Aluminum build v2

[jeruw]

I'm working on some redesigns for a new build in aluminum. As part of this, I've decided to try learning Autodesk Fusion 360. I have some limited experience in parametric 3d modeling from a previous job that touched on ProEngineer a bit. I've started re-designing some of my elements in Fusion this weekend and started to do test assemblies. This is so much nicer and so much more accurate than my previous Inkscape drawings! Being able to put in actual measurements and constraints on geometry is huge. And it's awfully nice to be able to assemble the parts and see how they fit together. Here's a render of a test fitting I did to see if I could get away with making all of the uprights the same dimensions. It looks a bit crowded to me, so I'll probably make the outer uprights (non-sensor side) smaller on this version too, or rework the base to be a little larger.

[jeruw]

Some additional renderings. I designed a smaller secondary upright as well as the shaft collar with 2 neodynium magnets to drive a hall effect sensor. I drew up a couple of the flanged bronze sleeve bearings as well and placed them in the uprights for the shaft to ride on.

I already uncovered an issue thanks to drawing this up and assembling it in Fusion 360. I thought I was going to be good to go with a straight shaft, a shaft collar that doubles as the magnet mount, and locking down the pivot arm near the secondary upright with a set screw. Sure, the shaft collar and pivot arm will prevent the shaft from sliding out laterally to the right, but I have nothing to prevent it sliding to the left. Well, that's not entirely true, the spring tension will probably keep it from sliding too far to the left, but I'm not sure that alone will be enough. I may need to extend the shaft at least 1/8" to the right as well and design a second collar.

Next up will be redesigning the pivot arms, spring retainers and handle base plate.

delta_assembly_2015-Mar-24_06-14-26PM-000_TOP.jpg (45.34 KiB) Viewed 1616 times

delta_assembly_2015-Mar-24_06-14-26PM-000_FRONT.jpg (29.63 KiB) Viewed 1616 times

test.png (119.57 KiB) Viewed 1616 times

[zdayton]

Looks good. Retaining rings like this might be useful. http://www.mcmaster.com/#retaining-rings/=wgit8p. Put them on the inside of the bushing. I would probably also plan on having both bushings facing in.

I think you should try to make the two bottom 'fins' the same part, or maybe almost the same part but one has a few extra operations, that would be easier from a manufacturing perspective.

[jeruw]

Thanks for the idea on the retaining clips. That's a good call. I'll get checked out on the lathe and that should make cutting the groove pretty easy. I'd thought about actually trying to turn the shafts in the lathe to take a piece of 1/2" round bar down to the spring's spec and then step it down to fit a 1/4" ID bearing, but that's a fair amount more manual labor and real bearings would add a bit to the build price. Maybe though.

You're right on the fins. Having one part to manage would be easier. I'll expand the base a bit to give a little more room and re-work the fin design slightly. I'll probably make the back of the fin flat to make it easier to clean up the tool marks. I don't have access to a spindle sander right now, so getting at a curve like that is kind of a pain. I should probably figure out how to do finishing passes during CNC .

[jeruw]

I also need to test to see if my hall effect sensor idea is actually going to work. I realized yesterday when I reviewed the blog post I'd referenced that the sensor was mounted in the center of rotation of the magnets. I'd intended to mount the sensor beneath the shaft, so the shaft collar would rotate the magnets near the sensor, but not around it. I have no idea if that's going to work now. I don't know enough about their function to say for certain, but I sort of assume that if I go from a 2 magnet set up down to 1 magnet that I should be able to get the 90 degrees of movement I need. As the magnet moves close to the sensor and then back away. The sensors will be here today so I can start testing.

[InvalidName]

This looks great. I think I will have to try out autodesk 123d as sketchup isn't up to task.

[jeruw]

This looks great. I think I will have to try out autodesk 123d as sketchup isn't up to task.

Grab a copy of Fusion 360. They're offering free licenses to hobbyists and 'startups'. It's pretty easy.

My only 'complaint' about the program thus far is it doesn't have real assemblies. I expected assemblies to work against the versions of parts, maintaining a sort of BOM, with the ability to update a placed part to a new version and bring in changed geometry. Unfortunately importing a part into another design to build an assembly actually creates a copy of that part unique to the new design, disconnecting it from it's original design. So if I redesign the base plate or fins in their original unique designs I need to start over with assembling them. Oh well. They have to reserve something for the people willing to pay for Inventor I guess.

[jeruw]

The hall sensors and magnets showed up last week. I also got in to the shop and cut a new set of fins. I was having some issues with the Shapeoko so it was slow going. These were the first parts I did the CAM in Fusion 360 for, and that was a bit of a learning curve. Crashed the tool a couple times and ruined a few bits before I realized I had to trim the gcode a bit. I'm still tweaking the base plate but hope to cut that this week to do some test fitting. I also need to cut a couple of shaft collars so I can do some testing with the hall sensors.

I've hooked them up and took some sample code to test the readouts. It's not entirely clear what the range of values will be in operation till they're mounted and the magnets are being rotated on an axis instead of just waved in the air nearby.

[zdayton]

I also just ordered some sensors. They are magneto resistive angle sensors, which I think is better than a Hall effect sensor for this application. They should be accurate to about 0.4 degrees after calibration.

[jeruw]

Which sensor did you get? I ended up getting some Allegro a1302 sensors for testing, they were cheap and easy to hook up (didn't need a more advanced multi-pin connection and breakout boards that I wasn't able to source affordably). We'll see how well they work.

http://www.allegromicro.com/en/Products ... 301-2.aspx

I'd be curious to know what you settled on for testing and what your plans are.