Project Log
so, current projects for stuff I am working, or have built:
QuadRotor / RoboWasp:
* Approx footprint: 1.25 meter (4 feet).
* Approx weight estimate: 15-25 lbs
* Estimated max thrust: 56 lbs (at 480 watt, or 12v 40amp)
* Power Source: Have a 9Ah (24Ahr) LiFePO4 battery.
* Bulk Construction: Wood, Cardboard, Urethane Foam, 3D printed plastic
* Status: pretty early / slow-moving.
** Have some materials;
** Have begun experimenting with 3D printed propellers and gearboxes for it.
* Uncertain:
** Switch from batteries as primary power-source to a small ICE
*** Chainsaw engines are cheapest, but reasonably large/heavy (approx 12 lbs, + fuel).
**** May imply increasing power-profile to compensate (24v 1.0-1.5 kW ?).
**** May need bigger motors/... if going to higher power.
**** Otherwise, need to find a reasonably small/cheap chainsaw engine.
*** An RC engine would put out sufficient power for the current size range.
**** However, RC engines are more expensive than chainsaw engines.
**** RC 4-cycle gas engines are particularly expensive (vs "nitro" engines).
*** Issue is mostly that a 9Ah battery will limit flight times to a matter of minutes.
**** Bigger batteries could also work, but:
**** A large bank of LiFePO4 or LiPo/LiON cells would cost more than using an ICE.
**** NiMH would be a little cheaper, but have a harder time producing large amperage.
***** Would need 3x10 or 4x10 NiMH D-Cells to deliver power, or around $200 worth of batteries.
****** So, around similar cost to using an RC ICE.
**** Should still be cheaper than LiFe or LiPo cells at least.
CNC Milling Machine:
* Outer size estimate: 24x24x40 inch
* Table Size: ~12 inch.
* Power Source: PC / ATX power supply.
* Status:
** Need to build box and physically assemble it.
** Main issue ATM is getting the wood cut up into the appropriate sized panels, and getting it assembled.
** Partial Reason: Need tools to cut and drill the wood (out of scope of small hand-tools and a hacksaw).
** It is outside what is viable to build out of cardboard.
** Spinning motors/etc, not much use without the box and structural parts.
Cardboard Box Robot:
* Size: Approx 20x24x12 inch.
* Weight: about 12 lbs (includes 6lb PbAc battery).
* Approx power: 1/10HP (uses about 100W of power though during operation, and about 9W when idle).
* Power Source: 9Ah 12v PbAc UPS battery.
* Construction: Cardboard, 3D Printed Plastic, Duct Tape and Hot Glue.
* Status: Built / Works
** Has a webcam for remotely seeing from the POV of the robot.
** Have a shell for it, want to paint it to look like a ladybug, but don't have any spray paint.
cardboard box robot was mostly so I could at least have *something*, and to test out some initial technologies for the further-reaching quadrotor project. it is actually the newest of the projects...
for the box-robot, it uses 3D printed gearboxes which reduce output to 400RPM, which drives the wheels.
the quadrotor propellers (early test forms) use 6000RPM output, as simulations implied that 6000RPM with a shallower pitch would be more efficient than 4000RPM with a steeper pitch in terms of producing thrust.
the "new" propellers (still need to 3D print them), will be about 14 inches and have an 8 degree pitch. they should (hopefully) generate some thrust and not break apart.
(ADD / UPDATE: got some lift briefly with the new propellers, but gearboxes soon failed pretty hard as all the gear-teeth turned to dust... some research: this is apparently way too much power to be running through the small contact areas on ABS gears... would at least need nylon, but even then I am apparently pretty near the point where metal is needed...).
had tested with some propeller designs I had got off the internet, but the thrust is nowhere near sufficient (and I reach full voltage across the motors and amps level off, meaning the propellers are not converting enough of the power into thrust, and are basically free-spinning).
sometimes does seem like all this is basically pointlessly frittering time away...
as evidence of how slowly all this is going:
when I started the quadrotor project, the newest Raspberry Pi was the Model B.
now, there is the Raspberry Pi 2 (and the B+ and A+ had been released in the meantime).
so, yes, very slow project...
