So as you charge the capacitor up because E = 0.5CV^2 the inductor keeps seeing higher impedance from the cap. Thus because the time constant = L / R it discharges much faster. Now the pulses on the discharge shrink to a tiny percentage of the on cycle and thus if you don't adjust your duty cycle according to this trend you end up charging very little as the cap voltage increases. Hope this helps, I like your project, watching with interest.
Don't give up on the single inductor boost converter yet. Current through an inductor increases with time when a DC voltage is placed across it according to V/L = di/dt. likewise when the transistor is turned off it decays through the diode and into the capacitor following the same equation. The voltage which sets the rate of decay will depend on the capacitor's voltage plus the diode drop. The switching frequency and duty cycle must be chosen so the transistor is not on long enough to saturate the inductor and it should be off long enough for the current in the inductor to decay into the capacitor. Another way to think of it is the energy stored in the inductor at the time the transistor turns off is Q=1/2L(i^2) which is also the amount of energy that can be stored in the capacitor after the transistor turns off. Energy added to the capacitor will be Q=1/2C(V^2). With multiple cycles the capacitor should charge to any desired voltage within its rating. Time requred for the inductor current decreases as the capacitor voltage increases as can be seen in the above equations.
Hey mate! I have also been working on a coilgun in the past months. I have a few suggestions you might find useful. 1. The classic single mosfet and transformer circuit is much better and effective that boost converters. I've initially tried to make a boost converter for cap charging, but the charging was slow, and I got 200V max. 2. You can use a pair of LEDs facing one another as your motion sensor. Basically one will be on, and the other will go to the analog of the arduino. I programmed my arduino in such way that if the voltage drops (object is in the coil), it triggers the switching mechanism. 3. As for the coils, use thick wires and aim for almost non existing resistance. It's best that the cap dumps all current at once. 4. Thyristors seem to be the best option for stage switching, I fried countless mosfets so far, and relays are too slow. Can't wait to see your creation!
@Isaac Eda It's basically a circuit where there is a transformer with a center tap winding, one winding gives the control signal to the gate of a MOSFET or IGBT, the other delivers the current through inductance. boginjr.com/wp-content/uploads/flysche1.png The schematic above is similar to what I've made, only instead using an IGBT. Also I've soldered a zener diode between gate and emitter. czcams.com/video/0jWpQ8HEYgk/video.html&ab_channel=LudicScience That video covers it up very well
I think (and have read it) that IGBTs are WAY better than Thyristors because )obviously) you can't turn them off whenever you want. So let`s say the projectile is in the middle of the coil right now and you would like to cut the power off for the coil it wouldn`t work because your Thyristor is still open (and will be until the capacitor is nearly fully discharged) and so you got a huge loss in velocity because of the resulting suck back (when you power the coil when you do not want i to). So IGBTs are the better option, because you can switch them on and of whenever you like and so you are also able to only get the "best part" of the discharge curve of the capacitor (3taus or somtehing like that) and then quickly turn the power for the coil off. But to all the people who want to build a coilgun: DO NOT forget the flyback diode (antiparallel to the coil) especially when stopping the coil while feeding it high current from the capacitor and than rapidly stopping it, because then the inductive kickback will fry your switch and probably even your capacitor by putting a VERY HIGH negative voltage across it and with a diode next to it, which can of course handle that much current. (Even though a IGBT is about 4 times as expensive as a thyristor when you compare them with their maximum current ratings, but you can easily put multiple of them in parallel to increase the maximum current you can switch with the circuit). Ps: I hope that he makes a new video about that topic soon with new thoughts or even some experiments (hopefully also about an induction coilgun ;) )!
Instead of driving an hbridge with an mcu, I’m looking to just make a basic oscillator with two mosfets, I’m still relatively noob so I think I’m just going to attempt with a fluorescent tube light driver I salvaged from a copy/printer machine. Love your channel and I know this an older video but I really enjoy your work!
Tattoo machines use electromagnets that use a physical switch to operate. Brushed DC motors use brushes and wires connected to pads as a switch. I know you know this as well but have you considered using the projectile to make a connection with the optimal coil as it travails through the barrel. Or hall effect sensor like a brush less.
I have. Brushless motors are undesirable and expendable because of their brushes. Brushes in a system like this would like get chewed up immediately. Hall effect sensors are used in BLDC motors to detect the proximity of the permanent magnets. In this system there are no permanent magnets and it is likely that the sensors would be overwhelmed by the magnetic fields of the adjacent coils. In my latest video I discuss why I believe a blind system, one with no sensing whatsoever, could work just fine.
Ay bro I know all those wires look cool and feel big brain but you need to get some single strand insulated copper wire and cut little staple looking pieces out of it and use that for your connections, always keep everything as close as possible but practical enough to get in there and move things if need be. As for the circuit board, DEFINITELY make one bro it feels and looks way cooler than breadboard, you'll feel like a pro just go to easyEDA.com it's free and online and they work with JLCPCB so you can get cheap as boards but legit quality I've made and ordered heaps man they ain't payin me to say this, I've been doing this stuff for years and they're a game changer But I love how passionate you are it reminds me of me so best of luck bro you're gonna learn so much stuff you didn't even know you didn't know 😂😂
I like how you use metal pliers and hold onto the metal as you discharge the capacitor
So as you charge the capacitor up because E = 0.5CV^2 the inductor keeps seeing higher impedance from the cap. Thus because the time constant = L / R it discharges much faster. Now the pulses on the discharge shrink to a tiny percentage of the on cycle and thus if you don't adjust your duty cycle according to this trend you end up charging very little as the cap voltage increases. Hope this helps, I like your project, watching with interest.
