Building an EDM machine. Part 1-some experiments

I've wanted to build one of these for years, even though I didn't have a real use for one before, no experience of using one, and only a rough idea of how they work. But there are just some things you can't make with a lathe or milling machine and the only way to do it is with one of these.

[@medievalhamster]

As someone who constantly explains that "I'm into EDM" does not mean that I like to dance, the running joke in this video really made me smile. Good job Andy.

[@AndysMachines]

😁

[@peterfitzpatrick7032]

They're not mutually exclusive... I like BOTH !! 😂 😎👍☘️🍺

[@robertqueberg4612]

Hello Andy, As a retired toolmaker, I am highly impressed. There have been times when a mold or die part needed a trip to another shop to remove a broken carbide drill or endmill. You have a good background in electronics, that helps you to “see” a project before the start. I will follow along with the series. The only thing that I thought of as a negative about an “electrical damaging machine” is the smell. Thank you.

[@Rolo-Tony]

That edm music at the end had me cracking up

[@CaskStrength777]

Piece of advice- flush makes the biggest difference with edm work. Ran ram edm for 2 years- your dielectric level should be higher than that- frothing shouldn't be happening at surface of oil when correct height. Contamination at burn area will create arc back and sputtering welding back onto your work. Nice job👍

[@AndysMachines]

Yes, you're right. I did add flushing (see the 2nd video) and got much better results.

[@jafinch78]

Awesome! Looks like another moment in re-enforcing uses for the old buzz boxes and other equipment that can be repurposed. Very clearly explained also. Keep us posted. Thanks for sharing!

[@bushratbeachbum]

Great project. Looking forward to seeing more!

[@gordon6029]

I’m really looking forward to this build.

[@pavano6970]

Super cool! The way you explain things is awesome! Keep doing!

[@JohnRR]

Wow what an excellent video. Very well done. Thank you.

[@gedr7664]

super exciting! looking forward to more

[@EleanorPeterson]

What an enjoyable video. Leaving aside the dodgy 🎵 bangin' chooonz joke, the calm presentation style reminds me very favourably of when I was a child and was allowed to join my dad on the sofa after midnight to watch the Open University on BBC2. (Yes. No PCs, no mobile phones, no tablets, no internet; if you were lucky your house had one huge [18"!] CRT TV and it lived downstairs in the corner of the living room. If you'd been a VERY good girl, your parents might let you stay up past bedtime to watch it. Sometimes there was cocoa☕. Ah, happy days...) Although I couldn't get my school-girlie brain around the university-level technical stuff, the mathematics and the squiggly formulae, I loved the 'intimate' narration, the hand-created pre-computer graphics, the often startling research-laboratory film footage, and the sense that I was learning something - even if I didn't know exactly what it was...🤔

[@xcyted4now]

New subscriber. Can't wait to see the rest of this build!

[@FilterYT]

Great video Andy, thank you!

[@shoobidyboop8634]

Nicely done illustrations.

[@stevebromley4230]

Great project. I love the pulsed power supply! A huge number of the old machines used a voltage somewhere in the 50's. Lots of the newer machines have a variable voltage. largely it's the amps that control how big of a chunk you blow out with each spark and the voltage helps control the gap between the electrode and part. So when you have the amps turned up over 100, you might only want 25 volts, to make the gap smaller. When you are down to .01amps you might want 300 volts, to even have a gap big enough for the burnt stuff to come out of the burn. One reason for 50-60 volts is nice is because mostly that voltage doesn't get through skin, even when covered with edm fluids. If you touch the 300 volt electrode, you know it. The 'gentile' flushing: ruffing seems happy with 8-25 psi, in most cases. Fine finishes might be any where from 2-40 psi. A small deep hole might need 3500 psi, IE pushing a .010 inch electrode through 3 inches. hope that is of some help.

[@AndysMachines]

Thanks for the information! There are a lot of variables I have to play around with.

[@suhailasaba1051]

The common idea that I know that making a spark needs high voltage in thousands. Would you explain how 60 volts can do the job ? Is it the power that matters in making a spark ?

[@AndysMachines]

You don't need high voltages to generate a spark if the gap is small and once an arc is created it doesn't need very high voltage to maintain it even over a larger distance. Welding machines can sustain arcs with as little as 10 volts but often use high voltage to strike the arc, or with a basic stick welder you just scrape the electrode on the work to start the arc.

