PleasE MorE InFo !!! The best I´ve seen so far, simple circuit, but still pulses are dictated by an arduino, as is sensing of too small distances, but there´s too little info for beginners like me... A circuit-diagramm, as well as the code would be nice...
The sound was music to my ears, and the light from the sparks was barely perceptible, exactly as should be !!! I mean, he surely is my hero, but I´d love to know more about the circuitry and the PWM-generator , as well as the cap-voltage-sensing/code... I cannot program an arduino as a PWM-generator, but there´s very cheap ones to buy, but I cannot judge whether such one is gonna pair well with the Mosfet I will use.
Thanks for the video, your results look very good. I have been playing with simple RC circuits and cannot obtain a sustained spark. I guess using a pulsed output would improve this? I realise the drive to the MOSFET is also control by open/short/current detection but does it use a fixed base frequency (25kHz) and what is the mark/space ratio or duty cycle of the signal? Perhaps you use PWM of the pulse based on measured cap voltage... be interesting to know. Trying to get an idea of a good pulse to start with if I build a pulsing circuit.
Sorry for never replying to this. If I remember correctly, I played with different frequencies and duty cycles, and determined that 30-40 kHz with 5-15 microsecond pulses worked the best. The duty cycle was chosen/discovered such that the bulk capacitors were allowed to recharge mostly between arcing. This gave me the most consistent results. If the electrode voltage was below about 60 volts at the beginning of a spark, it would not reliably initiate an arc. IIRC I ran the bulk capacitor around 90 V max, and during arcing it would measure 75-80 V with a DMM.
@@gizmoguyar "The duty cycle was chosen such that the capacitor was allowed to recharge between arcing" Was the RC-Circuit-frequency a straight multiple or straight fraction of the frequency of the mosfet-switching, or was it chosen so that both have nearly the same frequency? Thanks !!!
@@klausbrinck2137 It was neither. I empirically measured the RC time constant, and I used a MOSFET frequency of about 4-5 times that. This way I could be confident that at the beginning of each pulse, the capacitor was roughly at the same level of charge, independent of the size or length of the previous pulse.
@@gizmoguyar This answer was very satisfactory, finally I know about the relationhip of the 2 oscillating systems !!! Thanks a lot. The ratio may be different for my efforts, but, at least, now I know what I should pay attention at. I know nearly nothing about electronics, all I do, is to logically evaluate all statements expressed here, bring them in relation, and find the truth. My biggest fear was, that the "ON"-part of the mosfet-duty-cycle is gonna precisely overlap with the least-charged-capacitor-state of the RC-oscillation-cycle, for example. But your answer tells me, that I shouldn´t fear that much, generally... Thanks again !!! I endlessly envy your results, greatly done !!!
That's a pretty great-looking rig, I'm still learning about EDM in general, but liking yours as a DIY approach. This might be a silly idea, but is it possible to have a very strong magnet either inside the water or under the bottom of the tank? It might help keeping some metal particles away from the cutting area.
Well that's encouraging; pretty much the exact same architecture as I'm trying to get working. The only difference I have is using a Pi Zero, purely to give me a WiFi gap between the equipment and my laptop. The issue I'm having is with the IGBT keep dying; as the spark occurs, you get an instantaneous moment of ultra high current, which is frying the thing. Could you advise what MOSFET/IGBT and driver (if applicable) that you're using please?
I am using a normal N channel Mosfet here. I don't remember the part number, but It's rated for much much higher voltage and current than required (I think rated for 800V, 120 A continuous, pulsed to 1000 A). But I did add small amount of resistance to guarantee some sort of current limit (about 1 ohm). It's also avalanche rated. I did not have any issues with it burning up, but I definitely had some issues with noise and clean switching. I added a low impedance mosfet driver on the gate, kept all the wires short, etc. etc. all the standard stuff. I think if you don't limit the current you will always burn up your switches. The resistance and inductance in those short wires is super low. The "proper" way to do this is to add inductance in series with the switch, and control the duty cycle like you would in a buck converter. This requires extremely fast electronics, though. And a lot of math and careful design to prevent oscillation.
Hello sir I'm very much impressed by your work will you please care to share a detailed schematic diagram along with proper circuit diagram and other components and the Arduino code then it would be great help
How does the pulse shape look like in oscilloscope ? Can you show it? Is it just a simple one step pulse of some ratio of duty cycle? Or, a two steps one, a low or higher at first or vice-verse ?
Old fart in Ireland here, just a suggestion, would a IGBT welder work as a power control moduale for EDM it would seem to have all of the values required for this. Any body out there got any ideas on this ?
Yes, I'm sorry about that. I recorded this with my phone, and it was doing some strange noise suppression stuff (there were other noisy machines running in the room). ...one more thing I need to learn.
PleasE MorE InFo !!! The best I´ve seen so far, simple circuit, but still pulses are dictated by an arduino, as is sensing of too small distances, but there´s too little info for beginners like me... A circuit-diagramm, as well as the code would be nice...
I own 2 wire edms and a sinker edm. Thats not a bad little system Youve built there. The sound on the letter cutting was not bad at all. Good job!
The sound was music to my ears, and the light from the sparks was barely perceptible, exactly as should be !!! I mean, he surely is my hero, but I´d love to know more about the circuitry and the PWM-generator , as well as the cap-voltage-sensing/code...
