wasted time on it. How are you going to build an ‘amateur rocket’ for $10,000? How are you going to also build said rocket with 10lb of hot glue unevenly spread throughout the payload and 100% sideloaded electronics? While the self-landing system was impressive, this video coming afterward and just being a toddler ripping apart expensive toys and claiming to “not have experience, just a big wallet” is off-putting and honestly pretty embarrassing.

Depending on skill set it’s definitely easier to take apart a camera and add a heatsink, versus writing code turning a 4k 30fps into 4k 120fps camera module

Some of my combat robotics buddies have explained that knowing what parts of a robot can be jank and what parts have to be well done is a skill that requires many years of experience. I think that’s a very true and valuable statement. As someone from the rocketry community, I’d like to say that Joe is pretty good at this skill.

An argument against doing the entire thing yourself is that this requires an entire skillset that isn’t necessarily applicable to our hobby, and learning it just to get a half-assed camera for $50 out of it ends up being way less cost effective than just spending $200 per camera instead. Even if the camera that pops out of it at the end is exactly the same as the GoPro, firmware and everything, was the time it took to learn all of that worth less than the difference in cost between the DIY camera and the GoPro? No.

Cameras are more than just the digital tech. He’s used others and the Go-Pros have better video quality.

Having experience with custom camera solutions (both CSI and LVDS), this comment reeks of some armchair expert who hasn’t tried doing anything with camera sensors before, but saw the Raspberry Pi camera sensor so it must be easy!

This is way off-base. It’s a reasonable choice to not build your own video recorder module from scratch, but furthermore Joe is in fact pretty good at what he does.

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.

thermally conductive silicone rubber By using our website and services, you expressly agree to the placement of our performance, functionality and advertising cookies. Learn more