Vulcan Open Lab

From Adam Colton
Jump to navigation Jump to search

This is really a spin off of the 1K Workshop

https://www.youtube.com/watch?v=7zcnFeZ2tiE

Research complaint mechanisms (particularly in conjunction with 3d printing)

Early on, hot glue can be use very effectively. The first 3D printer can be hot glued together just long enough to print proper mounts.


Modules

  • Hand Tools: Nothing to learn, just buy these.
  • Basic Programming: No cost, useful.
  • Basic Math: Also no cost and useful. Can focus on problem solving and use the computer programming section.
  • Basic Electronics: Focus on theory and spice
  • Power Electronics: linear and switching regulators and power consumption models
  • Electronics Lab 1
  • Digital Electronics: Intro to Arduino
  • Electronics Lab 2
  • CNC Lab 1
  • Foundry Lab 1

Timeline

I won't start this project in earnest until 2024ish. From now until then, I will continue to try and build the base knowledge needed for this. I would really like to have experience with metal casting, 3d printing and cnc milling by that point.

  1. Research: Just coarse research questions necessary to get started
  2. Alpha prototype: Try to build the end state, no restriction on tools used or cost.
  3. Research: Any issues that came up during the alpha build
  4. Beta and Bootstrap: Plan the beta build and a bootstrapping sequence to get there. I can still skip some steps that will take a long time so long as I can prove them out.
  5. Final: Do the full build and all bootstrapping

Electronics Lab 1

Basic Power Supply

Just a buck-boost module with a short circuit protection. I'm not sure if a relay, BJT or MOSFET makes more sense. A relay is simple, but potentially slow. BJT is also pretty simple, the fact it will consume a bit of power bothers me, but the cut off will generally be 0.5-1A so it will consume less than 300 mA. A MOSFET would be best, but it's a pain to drive. Can use P-channel and just require that the input voltage is >10V or use a separate buck module to get an N-channel to 10V above the load.

Soldering Iron

I think the best technique here is to add precise controls to a basic model. This version will just have a Schmitt trigger, but it should be built in a way that will make it easy to add digital control later.

Helping Hands

Kind of a pain because it's a building project not an electronics project. Recommend a few ways depending on what they want to invest in or have access to:

  • Full build: Either drill the holes or use a rotary tool to cut and sand them. Need a tap set.
  • Board build: just use a wood board and cut or drill the holes and either hot glue or epoxy the pipes in place

Electronics Lab 2

Add digital control to the soldering iron.

Build a full bench power supply. Probably built around a variable transformer connected to a buck-boost with a final linear stage.

I'd also like to start on the lab computer. Need access to the GPIOs. It could be a RasPi, but it might make just as much sense to build a really basic desktop style computer. Eventually this will run everything in the shop, but for the first stage, it should handle modeling software and get used as oscilloscope and multi-meter. The multi-meter should be able to handle multiple probes simultaneously. It should also be able to do part identification.

CNC Lab 1

Build a small 3-axis CNC machine. Goal is to be able to do ABS-3D printing and be able to (very slowly) cut Aluminum.

Mini-drill

This should probably be it's own module. Build a very basic drill press to allow accurate holes.

My initial thought was for this to be built from really simple parts, but it may make sense to build this from high quality parts that can later be recovered. For instance, linear bears and ball-screw threads that can later be used on the 3D printer. This could allow for enough precision control to accurately tap holes.

Foundry Lab 1

Build an 15A/110V aluminum foundry. I'd like to be able to get above 2000°F so it could melt copper.

I really like the lost foam casting results from Makercise. I'd be interested to see if we could use the soldering iron with a long thin tip in the CNC machine to automate cutting the foam.

CNC Lab 2

These two components will share NEMA 17/23(?) steppers and lead screws. But ultimately I think it makes sense to build two separate frames. The print head should be light and move quickly to cut down on print times. The CNC should be reinforced to support side load.

3D Printer

I'm really curious if I can build the bulk of this out of MDF. I think I can make the table move smoothly with a sort of drawer-guide design built around the "railroad-wheel" design. With the wheels and rails both made of MDF. The wheels should only need to be sloped at ~1°. But this requires testing.

Like many of the parts, I can come back and build an aluminum frame from cast parts.

I really like the printers that have a belt for one axis. This allows for much longer pieces as well as queuing up a bunch of prints.

CNC Mill

I might go with the "mostly printed CNC" build. Is it worth polishing up the conduit? If so, set up a simple motor powered hacksaw first. Conduit reamer to clean up cuts?

Another nice-to-have that could be done before this would be a small electric screwdriver. This would not just be a convenience, but it could also be torque limiting which will help avoid cracking 3d parts.

I'd like for one of the early upgrades to be a 5 axis mill. This would allow us to make our own lead and ball screws from aluminum.

In this video he makes an end mill from a sewing needle. That could be a handy trick, particularly for PCBs.

Chips and dust need to be taken into consideration. I think this should have two modes. A vacuum to collect dust - the basic "brush" setup. And a coolant system that will filter out chips for cutting metal or glass.

Beyond

From here, the 3D printer can make any plastic parts that are needed. Aluminum parts will be made in 2 stages; a rough is 3D printed, then cast in aluminum. Then the aluminum part is milled to precision.

