Dyne's LD-F1 (D-O concept) build

[kresty]

Looking good! I have a 12V power supply in L3-G0 so he can "plug in". (Not sure what I'll do when I switch him to 18V).

[Dyne]

Looking good!  I have a 12V power supply in L3-G0 so he can "plug in".  (Not sure what I'll do when I switch him to 18V).

Thank you.

I'd love to have enough room in this droid to fit a power supply (though I'd likely use that space for a sound system). As it is I'd be lucky to fit a smallish wall wart inside.

[Dyne]

Since I can't attend Droid Con 2023 this coming weekend and don't want to be sad about that, I thought I'd post a bit about what I've been doing.


Firstly, I mentioned awhile back that I would release information on how to build the droid.  I want to do that before making major untested revisions like the static neck, so I've started putting together a snapshot of the droid's files and build info as it currently exists.  It's slow going because there are a few hundred parts and I'm trying to write assembly instructions as well as a bill-of-materials, which means I either need to remember what I used/did, or else find that info in what I wrote here (if I even did).

As part of the instructions, I want to include some nice exploded views of the various components.  I have been reassembling the droid in Blender so I don't have to rely on screenshots. I don't think the second image below is especially clear because it's so busy, but that'll improve when I render individual sections of the droid.  In the meantime, it gives some idea of how much stuff there is in here.





No clue when I'll finish that.



Secondly, I'm starting to think about my plans for rewiring the droid.  I don't think I ever actually posted the quick diagram that represents my initial "plan" from back before Dragon Con.  If you can call it that.



I mentioned this briefly back when I wrote post #37, saying I'd post a better version later when I had time to create one, but I was definitely in crunch mode and never quite got around to it.  (Just as a hint of how rushed this was, it looks like I specified using a door where the main power switch should go.)

I do know a few alterations that were made when I actually implemented this.  For one thing, the ESP32 and its logic level converter were put inside the head.  That means they draw their power and ground connections from the PCA9685 board (also in the head), not from the 5V battery elimination circuit on the two ESCs (which are in the body).  Instead, both ESCs connect to the RX8R receiver (with the 5V BEC lead on one of them disconnected), and there is no 5V bar ... the voltage regulator output only goes to the PCA9685.

This schematic will likely change quite a bit as I revise LD-F1.

[savagecreature]

Beautiful CAD renderings.
As always, thank you sharing them

[Jaffa007]

Exploded view’s always look great


Sent from my iPhone using Tapatalk

[Dyne]

Thank you.  Hopefully the renders will prove useful.


I took some time to revise the power plan to something that's closer to what's actually going on.  Firstly, I took the main power door out and replaced it with a main power switch.  Then I cleaned up the connections shown between the head and body.  

One difference from current reality is that I went ahead and removed the fuse block on the diagram, replacing the fuse holding function with my planned inline fuse holders, and the connections to the volt meter and fuses now going through a new 12V Wago connector.  I also replaced the existing Ground Bar with a planned Wago connector and removed the non-existent 5V bar.  Electrically, this is an identical setup to what the droid currently has, just with different physical components.

I added an ESP32 and logic converter pinout as well.  The pinout just makes it easier for me to see where the connections actually are supposed to be on the board. It's nothing that wasn't already shown on the plan, aside from now having the correct wire colors in that area and having the converter's ground connection with the PCA rather than with the ESP32.





LD-F1's head has now been detached in order to start cleaning up and securing his head wiring better.    I'll be taking the time to hot glue the connectors in place on the PCB pins, add clips to keep the wires from snagging, and mount the ESP32 more securely.  This will hopefully resolve his LED reboot issues.

That also let me investigate the source of the head slipping.  The relevant parts are shown in this image from post #17.



The cause is pretty much what I suspected: the Rotation Platform (yellow) is not rigid enough.  As the droid moves, the part flexes between the rotation servo and the Neck Core (orange) just enough to allow the gears to slip out of mesh briefly.  The noise dampening inserts also contribute.  I doubt I can make the platform thicker, and I suspect it was already printed solid, so I may have to get a metal version of that part made if removing the inserts doesn't help.

I also want to prevent the Tilt Servo Brackets (Red) from splaying or bending under the weight of the head, which I might be able to do by adding a brace across the front to tie them together, but if not they might need to be metal as well.

I put a bit of velcro around the Tilt Arm (blue part, in the sloped section above the tan part) to hold the tilt servo wires up out of the gearing.


Having finally gotten annoyed with one end of the coiled wire under the head constantly coming loose (it was originally hotglued in place), I finally drilled a hole in the rear underside of the large section of the head to feed it through

[Dyne]

I'm still working on compiling the files.  As I said previously, the initial files will mostly be LD-F1 as he currently exists, before the big changes I'm making. However, there are a few tweaks.

For example, back in post #12, I showed the Voltage Regulator that hides under the tracks. When I added this, I had to drill holes in the Rear Outer Wall to route the wires into the chassis because I'd printed the part before I realized I hadn't added one to the model. I've since done that.

There are also some new parts I haven't used yet, like an optional tube that the Front Axle can slip through inside the Electronics Box when the axle is installed. The purpose of this is simply to make sure the metal axle isn't able to short any wiring if something comes loose.



Here's another new file.  Back in post #60 I showed a pic of the inside of the head and a servo meant to raise and lower the antenna.  The idea I mentioned for that mechanism didn't really work out, so instead I made a little mount that rotates the servo 90 degrees.



This puts the servo movement in the same plane as the antenna's motion, which should make the hypothetical actuating mechanism a lot easier (a rack and pinion, or just a linkage of some sort).