I cannot program an arduino as a PWM-generator, but there´s very cheap ones to buy, but I cannot judge whether such one is gonna pair well with the Mosfet I will use.
Do you have the sketch for the Arduino
Nice work.
Thanks for the video, your results look very good. I have been playing with simple RC circuits and cannot obtain a sustained spark. I guess using a pulsed output would improve this? I realise the drive to the MOSFET is also control by open/short/current detection but does it use a fixed base frequency (25kHz) and what is the mark/space ratio or duty cycle of the signal? Perhaps you use PWM of the pulse based on measured cap voltage... be interesting to know. Trying to get an idea of a good pulse to start with if I build a pulsing circuit.
Sorry for never replying to this. If I remember correctly, I played with different frequencies and duty cycles, and determined that 30-40 kHz with 5-15 microsecond pulses worked the best. The duty cycle was chosen/discovered such that the bulk capacitors were allowed to recharge mostly between arcing. This gave me the most consistent results. If the electrode voltage was below about 60 volts at the beginning of a spark, it would not reliably initiate an arc. IIRC I ran the bulk capacitor around 90 V max, and during arcing it would measure 75-80 V with a DMM.
@@gizmoguyar "The duty cycle was chosen such that the capacitor was allowed to recharge between arcing" Was the RC-Circuit-frequency a straight multiple or straight fraction of the frequency of the mosfet-switching, or was it chosen so that both have nearly the same frequency? Thanks !!!
@@klausbrinck2137 It was neither. I empirically measured the RC time constant, and I used a MOSFET frequency of about 4-5 times that. This way I could be confident that at the beginning of each pulse, the capacitor was roughly at the same level of charge, independent of the size or length of the previous pulse.
@@gizmoguyar This answer was very satisfactory, finally I know about the relationhip of the 2 oscillating systems !!! Thanks a lot. The ratio may be different for my efforts, but, at least, now I know what I should pay attention at. I know nearly nothing about electronics, all I do, is to logically evaluate all statements expressed here, bring them in relation, and find the truth.
My biggest fear was, that the "ON"-part of the mosfet-duty-cycle is gonna precisely overlap with the least-charged-capacitor-state of the RC-oscillation-cycle, for example. But your answer tells me, that I shouldn´t fear that much, generally...
Thanks again !!! I endlessly envy your results, greatly done !!!
That's a pretty great-looking rig, I'm still learning about EDM in general, but liking yours as a DIY approach.
This might be a silly idea, but is it possible to have a very strong magnet either inside the water or under the bottom of the tank? It might help keeping some metal particles away from the cutting area.
Well that's encouraging; pretty much the exact same architecture as I'm trying to get working. The only difference I have is using a Pi Zero, purely to give me a WiFi gap between the equipment and my laptop.
The issue I'm having is with the IGBT keep dying; as the spark occurs, you get an instantaneous moment of ultra high current, which is frying the thing.
Could you advise what MOSFET/IGBT and driver (if applicable) that you're using please?
I am using a normal N channel Mosfet here. I don't remember the part number, but It's rated for much much higher voltage and current than required (I think rated for 800V, 120 A continuous, pulsed to 1000 A). But I did add small amount of resistance to guarantee some sort of current limit (about 1 ohm). It's also avalanche rated. I did not have any issues with it burning up, but I definitely had some issues with noise and clean switching. I added a low impedance mosfet driver on the gate, kept all the wires short, etc. etc. all the standard stuff. I think if you don't limit the current you will always burn up your switches. The resistance and inductance in those short wires is super low. The "proper" way to do this is to add inductance in series with the switch, and control the duty cycle like you would in a buck converter. This requires extremely fast electronics, though. And a lot of math and careful design to prevent oscillation.
Cool project, thanks for sharing. In the butterfly the triangles have round bulbs on the tips, is that an artistic feature or a design compromise?
Bro that was bad ass! Could you please put together a parts list with a schematic. Or am I being extremely lazy? Lol. Good shit though!
Please do so abd add in amazon affiliates links
Hello sir
I'm very much impressed by your work will you please care to share a detailed schematic diagram along with proper circuit diagram and other components and the Arduino code then it would be great help
looks like a great project - too bad the audio is so weak
Have you tried graphite yet? I’m new to the channel.
hello, i try edm mosfet circuits bu no luck :) can you explain the high speed mosfet circuit please
How does the pulse shape look like in oscilloscope ? Can you show it? Is it just a simple one step pulse of some ratio of duty cycle? Or, a two steps one, a low or higher at first or vice-verse ?
i need this wiring tutorial can you help me make similar one
Bravo......great job.........I use mineral spirit's........cheers
Old fart in Ireland here, just a suggestion, would a IGBT welder work as a power control moduale for EDM it would seem to have all of the values required for this. Any body out there got any ideas on this ?
nice work man! any chance you can release some BOM and build sheets?
I would have loved this but the volume is way too low
Yes, I'm sorry about that. I recorded this with my phone, and it was doing some strange noise suppression stuff (there were other noisy machines running in the room). ...one more thing I need to learn.
@@gizmoguyar 😁No prob, we all must live and learn, otherwise we may not live long enough to learn LOL
The real problem with EDM is getting the ARC power supply, the rest is just a 3d printer and a bucket
Nice
🌹🙏