another #My-Engines #DIY #EDM #machining #video. Finally he got his pulse EDM working. He had blown the protection diodes and the pulse transistors in the #electronics. His numerous 25Ω power resistors were inadequately heatsinked. He found that long on-times were the key to low electrode wear, backwards from what I would have expected. Sadly he doesn’t say anything about his parameters (voltage, current, frequency, duty cycle) or circuit design. Looks like he’s at about 5kHz. He’s using a diesel fuel pump to circulate his lamp-oil electrolyte (kerosene) through water filters.
on 02025-03-12another #machining #video, by #My-Engines, on die-sink #EDM based on Ben Fleming’s book, using varying sizes of capacitors in an RC setup (16μF, 32μF, 192μF) with the lower capacitances giving a finer finish. He has some detergent (“wetting agent”) in his water electrolyte, presumably an nonionic surfactant (he says his tap water works better than deionized or distilled). However, though big electrodes worked fine, his thin copper and brass sheet electrodes eroded too rapidly, and he wasn’t able to get a stable “burn” out of either them or graphite. Lamp oil worked better than vegetable oil or water for him. #electronics #bootstrapping
on 02025-03-12#machining #video by #AndysMachines on his die-sink #EDM #machining #manufacturing setup. He’s using an aquariump pump to circulate his distilled-water electrolyte through a coffee filter, which needs changing often. He switched to a photo-flash 160μF capacitor for his IGBT-triggered 60V sparks (averaging around 50V). The #electronics use two DRV8825 stepper drivers, though they have trouble with the noise; the die advances when the current averages below an adjustable limit of 2.5 amps, and then there are periodic retractions to permit better flushing. He measures the capacitor voltage with a voltage divider to estimate the current. #bootstrapping
on 02025-03-12#video by #AndysMachines on #EDM #machining, using an RC circuit to limit the energy of each spark and an IGBT to switch it, with 20μs on and 180μs off time (5kHz 10%), 60V, a 100μF electrolytic cap (which gets too hot), a heating element to limit the cap’s inrush current, a 3mm brass electrode cutting aluminum, and just distilled water for the dielectric. He makes it through 1mm of aluminum in 23 seconds, 18mm³/minute. #electronics #manufacturing #bootstrapping
on 02025-03-12#video of die-sink #EDM #machining in diesel fuel using an Arduino, a MOSFET for PWM, a halogen worklight in series to limit current, and a stepper motor. #TheIdahoanShow reports that motor oil, WD-40, and canola oil also work, and aluminum and brass (such as a cartridge case) both work well as electrode materials, but TIG welder electrodes tended to stick and not cut very well. Stainless “kind of worked”. His demo was to cut a plus-sign-shaped hole into a chunk of old file. #bootstrapping #DIY #jugaad #manufacturing
on 02025-03-12#RMRRF DIY #EDM #machining #video #toread
on 02024-05-18#video touting “BAXEDM”’s DIY wire #EDM kit for #manufacturing parts with high #precision; getting his “location” fit (ISO H7/h6) for his M-shaped “shaft” requires +0/-16μm tolerance on the “shaft” and +25/-0μm on the hole, so you need about ±8μm precision. He’s using a 300μm nominal kerf width, shooting for ±10μm precision, using 15mm-thick cold-rolled 304 stainless plate. He got the “shaft” almost 10μm under the nominal dimension.
on 02024-05-17#video of #Rack-Robotics’s DIY wire #EDM #machining rig, including #manufacturing it with #3D-printing from carbon-fiber-filled PLA; the power supply here is their “Powercore V2”, which is 200 watts at 70 Vdc output with “up to” 90% efficiency for US$400. He uses a jug of distilled water from the store. He says the Powercore is “expecting a 1Ω resistance”, so it doesn’t shock him when he puts his hand in the circuit, though I notice he doesn’t connect it across his two hands. He says the wire’s outer coating is zinc, not tin as I had guessed. He says the water circulation is 100 “psi”, and it seems to be moving at a pretty good clip parallel to the wire. He touts ±100μm tolerances on the current revision, which sounds pretty shitty for wire EDM, and for the next revision they’re still only targeting 50μm, so they’re really trying for the rapid prototyping market here, not the traditional ultra-precision market of wire EDM. The motors (“hybrid” steppers) are kept out of the water, using only mechanical linkages to move the toolhead and tension the wire underneath the water. They dunk it in a 10-“gallon” aquarium to hold the water. They’ve moved to Colorado Springs and gotten funded. (This video is from this week.) 180 watts output at 5kHz implies 36 millijoules per spark, which implies about a 14μF output capacitor.
on 02024-05-17#video of #manufacturing a DIY wire #EDM for #machining; basically an ad for #Rack-Robotics’s (rackrobo.io) maybe-someday-open-source Wire Tool replacement hotend to turn any 24-volt 3-D printer into an EDM machine with 300μm brass wire, which looks silver, so maybe it’s tinned. Not sure how the filtration and feedback works, which seems like almost the most important part? He never mentions filtering. Also demonstrates hand-carving a 600μm aluminum plate with a 1.5mm brass rod electrode, which ought to be impossible because it should short out to the workpiece when you’re holding it in your hand, but actually it works fine. Gets up to 100mm/minute with the wire on the 600μm aluminum, or 20mm/minute in 3mm aluminum, while steel is 4½ times slower, while titanium is only 2–2½ times slower than aluminum. The power supply (“Powercore V1”) goes up to 72 volts, but the demo is all at 65 volts; he says the hard part is actually the power supply, which is “high voltage” (ha!), high power, 5000 Hz. This is at #rmrrf which I think is the Rocky Mountain RepRap Fest.
on 02024-05-17