I'm not sure the direction to go. I'd kind of like to do a traditional set of power tools, but I don't think that actually makes the most sense.

The goal is a large volume, 5-axis CNC machine with reasonable precision and to improve the small CNC machine to be able to handle extreme precision.

There's also some intermediate things that I'll want:

  • ABS recycling
    • beefy shredder
    • extruder (also good for ABS pellets)
  • Plaster recycling; just need a tumbler and we should be able to use the foundry at low power to fully dry out the powder
  • Air filtration
  • Dust collection

What is the best material for casting? 3D printer filament can be a bit expensive. Can we print with wax? Is foam insulation cheaper? What holds better detail? Can I accurately combine them to create the bulk of a mold from foam, then add details in ABS?

Small CNC Upgrades

Enclosure with temperature and humidity control and air filtration.

Small lathe so it can be use for precision lathe work. I should also be able to fit a grinding wheel into the lathe and grind precision cutting heads.

Cut some acme threaded rods.

I'm not sure what the right move is for the motor. I think some combination of precision movement with feedback will be right. So the motor will be either a stepper or the pneumatic equivalent. I need a low backlash reduction. Then on the ends of the shafts, a large wheel with an optical sensor with very high DPI.

I don't much about 3D printers, so I'm not sure what print head to start with and what upgrades can be made, but that's an area to research.

Same thing for the small router cutting head.

Research

Tool Metal

I think we should be able to use the foundry to soften steel, cut or grind it, then harden and temper it.

It may also be possible to use electrolysis to deposit tough metals like tungsten on the surface to protect tools.

CNC Circuits

I like the idea of using CNC to place copper wire or thin copper strips to layout circuits. I think this could be done in combination with either 3D printing or a thin adhesive to create complex, multi-layer designs.

This could also be done by making sheets of copper foil, gluing them to a substrate, cutting away the unnecessary parts, then printing the next layer of substrate. Connections between layers becomes a challenge. The surface would need to be thoroughly cleaned. I'm not sure if metal-to-metal contact is enough or if we need some sort of solder. It would be complex, but it may also be possible to create connections between layers by using a continuous sheet of foil; glue and cut one layer, put down the substrate for the next layer, lay down the adhesive then fold the next layer of copper back.

If the substrate can withstand the heat of soldering, the whole thing could be integrated with a pick and place machine.

Machine Tolerance

Builds for the Gingery machines seem to take a long time. When watching videos, a lot of this seems to be in getting the necessary tolerances. What is the fastest way to achieve this?

I also believe there is a trade off to be had. A looser tolerance early on may allow to progress more rapidly, which unlocks access to equipment that can be used to efficiently retro fit a better tolerance.

Upgradeable Machine Bed

Leave the first sections the same; electronics, then 3d printing, then foundry.

From there we build a universal machine bed. A combination of a lathe and a mill bed. It has the most movement along the y axis. What would be the cross feed on a lathe serves as the x direction. But like a lathe, it's angle can be set.

The machine can be setup as a lathe by attaching a head and tail stock to the ways. There is an overhead mount for a Z axis that can be used as a mill.

The motor can also drive a table saw under the ways. It's use is sort of a power feed table saw sled. This is ideal from a safety stand point because the users hands are never near the blade and they can stand so that kickback won't hit them.

The motor can also drive a bandsaw wheel with the opposing wheel mounted on the z axis. Again, this is used with a sort of sled power feed.

I think this replicates most of the operations one would want to do. A mill can work as a drill press. The table saw can make miter cuts easily, the cross feed angle can be used to accurately set the angle of the cut. The mill can replace a jointer and planer (though it may be a little slower).

It may be possible to permanently mount the motor under the ways. It would be easy to connect to the headstock with belts. It would be in the right position to drive a bandsaw wheel. It would be 90° off from the table saw arbor, but either the motor could rotate or the arbor could include bevel gears. To deliver power to the Z axis, I think I should incorporate a vertical shaft. This complicates the build, but re-mounting the motor would be a pain.

Materials

MDF / Hardboard

This is cheap, easy to work with handtools and dimensionally stable. It will be used extensively early on.

Foam Insulation

Also cheap and easy to work. Good for volumetric work, particularly things that will be cast.

3D Printer Plastics

Obviously requires a 3D printer, but good for a lot of parts and for making castings.

Aluminum

I think the plan is to get a foundry up and running and make many of the later parts out of aluminum. Large quantities of aluminum scrap can be bought quite cheap.

I need to understand the alloys though. Is there a feasible way to sort scrap into reasonable quality alloys?

Epoxy

This is very strong, there are resources on building the body of machines (mills and lathes) out of this. I can use MDF for casting.

Proper use does require vibration. A vacuum pump may help, as might a torch.

Components

Precision Motors

There are 3 contenders here

  • Stepper : Cheap and easily available, can be used close loop or open loop
  • Brushed : Very cheap and fairly powerful, these are drill motors. Electrically noisy, requires closed loop control and they are hard to control precisely.
  • BLDC : More expensive, but powerful, fast and can be controlled very precisely. Does require and ESC to do that.

For closed loop feed back it will be either optical or magnetic.

Power Motor

A roughly 2hp motor.

Retrieved from "https://wiki.adamcolton.net/index.php?title=Vulcan_Open_Lab&oldid=6593"