one stabilizer updates: intro to the project and getting the CAD and stuff ready [updates]
So the other day I was wondering how do I want to test out the 3D printer - I want to use it. And I have some free time this week so I thought maybe I'll add on a small project which is to use servos and parts that I already have to create a camera stabilizer that I can use with my dslr. Then if it goes well we can link the filament, the parts, and everything needed to make it to this website and make it open source.
Basically, we're going to use 3-4 of these qwinout super200 servos (200kgcm torque rating) and the 3D printer filament here: https://www.amazon.com/gp/product/B0837QMVWR/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1
Which is a carbon fiber 3D printer filament, to make an inexpensive, hand-held camera stabilizer with a design inspired by the DJI Ronin 2. The DJI Ronin 2 is a state-of-the art camera gimbal system designed and built by DJI, a prosumer and professional drone manufacturer who focuses on aerial filming.
So there are many features we aren't going to include of course including like powering a RED camera and all the intelligent tech support that comes with the camera. For electronics we're going to use an ardupilot pixhawk 2.1 and we'll power the whole thing with our 6S lithium batteries from the helicopter project that doesn't seem to be taking flight much longer.
pixhawk 2.1 ~240 dollar value
one roll of CF 3D printer filament ~33
3x qwinout super200 servo ~80*3
I have an xbox controller for kicks but not needed
In total we're looking at like 550 dollars for this stabilizer (I'm building this for free because that's my challenge this summer to spend no money on my projects and these parts were chosen since they're lying around). But compared to a Ronin 2 which is like 8000 - maybe worth it? It's definitely not better than the Ronin-S so we'll have to give this product design exercise more features than the S so let's watch that video and read up on it: https://www.dji.com/ronin-s
Okay let's be honest, there's no way me in a basement is going to do better than the Ronin-S but we can try to make it work well regardless. This exercise is more so I can learn design and use the parts sitting around me to design a product-looking device you know. And hopefully create something to share with people! We're going to start with the servo mounts because those kind of dictate the rest of the design. And we're also going to try to do the CAD in a GOOD way meaning fully defined sketches and etc.
So from the drawings and measuring the physical servo and adding in some tolerance for the 3D printer itself, we came up with this design as a servo mounting area. We'll get back to this later after we measure the SLR camera and create the carrier we're going to use. I think we should start with the carrier and not the servo mount just because that actually will dictate the size of the contraption. But before we get there - we do want to cad the other side of the servo just so we know how to mount things.
Now let's get to designing the actual camera-holder portion of the stabilizer. From what I can gather online it's very simply called the plate. Here's a diagram of other parts just for fun.
Anyways. Let's get to designing the plate and the tilt axis as our stabilizer will follow the Ronin 2 in the camera-tilt-roll-pan configuration that it sits in working from the camera to the top of the stabilizer. This design will also need space for the batteries as well as the pixhawk 2.1 - these are things we should not forget when designing it. We should also remember we have a 300mmx300mmx400mm build volume that we can use on the printer. First, let's measure the camera.
This is a Canon EOS 70D and we have placed a microphone on top of it but we're going to design the stabilizer to work without the mic because the mic will likely only pick up the noise form the servos. So we'll try to record the audio separately when filming.
That is the one disadvantage to designing like this as we can't pick quiet motors like we normally would - but no matter this is cheaper. Let's measure the camera. A 160mmx120mm space should be good enough for this camera in its entirety, let's tack on a 1mm quick release plate and then a few extra mm for the lease in the tilt and we get a height of 190mm ~ish. The first thing I want to do is go find a camera quick-release plate CAD we can download. OR we can use a 1/4"-20 screw which would be fine as well.
Yea - maybe for simplicity sake let's use a 1/4" screw for now you know. Now let's design, what we're going to call, the Camera Carrier. We're going to make the carrier itself a little beefy because of how its shaped we want it to not break. So let's go with 1.5cm, actually maybe 1cm and then fillet.
Here's a very simplistic camera gimbal design with beams to hold together this front camera carrier. The beam design saves materials and makes it strong, lightweight, and easy to print. The beef-up in the corners gives us some extra material to counter any bending since there's no top cross-bar for extra support. We're trying to print everything in one piece and this should be able to print vertically but we'll generate supports anyways. These parts are all being designed with printability in mind because if we can't print them easily it'll be a waste of a design.
Next we get to the pitch axis carrier itself. The point of this is it wraps around the camera carrier and holds the servo that does pitch/tilt, and it connects to the roll axis on the other side. So first we need to get some dimensions down... it's 5cm from the face of the servo plat to the plane of mounting of the servo. And the width of the camera carrier is about 191mm. So the INNER dimension of the next axis must be 190mm+50mm roughly. Let's make it 240 + a little tolerance so 243mm. Because we're 3D printing the negative tolerance must be slightly high. We should remember we have a 300mm build volume which somewhat limits the other dimensions.
240+15+20 <300 which means WE ARE GOOD.
So actually we chose to make this thing a little beefy-er because it seemed a little too thin.
And then we took the dimension above, modified a little, and created the following component, this is the tilt servo mount and camera carriage mount.
Now there is one problem with this piece, most printers (including ours) will have a large amount of trouble printing something of this size. So we're actually going to split this thing along a vertical cut.
So there we go that would get printed into two different pieces. We're going to put them back together so we can put everything together in the top-level assembly. Let's start assembling the top-level before we get to the rest of the gimbal.
And yes, here's what it looks like so far! It's not looking bad! If we had any better motors and things I'd clean up the geometry to the point where it would look more like the ronin but that would be really hard to print and not as strong so this is what we're settling for.
We actually might change the design of the arm now that we're printing it as two parts. We can make it a lot nicer, sleeker, less boxy. The carrier can stay boxy but why should the other part?
So we made a modification to the carrier and then the arm and together this gives us a much larger range of motion and the space for larger cameras as well. The strange sizing comes from the servos we are using but in the future with better design we can fix this. The next stage is the upper arm. For consistency let's call the camera carrier the carrier, the thing that attaches to the carrier as the lower-arm, and the arm that will hold the lower arm (not designed yet) we'll call the upper arm. We'll start designing that next. There's not that many parts left.
This design doesn't look as good as I'd like to to - mostly because of that servo mount otherwise we could've made it a lot better looking probably. But If it's functional that'll be a step already.
But we're going to look at that later.