[@suhailasaba1051]

@@AndysMachines Thank you for your good explanation.

[@minercraftal]

Pretty much clear explanations, thank you!

[@gillywild]

Very interesting. Looking forward to following your progress :)

[@JuulCPH]

Wow I've never seen such a simple EDM machine! I did not expect it to work so well. Looking forward to the next iteration :)

[@channelview8854]

Truth is it is possible to go much simpler. A buddy of mine built one with manual feed. The electronics was one diode, a capacitor, and one incandescent bulb to act as a resistor. Period. It works as an RC oscillator.

[@azeddineabdoune814]

Very good ,sir,clear explanations! Thank you.

[@ethedona]

Cool! That worked amazingly well for a simple initial setup. I've seen some wire cut EDM parts that fit so closely they look like a solid piece when put together, its like magic.

[@AndysMachines]

Yes, you can do some clever stuff with wire EDM, like cut a shape out of a block that's circular on the top and square on the bottom.

[@ashpalmer1365]

Thanks, this was full of really good information.

[@ericlotze7724]

I was thinking on designing a die sink EDM machine and open sourcing it, but seems you beat me to it! AMAZING project.

[@AndysMachines]

I say go for it if you want to! You're welcome to take ideas from my project.

[@Anton-hz4ev]

Looking forward on how this project turns out :)

[@AndysMachines]

Me too! Still early days yet but I think this is going to be exciting!

[@FaraidunHawramy]

great tutorial . thanks so much

[@eltonm.t.873]

Nice video! I liked how the video was put together!

[@Stefan_Boerjesson]

Really interesting. Possible use is removing broken taps, drills,.... making square/rectangular holes.... Now prescribing to se the follow ups.

[@kaine2416]

I ran both wire and die-sink EDMs probably 15 years ago in a little plastic injection mold shop. All our sinker electrodes were cut from big blocks of graphite, machined on a 5-axis Röders CNC machine, and hand polished before being used. I often wondered if we were duplicating work since we could have the $1mil Röders just machine the mold cores, but I wasn't getting paid to ask those questions lol. It was necessary sometimes with the complexity of some of our cores, but they used the EDMs for even the simple ones. 🤷

[@gothamantiquities2595]

Hey Andy your idea when finished is going to be brilliant when perfected! I was thinking of really interesting uses and materials for you cool piece! Best regards from New York!

[@ingenia-tec5194]

Im looking forward this project :D It will be awesome !!! :D

[@ricardogarcia3900]

God Job. Congrats.

[@miloudlemlem1520]

a good tuto work thanks

[@forge20]

Pretty sweet EDM video. The machining bit isn't bad either.

[@user-dp2ru6vh2w]

❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤❤That is the right Man with the right thoughts and experiences ... Excellent

[@imwhatim7259]

Bro this best , useful and great information thanks

[@QuinnHale]

Bravo!

[@moemcl6210]

I started working on EDM machine a couple of years ago using a similar arrangement using a rewound microwave transformer, arduino control, blow heater as resistor and microwave capacitor for spark. Used voltage divide resistor across spark to sense and feedback this voltage after each spark. Voltage too high advance stepper motor, too low or short reverse. Spark, sense, respond, spark at many times per second. Have not finished as became time poor but control system complete but my slide and motion had too much play and needed to be redesigned from scratch so your quality slide arrangement is a great bases. Look forward to see your process.

[@AndysMachines]

That's very interesting! I hadn't though of using parts from a microwave, I have several old microwave transformers on the shelf. Yes, I'm also using a divider network to sense the feedback voltage directly with an A2D pin. It definitely helps to have a solid linear slide to start with, I'm hoping not to need to back the electrode out at all, just slow down as it nears the target current, I could get a feel for this just turning the screw by hand.