Don't give up on the single inductor boost converter yet. Current through an inductor increases with time when a DC voltage is placed across it according to V/L = di/dt. likewise when the transistor is turned off it decays through the diode and into the capacitor following the same equation. The voltage which sets the rate of decay will depend on the capacitor's voltage plus the diode drop. The switching frequency and duty cycle must be chosen so the transistor is not on long enough to saturate the inductor and it should be off long enough for the current in the inductor to decay into the capacitor. Another way to think of it is the energy stored in the inductor at the time the transistor turns off is Q=1/2L(i^2) which is also the amount of energy that can be stored in the capacitor after the transistor turns off. Energy added to the capacitor will be Q=1/2C(V^2). With multiple cycles the capacitor should charge to any desired voltage within its rating. Time requred for the inductor current decreases as the capacitor voltage increases as can be seen in the above equations.
Hey mate! I have also been working on a coilgun in the past months. I have a few suggestions you might find useful.
1. The classic single mosfet and transformer circuit is much better and effective that boost converters. I've initially tried to make a boost converter for cap charging, but the charging was slow, and I got 200V max.
2. You can use a pair of LEDs facing one another as your motion sensor. Basically one will be on, and the other will go to the analog of the arduino. I programmed my arduino in such way that if the voltage drops (object is in the coil), it triggers the switching mechanism.
3. As for the coils, use thick wires and aim for almost non existing resistance. It's best that the cap dumps all current at once.
4. Thyristors seem to be the best option for stage switching, I fried countless mosfets so far, and relays are too slow.
Can't wait to see your creation!
@Isaac Eda It's basically a circuit where there is a transformer with a center tap winding, one winding gives the control signal to the gate of a MOSFET or IGBT, the other delivers the current through inductance.
boginjr.com/wp-content/uploads/flysche1.png
The schematic above is similar to what I've made, only instead using an IGBT. Also I've soldered a zener diode between gate and emitter.
czcams.com/video/0jWpQ8HEYgk/video.html&ab_channel=LudicScience
That video covers it up very well
@Isaac Eda Sorry this is actually the better video
czcams.com/video/q9iSQSR9SHM/video.html&ab_channel=LudicScience
I think (and have read it) that IGBTs are WAY better than Thyristors because )obviously) you can't turn them off whenever you want. So let`s say the projectile is in the middle of the coil right now and you would like to cut the power off for the coil it wouldn`t work because your Thyristor is still open (and will be until the capacitor is nearly fully discharged) and so you got a huge loss in velocity because of the resulting suck back (when you power the coil when you do not want i to). So IGBTs are the better option, because you can switch them on and of whenever you like and so you are also able to only get the "best part" of the discharge curve of the capacitor (3taus or somtehing like that) and then quickly turn the power for the coil off. But to all the people who want to build a coilgun: DO NOT forget the flyback diode (antiparallel to the coil) especially when stopping the coil while feeding it high current from the capacitor and than rapidly stopping it, because then the inductive kickback will fry your switch and probably even your capacitor by putting a VERY HIGH negative voltage across it and with a diode next to it, which can of course handle that much current. (Even though a IGBT is about 4 times as expensive as a thyristor when you compare them with their maximum current ratings, but you can easily put multiple of them in parallel to increase the maximum current you can switch with the circuit). Ps: I hope that he makes a new video about that topic soon with new thoughts or even some experiments (hopefully also about an induction coilgun ;) )!
I just found your channel. Really cool stuff.
Instead of driving an hbridge with an mcu, I’m looking to just make a basic oscillator with two mosfets, I’m still relatively noob so I think I’m just going to attempt with a fluorescent tube light driver I salvaged from a copy/printer machine.
Love your channel and I know this an older video but I really enjoy your work!
Tattoo machines use electromagnets that use a physical switch to operate. Brushed DC motors use brushes and wires connected to pads as a switch. I know you know this as well but have you considered using the projectile to make a connection with the optimal coil as it travails through the barrel. Or hall effect sensor like a brush less.
I have. Brushless motors are undesirable and expendable because of their brushes. Brushes in a system like this would like get chewed up immediately. Hall effect sensors are used in BLDC motors to detect the proximity of the permanent magnets. In this system there are no permanent magnets and it is likely that the sensors would be overwhelmed by the magnetic fields of the adjacent coils. In my latest video I discuss why I believe a blind system, one with no sensing whatsoever, could work just fine.
Ay bro I know all those wires look cool and feel big brain but you need to get some single strand insulated copper wire and cut little staple looking pieces out of it and use that for your connections, always keep everything as close as possible but practical enough to get in there and move things if need be.
As for the circuit board, DEFINITELY make one bro it feels and looks way cooler than breadboard, you'll feel like a pro just go to easyEDA.com it's free and online and they work with JLCPCB so you can get cheap as boards but legit quality I've made and ordered heaps man they ain't payin me to say this, I've been doing this stuff for years and they're a game changer
But I love how passionate you are it reminds me of me so best of luck bro you're gonna learn so much stuff you didn't even know you didn't know 😂😂
Consider just buying the cheap 12v to 220v high frequency AC converter modules from eBay? They worked wonders for me and goes to 350v dc rectified.
Custom circuit board is easy as using dremel.
Look More of these
the full H Bridge gives + & - V to the primary so doubles output voltage
For your circuit issues in the future, Strip board:
en.m.wikipedia.org/wiki/Stripboard
w