Small CAD updates

• Finished up part of the tooling plate
• Drafted the chip evacuation belt
• Worked a little on the frame
  (tired today and got uni coursework so not much stuff done)

Re ran FEA as deflection was outside of design parameters

• The machine is now back to an estimated deflection of 0.182mm under 1KN load on x,y and 500N on z. (this should like never ever happen but i aim for under 0.2mm under those loads)
• Now need to essentially add in all of the “accessory” systems (basically anything which isn’t the core motion system
• Also need to design the 4th axis system

Finished the linear axes!!

• Y and Z axes polished and ready to go!
• Ran the FEA for the machine, and unfortunately found significant torsion under 1KN loads (0.2 mm above my goal of 0.1mm) - I need to review these results with a fresh mind
• Did some research into what fixturing surface to use and found a German supplier selling AW-5083 in the dimensions i require for a reasonable price
• Looked into fixturing options (part clamps) and started considering design options for the rotary 4 axis - should be simple
• Decided on a door design that gets it out of the user’s face in a fairly simple manner
• Also figured out a really good way to evacuate chips and coolant (conveyor belt) which is far better than my previous idea (vacuum extraction) : gotta get around to this in the next design session
  So much more to do, but so much progress!!

DFM for the CNC machine

I have an existing concept design for the machine that i spent about 30 hours on over at fallout. While i am satisfied with the architecture of the design and in theory it could be shipped, i’m not happy with how it goes together and how sloppy the design is, so i’m doing a second pass from the ground up, and designing everything so it’s clean and easy to assemble.
I’m currently working on the motion system and the frame - once i’m done with that i’m going to move onto all of the secondary systems such as the chip collector, electronics bay, doors, and control panel.
I do need to re run the FEA on the motion system once complete as there have been significant changes in the structure (for the better)

LTSpice Design

• I decided to download and learn LTSpice to simulate my inverter design - a push pull drive inverter made from an old power supply. It took me a while to get the center tap drive to work correctly as I modeled the transformer with inductances that were far to small ( i was thinking of SMPS inductors when i picked the simulation values ) when what I have is actually a big iron core transformer with way bigger inductances - which caused issues with the sim. (claude helped me realize my mistake, i was grasping at straws) but after I adressed that, it worked great and i decided to adress the voltage spike and ripple dampening - a 33V TVS diode and an RC snubber are necessary to protect the transistors from the spikes generated by the switching of the transformer (55V to my 60V rated transistor). The simulation performs as expected which i’m quite happy with!

• Claude was previously used to figure out estimated ratings for some obscure old bipolar transistors I have that I am using in the inverter.

• The first two screenshots are the collector voltages without and with RC snubber and TVS diode ( big difference ) the third screenshot is a general analysis on the power levels in the circuit (primary and secondary voltages + current draws)

It’s glorious, it’s dumb, it’s an overengineered water gun built from junk.

• I actually started this early this summer and now decided to submit it to stardance so other people could check it out and possibly build it (and so i can get some funding to upgrade it as i’m a broke student)
  The watergun uses a compressed air tank to provide the energy for the water to be shot and uses a solenoid valve to precisely shoot only a little water (making “bullets”)
  It uses an expresso pump to repressurize the accumulator chamber so that the air chamber doesn’t need replenishing.