People making videos about machining have a problem: the coolant gets everywhere. When you take a video to show the process of creating a device, the milky gunk that keeps everything cool gets all over your camera lens. AvE is experimenting with an interesting fix for this problem, with a self-cleaning camera lens. (Video embedded below, some salty language.) His prototype uses a spinning piece of clear PVC mounted on BB gun pellets, driven by compressed air. The camera can see through this spinning piece, but when the coolant hits the spinning piece, it is thrown off.
That’s not a new idea: most ships use a thing called a clear view screen or clearsight to help the crew see out of the window in bad weather. These are usually driven by an electric motor at the center, though, and AvE’s version is instead driven by the compressed air. That means that the camera can see right through the rotating part without obstruction. The next step, AvE says, is to make it thinner and mount it on a filter mount so it can be attached to a GoPro or other small camera.
This is definitely a project to watch. Anyone who has used a GoPro in the rain faces similar problems, so this might find a use outside of the machine shop as well.
32 thoughts on “Self-Cleaning Camera Lens Makes For Speckle-Free Video”
This here is a good idea.
The prototype is a bit wobbly, but as a proof of concept it is fine.
Though, his idea of using a gear motor for an electric version might be a bit insufficient.
I myself would instead put permanent magnets on the edge of the disk, then have electromagnets on the stator, turning it all into a brushless DC motor. (less moving parts that can break.)
Then probably build it with a metal ball bearing race instead of an acrylic one. With a skirt extending out in front of the mechanism, so that it flings water/liquids away from the assembly, instead of letting it fling around within it.
If the TMD (Tip Magnetic Driven) computer fans were anything to go by it’s worth being careful with this design. There was a recall on some of the early batches exploding. Since he’s probably not going thousands of RPMs I guess it’s safer though.
I’m sure I’ve seen a video where they achieve the same effect with a simple* air-curtain over the lens, no moving parts, for filming CNC machines.
I think AVE actually links to that video as his inspiration
yes, it’s here: https://hackaday.com/2019/01/27/air-knife-keeps-gopro-lens-clean-in-messy-environments/
That seems the best option if you’ve got compressed air on tap.
An electric powered clear view screen would be a better option for dealing with mud on a GoPro if you’re out biking etc.
Except as AvE shows, from the project originators own videos, the air knife doesn’t work well at all.
I think [tarkka]’s version in the HAD article seems to work pretty well – certainly a massive improvement over nothing, and in the test video they deliberately put the camera in direct flow from the coolent.
AvE mentions this directly and explains why it doesn’t work very well (boundary layer), which is the explicit reason he built this project.
Come on guys, at least WTFV.
Some comments here absolutely make no sense.
I watched this a week ago, but I couldn’t work-out why the camera has to look through the centre of the disc? Wouldn’t it be easier to put the axle of the disc to the side of/above/below the camera and just use a bigger disc?
Size. If you have to place the camera into odd places (eg into the cnc machine’s workspace) it comes in handy that its small.
These rotating windows are common on professional CNC machining centres
I think that NYCCNC uses one. And fishing boats have had them for ever.
And, briefly, motorcycles. https://ironhorsespares.co.uk/blog/turbo-visor-accessory/
ave has one on his machine as well for seeing in from the outside. i think this is so that he can place the camera inside the machine for better shots and less glare to deal with.
Ok, that’s the strangest motorcycle item ive ever seen.
I had a trip to Fingerlakes BMW rally last year, got caught in 4 separate typhoons from a hurricane on the way back.
Kept riding in rain so heavy I could barely make out lines on road- since there was nothing for miles to pull under. This woulda helped, but its too goofy to put on. Id rather find a bridge and sit it out, riding in rain is extremely dangerous. Especially on an antique motorcycle!
Looks like I am likely to find out about your last point this weekend. Wettest June in the UK since records began. And I am riding my 1921 Ner-a-Car for 60 miles round Banbury on Sunday.