Recently I was handed something that made me wish there was slightly more space available in the droid: A Toyota 12V window regulator motor.  This is strong enough to easily be able to lift the head even without additional gearing (after all, it can lift an entire auto window), without also being clearly impossible to fit inside the droid's central body.



Add a spool of wire or fishing line to the motor's output, attached to the back of the rear shell pivot or the back of the lower shaft hubs (so winding the spool pulls the line to lift the stalks), then use a motor driver like the BTS7960 and a pair of limit switches.  Possibly controlled by ESP32, possibly a simple setup with the switches and a pair of diodes wired up something like what's shown around 4:25 in Michael Rechtin's 3D Printed linear actuator video, which I linked to back at the end of post #18.

I don't think I'll do this as I still likely can't fit it inside (at least, not without removing the lower stalks, which would make it pointless).  But it's definitely more plausible than what I was trying before.


Edit: On the topic of the head losing its center of rotation, I think that by removing the vibration dampening inserts from the Head Rotation Servo and using M4 screws instead of M3, I've tightened things down enough that it will have a much harder time skipping. I won't know for sure until I get around to putting the head back on.

While I was putting the new screws in, I noticed that the two screws that secure the forward end of the Starboard Tilt Servo to its bracket had lost their nuts and backed almost completely out, no longer securing the servo to anything. Most likely this was because the screws were a bit too short for lock nuts and they vibrated loose and fell out. I've replaced them with longer screws and put my last M3 lock nut on one of those. I'll put lock nuts on the others when I get some more.

I made a minor change yesterday. I dremeled off the excess length of the Rear Axle. It's always been a bit too long, which interferes with the Track Covers.

I've also realized that without the head attached to the rear shell, it would be much, much easier to motorize the opening. Which means I could potentially put something fun in there after all.

[Dyne]

I have a working (if ugly and somewhat overly complicated) antenna mechanism.

First, I reversed the servo mount mentioned last time (orange) to move the servo further to the starboard side of the head, since I needed a bit more space between the servo rotation plane and the antenna.  I didn't need quite as much space as that mounting position provides, but if I moved it less then I'd completely block the screw holes that attach the mount to the head.  

Instead I made the hub of the printed servo horn (yellow) taller to make up the difference.  At the end of the servo horn is a small standoff sitting against a 623zz bearing.


Surrounding the bearing is a bearing retainer (red).  This is basically a slotted track that permits the bearing to slide forward and back.  This is necessary because the servo's rotation carries it toward/away from the antenna (black).  The retainer also has a small protrusion that is screwed onto the end of the antenna mast.


The result is that rotating the servo moves the retainer up/down, and the antenna goes with it.  It's not a huge range of motion -- a maximum of 34 millimeters, which is the diameter of the arc described by the center of the bearing -- but it's better than nothing.




Some of the wobble you see there is because there's a little bit of interference near the top.  I think there is slightly less clearance between the antenna and the head geometry up there.  I was seeing similar interference in the same place on the previous prototype.  The rest of the wobble is because the Antenna is a bit loose in the hole in the head.  I can't do much to stabilize that on the inside of the head -- as you can see in the video, the top of the servo range puts the retainer very close to the part of the head that the antenna goes through.

Instead, I have cut a small piece of styrene tubing that has an ID that more closely matches the antenna mast's OD, and glued it to the outside of the hole in the head.  I might make a shim as well.
 

I could revise the hole diameter in the head model, but may not ... it could cause issues if the exact tubing isn't available, or the hole doesn't print as cleanly (it's horizontal in the normal print orientation).  


In any case, with the antenna working, I'm considering putting the top door on the head (blue) and making that functional as well, mostly just to give LD-F1 a little more life.  I would need to make some sort of hinge (probably involves cutting a few holes into the head) and come up with a mechanism to attach the servo to the door.  Fortunately, the servo mount that I built in for this purpose is a lot more conveniently oriented than the built-in antenna servo mount was.



In the meantime, I've redone the receiver-to-servo control lines.  This involved a lot of faffing about with JST SM connector crimping and rehousing, which is always a bit of a pain.   Part of the goal was to remove the loop of extra green, yellow, and white wire in the head (which you can almost see as a background blur on the left of the Antenna closeup photo), but mostly it was to add new lines (red and blue) to support these new servos, as shown on this updated power plan.  There's also a new wire for future use (I have long had something in mind that I could put in the side bay instead of those vents, but I'm not ready to commit yet.)


BTW, I swapped the Left and Right Tilt Servo on this diagram compared to earlier versions, in order to match the actual wire color order.  I also added black rectangles indicating JST connectors for the servo signal wires and the LED sequin.


Incidentally, we've now passed the anniversary of what I consider the true start of this build (back in Post #13 on June 15th, 2022, one year ago yesterday. This post was started and would've been made on the actual anniversary, but I needed to get more screenshots and pictures.)

Almost everything prior to June 15th consisted of 3d sketches, concept work, and models that ultimately were rebuilt from scratch (body shells) or weren't used at all (flexible tracks).

[kresty]

I wonder how a power car antenna does it?

[Dyne]

I wonder how a power car antenna does it?

I believe they have a spool that works somewhat like a measuring tape (i.e. flexible enough to coil, rigid enough to extend the antenna without collapsing when unspooled).

My friend suggested something like that, but I don't think I have anything that'd work as the tape (except maybe filament, and even then it'd probably require the spool to have a large diameter or it'd break unless it's nylon). It probably also relies on the tube having a large enough ID for said tape to run all the way to the top, though, and this tubing has a pretty narrow hole, almost too small for the screw attaching it to the retainer.