[@moemcl6210]

@@AndysMachines I seem to remember that the trick was moderating the sensed voltage across the spark was a band around 20 volts. So this was the 'no movement' optimum spark gap and outside this would cause movement in or out. I believe some forward and back cycles are important for the flushing of metal particles but this is even second or so not the micro or millisecond spark cycles. Another thing that helped considerably was pumping fluid at the cut area (tried different cheap 12 volt pumps from eBay). I started the journey with 'Build an EDM by Robet Langlois' among many others. I started with dedicated PCB and digital ICs for control which taught me heaps about making my own PCB and all that discrete component stuff but I was finding that any change meant modifying board, trying then tweaking some more. I then made the leap to micro controllers ie Arduino which was yet another large learning curve (not much programming experience since my uni days with BASIC on an Apple IIe 30+ years ago) but gave more flexibility once the basics were mastered. I do hope to have a chance to get back to this project as it opens interesting manufacturing processes to compliment CNC stuff. I love 'making things'.

[@warrenking1815]

Nicely done, will be really great when you get the x y done.

[@niw-w1f]

Nice video! Freaking just broke another tap, and I can not afford to lose this workpiece so I am going to build an simple EDM now. In my case I just need to get me some linear rails, extruded alu profile, leadscrew+nut, and some machining, all the other parts are available so why not give it a try :). Although I could use my portal CNC mill as the whole guide thingy with an additional head, hmm interesting that would mean its finished today! Subscribed and waiting for more interesting videos!

[@wahabfiles6260]

I lold at Electronic Dance Music shock at 0:16 :)

[@vizionthing]

Outstanding - Though I wish I wasn't this early!

[@chrisfairbrother9197]

Hi Andy, many years ago we built a spark eroder at work and used the old fashioned bar heating elements as the resistive loads. Each bar had several metal clamps (jubilee clamps) along them. This enabled varying resistive loads to be switched in and out to vary the current as required by the cross-sectional area of the electrode. I may be wrong, but I remember the electrode was negative. We used copper good, elkonite (tungsten copper) better and graphite the worst for wear if the electrode was very large.

[@AndysMachines]

Hi Chris, thanks for the comment. In wire EDM the wire is almost always negative, it's supposed to cut faster but also wears faster, however as new wire is continually fed in it doesn't matter. In sink EDM the electrode is usually positive I believe, but sometimes the opposite polarity works better depending on the work and electrode materials. That's interesting you found graphite to wear faster, I thought it was supposed to be better. I've got some graphite rods but haven't tried using them yet (it's really messy stuff to machine).

[@chrisfairbrother9197]

@@AndysMachines this machine we built was a long time ago when spark eroders we not available or extremely expensive. Graphite has improved over the years. In later years, we used tons of the stuff to spark crt monitor cavities and even whole car bumper cavities. Try getting in touch with Erodex, they may let you have some samples.

[@clivesteer5235]

Hi Andy. Great to see the project moving forward. We may have been in contact before as I've the same machine as you. My project is on the back burner at the moment due to other projects but after seeing your vid I may restart as I've only the power electronics to do.

[@AndysMachines]

I don't actually know much about the original machine, I only had parts of it and none of the electronics. I think it was made by Able and is quite old.

[@raymondzhao9557]

GREAT

[@DIYGuitarMods]

I'd be very unterested in seeing if the meanwell hlg series constant current PSUs would work for this application. seems like a good fit! They seem to be used mostly for lighting applications but if the dimming function can switch fast you could send your micro controller PWM to that and it also already would be short circuit protected since its constant current

[@SynthRockViking]

I like both edm's😌

[@enriqueruizalfonzo636]

Super. .. genial

[@d6c10k4]

You might look for an old Charmilles D10 ram EDM that you can get working. I would think one could be purchased pretty cheap and with your electronic skills you should be able to fix anything that might be wrong with it. It will have the X-Y table, the servo to control ram, power controls, and fluid system. It's a benchtop size machine so doesn't require much space. I used to run a D20 Charmilles which is a slightly larger floor standing model but otherwise much the same.

[@AndysMachines]

These look like nice small machines with similar functions to what I'm building. I have seen a few EDM machines come up at a local machinery auction, but all far to big for home/hobby use. Though buying one would take away all the fun of building it from scratch!

[@Graeme_Lastname]

A micro controller a few rails and some steppers. No need to apply force. You see very little about this subject. Thanks m8. :)

[@jimsvideos7201]

It dawns on me only now that this might be ideal for cutting keyways, rather than investing in broaches or a shaper.

