I'm so happy you're continuing work on this project! I just had this idea about a week ago and yours seem to be the only videos on RU-vid that are making headway in developing this tech. I think there are huge possibilities here and I've begun to work on my own system. Thank you for the time you've put in on this project. The quality of your results speak for themselves!
If you're interested in doing the microfluidic circuits, I would recommend cutting steel and using NaClO3 as your electrolyte. The finish is the best I've seen, bar none. There's also a few papers out there on how to configure your current density for better surface finishes. It's funny actually, I've been trying to replicate your results and I'm only just now seeing your recommendations for a the pump and voltage booster after using the exact some ones in my setup. Great minds think alike I suppose. I've been trying on and off to get away with using a 3DP nozzle as my jet fixture and it's just totally trash at maintaining a good quality free jet. Gonna have to switch back to needles again.
I'm happy I stumbled upon your videos, my project for next year is to mod the remains of my old printer to an ECM machine, one that can achieve high resolution with reasonably deep cuts. There are extremely thin hypodermic needles I want to try (if you ever get a spinal tap you'll appreciate they exist, lol) but I can already imagine the challenges that come with them in regards to max current and flow. So much to try, error and learn, very exciting!
Oh my god! Are you the son of the guy from Applied Science? You talk like him! Even your videos end like his! New favorite channel! Kudos buddy for this amazing work, I'll be soon trying this too!
You can improve z resolution by switching to 2mm pitch leadscrews which gives you .01mm resolution on full steps, and even less when doing microstepping. If you don't believe in microstepping, gearing z might get you even higher resolution.
Subbed, liked, and comments for the algos. Didn't have high expectations with such a low view/ sub count. Pleasantly surprised, especially with your python script. I'm going to try this on my lowrider cnc to cut out aluminum replacement end plates and start working on rigidity improvements.
Great stuff!!! One thing I would like addressed further is the role the NaCl in solution, which dissassociates into Na+ and Cl- when dissovled in water, playes in the ECM process. Thanks!
Sorry if this sounds dumb, but the only reason to use a pump is to blow away the residue, isn't it? Otherwise just for a shallow engraving/etching a solid needle and a fully submerged piece + needle tip should be enough and simpler than adding a pump, and will perhaps allow for finer details
Have you tried a lithographic/photo-lithographic method? I used a crude form (a pattern cut from tape) to make a spring steel reed valve with a complex shape that would have been very hard to make any other way.
So this will work on copper. This can be used for PCB milling in this case. No more broken mills :-) would be great. You Eagle CAD for this. Can you show more of your setup.
a good video but, at 5:13 it's complete nonsense in explaining the need to supply current (having lessons of oxydo-reduction chemistry, I would eplain it like this) by applying current, you force down electrons in the cathode and pumps out the same amount from the cathode ; the overall system stays electrically neutral the reaction happens in two times : 1) a) at the anode the aluminium is oxyded ( anode => oxydation) Al(solid, noted s)=>Al3+(in the water, noted later aq) +3 e- b) at the cathode the water gets reduced, in dihydrogen (H2) producing HO- ions : 2*H2O(l, liquid) +2e- => 2HO-(aq) +2H+(aq) +2e- => 2HO-(aq) + H2(g,gazeous) to ensure a global electrical neutrality, the reaction at the anode happens two times whereas the one at the cathode happens three times(in the same amount of time) : giving us the final equation : 2*(Al(s))+3*(H2O(l)+2e-) => 3*(2HO-(aq)+H2(g)) + 2*(AL3+(aq) 3e-) 6 electrons are cosumed and produced, everything right : 2AL(s)+ 3 H2O(l) => 6 HO-(aq)+ 3 H2(g) + 2 Al3+ (aq) 2) the aluminium ion has an acidic behaviour and the HO- ion, is by essence a basis (oposite of an acid if you didn't knew it) the al3+ and HO- react in thefloolowing acido-basic reaction : Al3+(aq) + 3 HO-(aq) => AL(OH)3(s) (one aluminium linked with three oxygen, themselves linked with an hydrogen) The need for supplying current comes from the fact that the reduction of water into dihydrogen isn't a spontaneous reaction (totally the opposite, its the inverted reaction of H2+O2=> 2 H2O that is spontaneous, making hydrogen an extremely flamable gaz). You need to force the production of dihydrogen by pumping electrons in the cathode. About the reason for the size of the holes depending on the concentration of salt, I have my own (not proven in a lesson, my classes doesn't goes that far) : - the oxydo-reduction needing a current, that reaction is highly dependant on the resistance of the solution (thanks to Kholraush's law, the conductivity (invert of resistance) is proportional to the concentration of each type of ions) - a chemical reaction is always an equilibrium, if too much products are produced, the reaction will slow down and even come to a stop (you can even revert a reaction by increasing the amount of one of the products, it is a matter of the reaction's coefficient my idea is that with higher concentrations of salt, the resistance of the solution decreases, producing more aluminium and the reaction at the bottom of the hole slowes down whreareas the reaction on the sides get more proemeinent, making in the end a bigger hole
Hello,I love your project and interest to reproduce it. Is it possible to somehow calculate how deep will be the cut.Therefore it will be possible to mill to a certain precise deepth ?
have you looked into using potassium hydroxide or sodium hydroxide instead of sodium chloride. potassium hydroxide is supposedly the best electrolyte because the anode/cathode stay clean when producing hydrogen for a HHO generator. but maybe this quality is not what you're looking for when drilling/milling. something to try possibly.
