Ghostpilots Aluminium Delta Throttle build-log

[Ghostpilot]

Every journey has a first step.
Inspired by both Jeruw's and SC-Maik's alu-builds, I set out to make my own.
My first step was to make some bearing blocks for some bearings I had in my parts box. I chose to base them around the general shape of the Actobotics bearing blocks with two slightly raised legs.

[aniron]

A good start!

Did you use the general dimensions of the actobotics parts as well?
I have a .scad of a base plate i modelled after SC-Maik's files if you are interested.
Might be easier to modify to your dimensions,

[Ghostpilot]

The shape I drew on my own with metric offsets and radii, but the c/c (center to center) distance I took from The Acrobatics specs. Apart from the custom bore size and height from the bottom surface that brings, the footprint should be a drop in replacement for an Acrobatics bearing block. It's metric threaded so the size of the mounting screw holes in the base have to be adjusted accordingly. I choose M4 threads, but M3 would hold too. I wonder if I should put a set screw from the bottom into the bearing bore, to be sure it won't wiggle out over time. How tight fit is an Acrobatics one? Can you easily take out the bearing or is it press fitted in there?

[aniron]

I have not specifically tried to remove the actobotics bearings, but they appear quite solidly seated.

A set screw might be good if yours are a looser fit.

I'll put up a post tonight (hopefully) in my own build log with the .scad files. You might need to modify it for your own sensor assembly (mine is a tube clamp with the same footprint as the bearing blocks, with enough separation to fit an axial connector between the sensor d- shaft and the .75" tube.).

Seeing that you are in the cnc business, do you have a ballpark what it would cost to get a baseplate machined in (thick enough) aluminium?
I suppose spring anchors would need to be separate parts as opposed to the 3d printed version, to make it a 2d shape.

[Ghostpilot]

I may work as a CNC machinist, but the workshop I work in belongs to the company I work at, so all orders are internal work orders.
I rarely if ever handle orders where I have to calculate a price on the outcome. However, I could look into the material cost for such a base plate.
Define "thick enough". If I recall correctly, the current baseplate thickness on both the printed and aluminium ones (to date at least) seems to be 1/4" or 6,35mm. That is thick enough to countersink metric M4 screw heads to be flush with the bottom surface without compromising strength. If it's weight you are after, it would weigh more if the thickness were bigger of course. However, it doesn't have to be made in one piece. A 6,35mm base plate could be mounted into a thicker frame same as Jeruw's build to add weight and stability.
Spring anchors could be screwed down from the bottom, so it will still look like (well mostly) it is one piece.

The anchor points for the Z home position spring (I'm sure it could have a shorter name), do they need to be able to swivel (spin around)? If so, it could be done by simply countersink a hole from below in the base and make a plug that looks like a pin with a hole through from the side (like the eye a sewing needle) and a head like a normal screw. If inserted into the countersunk hole from below it will be able to spin around its own axis but will not be able to go through the hole. I can make a drawing later to explain what I mean.

[aniron]

Okay, I see. I'll just get a quote somewhere locally. If you are checking the ballpark cost anyway for your own information, I'd still like to know.

I have simply assumed that the Al plate could be made thinner without losing too much rigidity,
and was planning on possibly having it made together with some parts for my 3D printer (5mm thickness, replacing lasercut acrylic)
if there would be enough gain in making a larger order (that would still fit the machine in one go) to come out cheaper than having
the throttle parts made thinner and ordering printer parts elsewhere from someone making batches of them.

I'm not extremely concerned with the looks of the lower throttle mechanics, as my ultimate plan is to make them innards under a faceplate.


I was considering a mechanism similar to that on the MFG crosswind pedals that would allow for the spring to be adjusted.
(threaded rod with hole through one end, spring fits through hole. Other end (through hole) with nuts for adjusting the tension.) That would also provide a swivel point.
On second read, I think this is what you meant?

[Ghostpilot]

Here is what I thought could be used as a simple rotating Z-spring anchor. Simplest possible.

[Ghostpilot]

I have simply assumed that the Al plate could be made thinner without losing too much rigidity,
and was planning on possibly having it made together with some parts for my 3D printer (5mm thickness, replacing lasercut acrylic)
if there would be enough gain in making a larger order (that would still fit the machine in one go) to come out cheaper than having
the throttle parts made thinner and ordering printer parts elsewhere from someone making batches of them.

If you got aluminium parts laser cut from big plate material it would make sense making several parts together.
If using a CNC mill, the parts will most likely be made one by one. I usually don't gang-machine different parts in one go in the CNC mill. If I would make a program where several different parts are machined at the same time, it will be difficult to make just one of those parts again because they would be all programmed together as one unit. Another argument is that it would be extra work to gang machine parts if one or more parts are repeated within the group more than once. If made one by one I can just repeat the program for any part I need more of. Even if I would machine parts from plate material on a vacuum table I still wouldn't program them to be gang-machined. Think of it as making gingerbread cookies from a big dough, there has to be material between the parts or they will slip from the vacuum table when cut free. That limitation does not apply to laser cutting (plasma & waterjet too) where parts can share an edge that will only be cut once and still divide the two parts.

To answer the first question last, yes you can make the parts much thinner if using aluminium over printed plastic. However, it will be much more pleasing to the eye if the thickness of the alu parts is enough to countersink all screw heads to be flush with the surface.

[Ghostpilot]

I should have contained my impatience, but alas I went ahead and made the slightly older base plate with slitted sensor mounting holes, instead of using the newer version with only round mounting holes.
I wanted to see some progress I guess. Just waiting for springs to arrive while doing nothing didn't seem like wisely spent time. Perhaps I can make a plug to cover up the visible portion of the slits later.
I haven't made any PCBs nor PCB holders yet, so I just put the first three bearings blocks (which had faults) as place holders for now. That's why there are only bearing blocks on there right now.

The only new feature I added to the base is the central hole countersunk from below where my rotating Z-spring anchor will sit.
I will make it spin on small thrust bearings I found in my parts drawer, just because I can. It's totally unnecessary but I really wanted to use them for something.


Thrust bearings. I think I salvaged them from a disk drive.

[Ghostpilot]

The idea about using a thrust bearing to make the z-spring anchor swivel motion frictionless was more a "because I can and already have the parts"-kind of thing.
Now when it's finished I am very pleased with the outcome.

Here are the parts. The centre hole is countersunk from below and a tiny ball-nose mill was used to make a track for the bearings to roll in. The z-anchor itself is a puzzle piece mirror of the hole and it also have the ball-nose milled track for the bearings. Last but not least the thrust bearing itself.


Here I have mounted the thrust bearing on the z-anchor part. Those two together are then inserted into the centre hole in the base plate from below.


When the z-anchor assembly is inserted into the base plate properly, I just had to fasten the z-spring on the top side. When the spring is in place the z-anchor won't fall out because the spring is holding it from the top. The swivel action is so smooth that I can hold the z-spring and set the base plate spinning in place on the z-anchor alone, because of the thrust bearing. If I put a similar spring anchor on the top plate, the z-spring will be able to spin freely regardless of how the top plate is tilted... I think.