[@Reman1975]

I'm watching this and pondering how long this would take to zap an axle hole (Complete with keyway) into something like a 1/2" mild steel gear? If the electrode was a hollow profile, like a cookie cutter, it would cut down the amount of metal that needed electrically eroding a fair bit, and would also decrease the power requirements. This would be at the expense of the electrodes life span, but for short runs of items it could be worth it. Machining a small batch of gear centre holes and keyways is quite a time consuming task for a small workshop. If the batches aren't big enough to invest in an item specific fixture for the lathe it can easily take a few minutes per part just getting it centred in preparation for boring out the axle hole, and machining to size. After that you still need to broch the keyways. Even a very experienced engineer with all the required equipment on hand may need to invest 30 minutes on each gear just to get the centre hole profile correct. On the other hand, even if this EDM setup might take longer, a simple gear blank locating hole could be bored into a base plate, and relative unskinned labour could run the machine, or an engineer could set the machine running and go off to do something else. Between gear blank swaps.

[@klausbrinck2137]

So, just to summarize, and know if I got it right: Your resistance is "a couple of ohms", so, let´s say 1.5 ohm, gives an RC-Oscillator-frequency of 1000Hz (when combined with your 100µF-capacitor)... You overlap the "fast" 5kHz-Pulse-Generator-frequency over those "slow" 1000Hz, so that the RC-oscillator-frequency becomes irrelevant, but the capacitor-discharge itself remains very well relevant, since only a capacitor itself could deliver the current fast enough. Then, your sink the electrode manually, and go by your ear and by your vision, in order to choose if you´ll sink the elektrode even further, or back up the electrode... In the next video, you add an (indirect-)Gap-Current-sensor (sensor actually measures the capacitor-voltage, that´s inversely proportional to the gap-current), to generate the information needed, in order to guide a stepper, that will forward the electrode from then on, instead of manually driving it. Did I get it right ???? Thanks... Your answer would be greatly appreciated.

[@AndysMachines]

Yes, that's basically it, the capacitor is a reservoir that can deliver the very short high current pulses. The gap voltage is maintained by measuring (indirectly) the current, current and voltage being relevant to each other.

[@klausbrinck2137]

@@AndysMachines Thank you again !!!

[@kaanabukay4791]

Dear Andy, It is very simple and usefull machine. Congratulation. Can you share electronics circuit schema if posible. Best Regards, Kaan ABUKAY.

[@policedog4030]

Three cheers for this noble effort! The Chinese have a saying - something like "The work will teach you how" - While I've just been wondering why no one is making a bench-top EDM; luckily you're a person of action and wade right in. I wanted to suggest using carbon gouging rods for the electrode; I think they would be cheaper and safer than brass which commonly has traces of lead and arsenic. I was surprised to see the brown color as isn't it just an oxidation of the aluminum? If so the color of the removed material should be white; could the brown be coming from something in the brass electrode? McMaster-Carr here in US sells a pack of carbon gouging electrodes in many diameters for US$15 or so; also carried at pro welding shops. edit: I should have been more clear these are the BARE carbon gouging rods so not the copper foil covered type

[@AndysMachines]

Thank you! Yes, I have some graphite rods, but it's messy stuff to machine, so I'm not using it for testing. I think the main advantage of graphite is that the electrode should last longer. I also have some pure copper to try. I think the brown colour is just tiny particles of aluminium, probably with an oxide coating, so it could be more oxide than metal. You get a similar colour when when wet sanding aluminium with fine sandpaper (1000+ grit) and the finer you go the blacker it looks. Even though aluminium oxide (Al₂O₃) is clear/white, apparently when it's very tiny particles it appears black due to scattering of light or something. I also tried steel workpieces and they produce dark clouds too, but then you might expect that. I wonder what titanium would do?

[@paulbendel986]

Nice setup! What is the pulse generator circuit? Are you driving the IGBT gate with another transistor?

[@AndysMachines]

I'm using a micro-controller running at 5v to generate the pulses, this is fed through 3 transistors using a separate 15v supply to drive the gate of the IGBT, I'm using a pair of transistors to pull the gate high and low as quickly as possible to discharge the gate capacitance, through a small resistor to avoid ringing.

[@akfarmboy49]

I’m in need of a EDM.

[@Raul28153]

This is interesting. the reading I've done all demands an electrode that rises and falls in microscopic distances at a high frequency. So much work goes into this one feature as to make the production of one seem overly daunting.