(Well, if I can make the blessed thing run. I have spark and I have fuel. It’s a 2-stroke, what else could be wrong?)
AvE is definitely aware of this. While he works to hide precise details of his background, he seems to have some knowledge of spin windows used on ships which led him to get one to install on his CNC machine which has led to this idea as he discusses issues with an air blade based on a channel he watches.
All this is in his videos but the organization of his channel is not terribly good so you have to hunt around to get all the backstory.
Why not use zoom?
Because that window you could probably film trough get some coolant too…
It seems to me that the simplest approach would be a thin disc of clear acrylic, glued to a hollow cylinder. Inside the hollow cylinder is a ball race, even a metal one if preferred, then another hollow cylinder. The camera is placed inside the inner cylinder, looking through the clear disc. The inner cylinder is the stator, and the outer cylinder and disc rotate around it, driven electrically or pneumatically as preferred.
This helps to protect the camera, as there is no easy way for the coolant liquid to get inside the inner cylinder, it would have to get around the back of the outer cylinder first, then travel back to the disc to get inside the inner cylinder.
That’s basically the same idea as I posted in the comments. As well, if you used an air bearing, the escaping air would keep shmoo from making its way inside.
Perhaps NASA could build a larger version to sit atop the solar panels of a future mars rover.
Probably 2 reasons:
1. Spinning it uses power, maybe more than they have. Especially at night.
2. Dust on Mars is electrostatic, spinning a plastic or glass disk is likely to build a charge, which would make the dust stick more.
Also, they seem to have a pretty good track record without it.
1) It wouldn’t be ran all the time. It would only run for a moment when needed.
2) I am thinking of Spirit and Opportunity. Opportunity is thought to have been lost when it ran out of power due to a dust storm covering it’s panels. Sure, that wasn’t what did Spirit in but even when both were still active, every time there was a dust storm it was an open, stressful question whether the winds would clear them off again and so whether they would survive or freeze.
Don’t get me wrong, NASA still has a “pretty good track record” because neither was meant to last nearly as long as they did. But either could have gone so much worse if the winds had just blown a little differently. To think, Opportunity could still be active today if there had been a way for it to dust it’s own solar panels!
Of course the rover would still have to survive the duration of a dust storm. this wouldn’t guarantee anything. I am thinking that the rover could keep track of the time of day. If it goes for too long without light during time when it expects daylight it would automatically go into a mode where the only thing on full time is it’s heater. Once every hour or so it would activate it’s receiver in case NASA is sending it a command such as an override in case it goes into this mode by mistake. Once every couple of hours or so it would spin it’s dust shield for 15 seconds or so. Or if that takes too much power it could do it once every several hours or even once a day. Dust storms do last days right?
A great idea, but I think he’s going about it the wrong way. Spray deflectors are fairly common in the film industry (at least here in the UK, not sure about sunny Hollywood). The big problem he’ll soon encounter is that the perspex is effectively consumable and will soon cloud up, particularly when pointing it at something firing out metal chips with the liquid and this design is going to make it very difficult to replace.
A simpler solution is to use a much thinner round perspex window, then rather than making it into it’s own bearing just apply 3 points of contact around the edge. 2 can just be v-groove or similar bearings and then drive the 3rd, could even be directly attached to a small motor. To replace simply unscrew one bearing. Unfortunately I can’t add a diagram, but I hope people will get the idea.
Keep your stick on the ice.
I’d just use a large diameter commercial ball bearing, like a Z9089, print or mill an adapter to affix a larger clear disk and a belt to the inner race, and rely on the disk spinning to keep the bearing clear of cutting fluid too.
Err, actually, 6806Z…. I have the actual part, ripped off a 3D printer cartridge fuser drum, but was reading it upside-down
I suppose setting up some kindof laminar flow in front of a window would do the trick in some situations.
Isn’t that the principle behind rotating glass disk boat windshields?
… aaand after watching the video, I know that that’s precisely where AvE got the idea.
Please be kind and respectful to help make the comments section excellent. (Comment Policy)