I'm interested if one could create a tool of an inverse complex shape and use it to cut that shape out of a solid stock of metal. Does the ECM tool need be a continuous plane, or could it have high and low spots that need to be precisely transferred to the working piece? For instance, if I wanted to make a mold out of metal such as aluminum or steel, could I print the shape on the 3d printer, apply a conductive coating to the print, and use that complex printed surface as a tool to cut the shape into metal? I'm assuming the answer is no, since in the demonstration you seem to be using the ECM setup like a mill rather than a press.
Nice info. Something I missed is info on how fast your needles get damaged? For example if you use a needle to drill a 3mm deep hole in alu, how many holes could you drill before the needle gets too much damaged ? Or don't they damage a lot ? I also wonder if you have to ket the electrolyte go through the needle itself. Can't you also just spray from the sides like in edm ? (anyway, the needles aren't hard to get) . And 3rd question. You talk about using a 12V power supply in the beginning, but how many amps are aproximately being used from that 12V PSU ?
I have been working on desktop metal printing, but I plan on doing an Ender 3 conversion once I figure out how to modify the firmware slightly for very small z-moves
@@MorlockEngineering wow, how did I not think of that.. I tried epoxy castings but they were not inert enough for the chemistry I wanted to use them for.. but I also have a Resin printer so casting pdms should do the trick. However, I still need an excuse to build a jet ecm, for which I know how to get ridiculusy slow and minuyr movements :)
Current does matter, but for jet-ECM because of the large inter-electrode gap you need a high voltage. As a result of needing higher voltage, but not wanted to boil the water, the current must be limited to a degree. So the cutting is slower.
CNC laser engraver would be an easy step. The CAD file is on the github, so you could do that easily with some modification. Use inkscape to generate the gcode
The explanation of the electrochemical processes is wrong. For example the reaction in second line is completely wrong, you don't need any electons for it - let's check the charge balance - the overall charge on the left and right side of the arrow must be the same. Here, you can see that on the left side it's -2, while on the right side it's zero - two negative and two positive changes cancel each other. Also you don't end up with 3 unused electrons - you use them for the oxidation of second atom of aluminium. The overall reaction is: 2Al + 6H2O -->2Al(OH)3 + 3H2. You use the voltage to transfer 6e- from 2 atoms of aluminium to 6 H+ ions to create 3 molecules of H2 and 2 Al3+ ions.
FYI, there's a cheap CC/CV power supply that you can control via USB+Python that you might find particularly useful. It was your original video on etching that got me thinking about its applications, particularly with regard to monitoring the output in order to monitor kerf width/MRR and detect shorts when you move too fast. I've only gotten as far as poking a piece of aluminum with an electrode and seeing "yup, it removes stuff," but I did do an example bit on controlling the supply, which might be of use to you. demo: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-BoTV7003glU.html code: github.com/swolebro/swolebro-youtube/blob/master/arduino/dps5020.ipynb also, important assembly tips: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-4kIKpnaqWLw.html I hope to revisit this sometime, but got too much other nonsense going on at the moment.
I defiantly want to revisit ECM as well. I am waiting for an Ender 3 to become available so I can make a video on converting it to an ECM machine with a swappable tool head.
@@MorlockEngineering Whoops - I'm partly responsible for four of those Enders being out of stock. Hahahah. Every time someone asks me which printer to get, that's the answer. They really are great. I'm not so sure about using a bed-slinging printer for ECM though. I guess with a small needle jet of water, it's not that much to drain from the bed, but with a pump flooding the work, you'd have to make sure you drain it off fast enough that you don't have a reservoir of water sloshing around making you miss steps.
Hah, I was just going to ask if a CC supply with feedback would be advantageous here. What kind of current does this process need? If under 30mA produced acceptable results it would improve safety when dealing with the higher voltages. Feedback for short detection and even "touching off" the top of the material should be feasible.
@@shonkydonkey If you want to remove any significant amount of material, you're looking at 5-15 amps. More if you can. It's the amps that remove material, not the volts. For detail work or etching, milliamps will do.
@@swolebro Hmm... EE thinking out loud here... the current by itself isn't that big of a problem, but I know Morlock Eng. was talking about 300V+ which does indeed start to get expensive. Rectifying AC mains is a fairly cheap way to get ~100V 10A, but 15A and boost to 300V isn't easy. That kind of power would need a welder or dryer outlet or something since 15A at 300V would need at least 45A at regular US mains voltage. Wonder how much use 100V and 10A would be. It should be feasible to make a linear constant current supply to step that down, but it would need some pretty big (expensive) FETs. It might need a decent amount of bulk capacitance to keep that kind of current going though. Wonder what effect a switch mode power supply would have on this. It would probably be a cheaper way to do this. Then there's the safety related questions when dealing with 100V+ and enough current to kill you pretty good...