[@AndysMachines]

The electrode will rise if the current exceeds a certain level, but not at high frequencies. I find you only need to pause the downward feed of the electrode to maintain the proper gap voltage, obviously I'm not cutting as fast as some industrial machines though.

[@Raul28153]

@@AndysMachines That makes perfect sense. I have a PC circuity board and a book put together to walk me through the manufacture of a more sophisticated sinker EDM. I've been so busy I haven't cracked the book.

[@yetihehe]

@@Raul28153 High speed electrode movements help with recirculating of fluid in gap.

[@eletromatica]

Congratulations on the project. I am trying to build an edm machine for my workshop and I saw your video. I built a 1khz pwm source from 0 to 150v and from 1 to 20A with 2 irfp260n and I'm using my cnc router using the g38 cnc command in conjunction with the current monitor circuit to adjust the gap. I did several tests with capacitors and I still haven't had a good result. My question is whether there is any recipe for dielectric fluid because here Brazil is very expensive

[@AndysMachines]

I just use water (distilled) as the dielectric, it's cheap and safe. It needs to be de-ionised or distilled, ordinary tap water is too conductive.

[@klausbrinck2137]

10kHz is minimum

[@DaxxTerryGreen]

Well done Andy. Where are you btw? I'm in Asheville, NC, USA

[@wantafastz28]

Any diagrams to your complete circuit for the pulses? Awesome work

[@AndysMachines]

That was just a test, the final circuit will be different, the only thing not shown on the diagram was the microcontroller producing the pulses, this could easily be done with an arduino.

[@andrewstambaugh8030]

Could you use a VFD instead? or is that a bad idea?

[@250smacks]

If EDM isn't on a cnc, what advantages does it have over plunge ECM? Given you can use 3d printed electroplated electrodes

[@AndysMachines]

They each have their own advantages/disadvantages. EDM is faster and gets better tolerances. With ECM you don't have the electrode wear, but you tend to get 'cutting' where you don't want it and cuts also flare out / corners become rounded.

[@tonyhill8300]

what would be an Ideal voltage?, what part do amps play? is that dependent on surface area?

[@AndysMachines]

I don't know what's ideal, I think it varies from machine to machine. You need enough voltage so that a spark can jump the gap between electrode and workpiece, too small a gap and you risk them welding together. I found less than 40V just doesn't seem to work. Too much voltage and the gap gets bigger meaning the cut gets bigger than the electrode, and sharp corners get rounded. More amps are needed to cut faster. Normally different voltage and current settings are used for roughing and finishing cuts.

[@Spark-Hole]

Power Mosfet would be better than IGBM. It would turn off faster and that reduce electrode wear. Anyway, thumbup.

[@AndysMachines]

Yes, MOSFETs can be used and I did consider that, but it seems IGBTs are preferred for applications like this as they can handle higher voltages. The particular IGBT I used, it turns out may not be the best suited type, but it seems to work just fine. I can always change it later if needed.

[@ParsMaker]

nice work, But I thought in order to have effective sparks the voltage hast to be over 1K and current should be over 10amp, please make more video on this project

[@AndysMachines]

Voltage is usually between about 50 and 400V. You probably could use 1kV but the higher the voltage the bigger the gap gets between the work and the electrode which is usually not desirable. Current can be anything from milliamps to over 500A depending on the size of the machine. There will be another video coming soon!

[@ParsMaker]

​@@AndysMachines I look forward to see your new video, Im thinking to make one but I DONT know much about the edm

[@thimothybennyvlogs]

Hi Andy, could you give specifications on the ohms required for the resistor, also the type of transistor used, is it pnp or npn?

[@AndysMachines]

The resistor I used was a heating element which is the equivalent of a very high power resistor, I think it was 400W and around 1.5 Ohms. It is an IGBT transistor but in the same circuit you could also use an NPN transistor (a very high power one) or a P-channel MOSFET.

[@akshayrnair8250]

Thank you very much

[@vinnycordeiro]

That's very interesting. I wonder how would you make submillimetric holes in brass or tool steel (maybe I'm thinking about making my own 3d printer's nozzles, just maybe... 😆)

[@AndysMachines]

Yes, small deep holes is something EDMs are good at. You just need a thin wire for the electrode. You can actually drill 3D printer nozzles on a lathe down to 0.3mm without too much difficulty, but the deeper the hole gets relative to it's diameter, the greater the chance of breaking the drill.