QuadRotor / RoboWasp:
* Approx footprint: 1.25 meter (4 feet).
* Approx weight estimate: 15-25 lbs
* Estimated max thrust: 56 lbs (at 480 watt, or 12v 40amp)
* Power Source: Have a 9Ah (24Ahr) LiFePO4 battery.
* Bulk Construction: Wood, Cardboard, Urethane Foam, 3D printed plastic
* Status: pretty early / slow-moving.
** Have some materials;
** Have begun experimenting with 3D printed propellers and gearboxes for it.
* Uncertain:
** Switch from batteries as primary power-source to a small ICE
*** Chainsaw engines are cheapest, but reasonably large/heavy (approx 12 lbs, + fuel).
**** May imply increasing power-profile to compensate (24v 1.0-1.5 kW ?).
**** May need bigger motors/... if going to higher power.
**** Otherwise, need to find a reasonably small/cheap chainsaw engine.
*** An RC engine would put out sufficient power for the current size range.
**** However, RC engines are more expensive than chainsaw engines.
**** RC 4-cycle gas engines are particularly expensive (vs "nitro" engines).
*** Issue is mostly that a 9Ah battery will limit flight times to a matter of minutes.
**** Bigger batteries could also work, but:
**** A large bank of LiFePO4 or LiPo/LiON cells would cost more than using an ICE.
**** NiMH would be a little cheaper, but have a harder time producing large amperage.
***** Would need 3x10 or 4x10 NiMH D-Cells to deliver power, or around $200 worth of batteries.
****** So, around similar cost to using an RC ICE.
**** Should still be cheaper than LiFe or LiPo cells at least.
CNC Milling Machine:
* Outer size estimate: 24x24x40 inch
* Table Size: ~12 inch.
* Power Source: PC / ATX power supply.
* Status:
** Need to build box and physically assemble it.
** Main issue ATM is getting the wood cut up into the appropriate sized panels, and getting it assembled.
** Partial Reason: Need tools to cut and drill the wood (out of scope of small hand-tools and a hacksaw).
** It is outside what is viable to build out of cardboard.
** Spinning motors/etc, not much use without the box and structural parts.
Cardboard Box Robot:
* Size: Approx 20x24x12 inch.
* Weight: about 12 lbs (includes 6lb PbAc battery).
* Approx power: 1/10HP (uses about 100W of power though during operation, and about 9W when idle).
* Power Source: 9Ah 12v PbAc UPS battery.
* Construction: Cardboard, 3D Printed Plastic, Duct Tape and Hot Glue.
* Status: Built / Works
** Has a webcam for remotely seeing from the POV of the robot.
** Have a shell for it, want to paint it to look like a ladybug, but don't have any spray paint.
cardboard box robot was mostly so I could at least have *something*, and to test out some initial technologies for the further-reaching quadrotor project. it is actually the newest of the projects...
for the box-robot, it uses 3D printed gearboxes which reduce output to 400RPM, which drives the wheels.
the quadrotor propellers (early test forms) use 6000RPM output, as simulations implied that 6000RPM with a shallower pitch would be more efficient than 4000RPM with a steeper pitch in terms of producing thrust.
the "new" propellers (still need to 3D print them), will be about 14 inches and have an 8 degree pitch. they should (hopefully) generate some thrust and not break apart.
(ADD / UPDATE: got some lift briefly with the new propellers, but gearboxes soon failed pretty hard as all the gear-teeth turned to dust... some research: this is apparently way too much power to be running through the small contact areas on ABS gears... would at least need nylon, but even then I am apparently pretty near the point where metal is needed...).
had tested with some propeller designs I had got off the internet, but the thrust is nowhere near sufficient (and I reach full voltage across the motors and amps level off, meaning the propellers are not converting enough of the power into thrust, and are basically free-spinning).
sometimes does seem like all this is basically pointlessly frittering time away...
as evidence of how slowly all this is going:
when I started the quadrotor project, the newest Raspberry Pi was the Model B.
now, there is the Raspberry Pi 2 (and the B+ and A+ had been released in the meantime).
so, yes, very slow project...