[@vinnycordeiro]

@@AndysMachines I'm no machinist, just someone who watches a lot of machining content on KZbin. I already saw brass nozzles being drilled by a 0.4mm bit, but for hardened steel (necessary for abrasive filaments, like glow in the dark or carbon fiber-filled ones) I believe this method is easier than drilling a hole.

[@AndysMachines]

@@vinnycordeiro Yes, certainly. EDM doesn't care about how hard the material is. But drilling hardened steel is much more of a challenge, even if it's not a tiny hole.

[@SillyPutty3700]

Question. Your hole will only be as round as your electrode correct?

[@AndysMachines]

The hole will be the exact shape of the electrode. but it will be slightly larger due to the spark gap. If you want perfectly circular holes you can also spin (or orbit) the electrode.

[@KyrychenkoAnton]

I wonder how this simple electronics circuit differs from baxedm arc generator, that you can buy to build your own machines with it. Its a quite cool project, and in future I might go building one with it. Just curious about how different it is since in recent video he compared such simple designs as you have with his generator as "bicycle vs sports car", so I guess there's smth different going on, but can't find info of how exactly..

[@AndysMachines]

Fundamentally they work the same way, but his has a load of extra circuitry meaning it can run at higher currents without blowing up, it's probably a lot more efficient too. In his older videos he did go into more detail about how he built his EDM machine/ spark generator, but now he is marketing it he seems to have deleted those videos. A pity, but I suppose you can't blame him.

[@KyrychenkoAnton]

@@AndysMachines I see, of course, he put a lot of work into it.

[@Core3DTech]

will this power setup work for Wire EDM as well?

[@AndysMachines]

Yes, it will work, though you may want to swap the polarity for faster cutting since wear on the wire is not so important with it being constantly replenished. However 'proper' wedm machines have much higher rated and more sophisticated power supplies.

[@frsc5063]

Hi, in your diagram You put the negative pole to the workpiece ?! In TIG welding is - to protect the electrode - the negative pole always on the electrode ... Maybe even this make here a difference...

[@AndysMachines]

Yes, it's usually this way around with sink EDM, but see the next video where I experimented with swapping the polarity.

[@tonyhill8300]

What effect does the capacitance of the capacitor have? can you go too large or small? assuming it is within the voltage spec. also the IRG4PC40S IGBT has a working frequency of less than 1 khz , according to the data sheet but you say your using 5khz as the carrier frequency?

[@AndysMachines]

I haven't tried altering the capacitance, I just picked a value I thought might work. Some EDMs have banks of capacitors you can switch in and out to vary the total, but I think that's more relevant for the non-pulsed type were it's the value of the capacitance that determines the size of the spark. With a pulsed EDM the length of the pulse controls how much energy is delivered per spark and the capacitor only acts as a reservoir. As I understand it, longer spark (or bigger capacitor) equals faster cutting but poorer surface finish and smaller spark equals slower cutting but better finish. 1 KHz sounds extremely slow, are you sure it's not 1 MHz? If you look at the scope trace at 2:57 you can see it's switching fine at just below 5KHz and would go much faster. I looked at the datsheet and the longest switching time delay I can see is 1000ns (1μs).

[@tonyhill8300]

@@AndysMachines pretty sure "Optimized for minimum saturation voltage and low operating frequencies ( < 1kHz)" thats straight fro the data sheet?, maybe it means it will run at less that 1khz. I notice you have a Capacitor across the power supply is that just a reservoir? Have a look at this for the pulse generator, it will output up to 32v which is ideal for supplying the IGBT (15v nominal) it also retains its setting!: www.ebay.co.uk/itm/1Hz-150kHz-PWM-Signal-Generator-Pulse-Frequency-Adjustable-Module-Test-Equipm-b/193682088378

[@AndysMachines]

Ok, I hadn't seen that. This may not be the best IGBT for this application, I just happened to have an old one and with my limited knowledge of IGBTs it seemed suitable from the specs I looked up, and appears to work just fine. If I need to replace it I might do some more research. You could also use a MOSFET, but I didn't have one rated to a high enough voltage. Yes, the power supply capacitor is just the normal smoothing one, it's actually part of the power supply I used. Yes, that pulse generator will certainly cover the range needed and has a nice push-button interface, however it may not drive an IGBT directly even if it can supply the correct voltage. You normally need to charge and discharge the gate capacitance as quickly as possible through a small resistor (~10Ω) to reduce 'ringing'. So you may need an additional driver in between. Also it can output frequencies much higher than needed and duty cycles much lower (down to 0%) so you need to be careful to avoid combinations that produce very short 'on' pulses, otherwise the IGBT will not have time to turn fully on (due to the gate capacitance again), it will be partially conducting and could get very hot and let out the magic smoke.

[@tonyhill8300]

@@AndysMachines apparently you only need a IGBT driver if you frequency is above 20khz, you can just use a 10k pull down resistor: kzbin.info/www/bejne/iKm1e4pmbqaSjqM Just need to check that pwm generator can supply enough amps!

[@AndysMachines]

GreatScott!'s videos are great, I've watched a lot of those. Often you can get away just using something simple. What I usually do is look at the gate drive signal on an oscilloscope, this can tell you a lot, for instance at 4:24 on his video you can see the ringing I mentioned.

[@dereknbartram]

What voltage was your pulse generator sending to the IGBT? I'm trying to recreate this using a IRG4PC40KDPbF but not having any joy.

[@AndysMachines]

The gate drive voltage was around 15v, which I think is typical for most IGBTs.

[@dereknbartram]

@@AndysMachines thanks, much appreciated. After some additional testing, it looks like the IGBT is dying as when there's a short the capacitor discharges through a short circuit with only the IGBT providing any resistance - that means well beyond the 84A peak (120A for your chip). Have you had issues with components burning out?

[@AndysMachines]

@@dereknbartram The first IGBT I used did blow up fairly quickly, I don't remember what type it was but probably not correctly rated. The current capacitor and IGBT do get hot after running for a while.

[@VigilanteSystems]

How about using Kerosene for the dielectric Fluid?

[@AndysMachines]

Yes, Kerosene or diesel fuel is supposed to work well, though so far I've stuck to distilled water for safety.

[@hpevans9041]

Hello, great video. Doesn’t the distilled water short out the work piece and electrode? I thought maybe wrongly that the di-electric acted to cool the work piece and help remove the waste material. I was told EDM was first developed by the Russians for use in their space industry.

[@AndysMachines]

Yes, the dielectric helps cool the work and flush away the waste material. Distilled water is much less conductive than regular water, though there is still a small amount of current flow.

[@klausbrinck2137]

Destilled water is similarly conductive to air, since it contains no salt/minerals , that could carry the current. It were the Soviets, not the Russians. It´s not exactly the same. The Soviet Union was made of many countries, even if the biggest one of them was Russia.

[@preciousplasticph]

Have you tried ECM machining? It's similar

[@AndysMachines]

Yes, I've done ECM on a small very basic scale for engraving writing on control panels etc. For those who don't know, instead of using an arc to burn away material it uses electrolysis to remove (or deposit) metal. You can use a mask to screen off areas you don't want affected and also shaped electrodes similar to EDM since electrolysis goes much faster when the electrodes are closer together.

[@sterlingdeye5660]

Beware headphone users

[@gearhead1302]

😂 the sad welder

[@luderickwong]

I always have a feeling that the whole process looks like mig welding machine work in a oil bath instead of gas shield? I know it shouldn't be understand this way but my IQ are just limited to that 😄 Some say that these machine have better energy efficiency than using small drill bit on drill press because it is not a mechanical motion?

[@whizzo94]

Which microcontroller did you use? Could you use an Arduino, Uno or Mega?

[@AndysMachines]

I used an Atmega328 which is the same processor as in the Uno, so yes you could use that instead.

[@michaelmaxim5689]

Try using a paraffin oil instead of the distilled water it works much better. Like what you can get to use in oil lamps.

[@AndysMachines]

Yes, diesel (fuel oil) is supposed to work well too, even vegetable oil. For now I'm sticking to water as it's less messy and not flammable.

[@michaelmaxim5689]

@@AndysMachines Fires are not really an issue because the flash point is too high.

[@tonyhill8300]

built this circuit and it makes no difference whether the IBGT is in the circuit or bypassed?

[@AndysMachines]

It will work without the switching circuit (workpiece connected directly to ground) In this case the capacitor will charge up through the resistor until the voltage across the gap is big enough for it to arc, the entire charge from the capacitor will be dumped and then the cycle will repeat. You will have no control over the spark except by changing the value of the capacitor, you can have banks of different values of capacitors to switch in and out. With the switching circuit you can fine tune the spark for less electrode wear, faster cutting and better surface finish, but depending on what you want to do the resistor-capacitor only version might be perfectly suitable. eg. for burning out broken taps surface finish and electrode wear don't really matter.

[@TABE-O]

Ripper!!

[@marcobogos6510]

When can we expect part 2?

[@AndysMachines]

Soon. Maybe next month, I've made quite a bit of progress already.

[@titter3648]

How will you detect short circuits?

[@AndysMachines]

By measuring the the current and if it's above a set level backing the electrode out.

[@KkKk-yd1sv]

What is the Distilled water ?

[@SillyPutty3700]

If you are in the US or Canada I can hook you up with an EDM table and oil tank. with the vertical stepper motor already in it

[@AndysMachines]

I'm actually in the UK, but thanks for the offer.

[@ClosestNearUtopia]

So the holes, which were drilled with unbelievable tolerances at the Gaza Piramids, could it be done with EDM?

[@AndysMachines]

Not unless the pyramids are made of metal, or some conductive material.

[@ClosestNearUtopia]

@@AndysMachines theyr made out of piezo electric material, which in some way, does cunduct, and even produce energies under its own weight.

[@ClosestNearUtopia]

@@AndysMachines thanks for tour reply btw.

[@evilkidm93b]

that was funny

[@bobweiram6321]

I wonder if using a graphite, tungsten or titanium electrode will wear out less.

[@AndysMachines]

Yes graphite should last longer. I have some to try, but it's very messy stuff to machine. Even just pure copper should be better. I don't think brass is the best choice, it's used in wire EDM machines, but then the wire is constantly replaced.

[@bobweiram6321]

@@AndysMachines I am assuming an ideal anode in electrolysis also serves as a good electrode in EDM. What properties makes for an ideal electrode? I'm thinking a relatively inert material with a high conductivity would be ideal, like Gold (cost not withstanding).

[@klausbrinck2137]

@@bobweiram6321 From the countless experimental EDM-videos I´ve watched, I´ve come to the conclusion, that pulse-duty-cycle and -frequency is far more important than the material, and that the material is always the easiest possible, like copper or brass. In a video, a guy brought wear down from several 10s% to 0,5%, through trial-and-error with duty-cycle and frequency.

[@etmax1]

do the drilling clamped down properly and with a bit of oil and it will be much better

[@AndysMachines]

Yes, this was just a 1st test. Even though there's no real force on the workpiece it does help to clamp it down as small vibrations can move it slightly. I haven't tried oil as the dielectric yet.

[@bjl1000]

Part 2 ?

[@AndysMachines]

Working on it right now...

[@user-cl8kl2ss9i]

Электро эрозионный станок .

[@vajirasamaranayake6956]

Can I send the Sakit Daigam

[@MArco09380]

Where could I find the Drawings for the electoral system?

[@AndysMachines]

The circuit diagram and explanation is in part 2 of the video. I used quite a basic circuit, there are more sophisticated ways to do it.

[@reprapmlp]

At 8:26 you say "18 millimetres cubed" but the on-screen text says "18 cubic millimetres" per minute, and the text matches the actual rate.

[@AndysMachines]

Yes, I meant 18mm³ (cubic millimetres), not an 18mm x18mm x18mm cube. Sorry about that, cubic millimetres isn't a unit I use very often.

[@tinkertom3683]

If you clamp two pieces of wood then drill you get a clean cut lol

[@wxsb-xp2tv]

low esr capacitor

[@bangtutor6173]

Lol elektronik dance music

[@yurycz8933]

+

[@thongvo2865]

R1=??

[@AndysMachines]

in this case it was only a few ohms, you can calculate it using V=IR to limit the short circuit current from the power supply.

[@thongvo2865]

I thank you. I do not know English. I am using google translate. I often have problems with broken drill bits, broken taps and hardened steels larger than 60hrc. Maybe this is the cheap solution for me. I can't afford an industrial EDM machine