CRAZY coilgun without massive capacitors part 1

[@Gyro1]

Just a little correction. In the section about the MOSFET I have mentioned That the gate to source voltage needs to be 20V. That is actually the maximum voltage. In fact the 10V would be more than enough for this particular MOSFET. You can find it as graph named typical output characteristic in datasheet where VGS is the gate to source voltage. Thanks everyone for letting me know.

[@Azagro]

24V is not just way too much, it's actually destructive, for the 10V it may be enough, but not good enough. i.imgur.com/9UwFDSt.png Check this graph, it comes from the IRF3205 Datasheet (use datasheets when dealing with high power or high precision). 15V seems to give the best Id at 15V. Also be careful with the pulse duration: i.imgur.com/XbvDCJJ.png More info on MOSFET and power drives on my channel if you wish to operate safe and effective.

[@Gyro1]

Thanks

[@cyarks7468]

why didn't you use an opptokoppler at the gate to trigger you would be able to trigger the oppto with the Arduino and the oppto triggers the batterie to the gate

[@Azagro]

The same reason you don't use opamps for gate driving.

[@thegoodhen]

I think a better solution would be to use a logic-level MOSFET. For instance, STB80NF55L can be triggered by a voltage as low as 1V (If I'm reading the datasheet correctly) and can pass 80A continuous. The cost is under $2.

[@calronncardin2155]

Can I power the gun using a dc-dc step up converter instead of the li-po batteries? I'm wondering. If c Could perhaps attach a 3v volt supply to the converter (the one I have converts 3v to 32v) to work, or do I need to purchase li-po batteries?

[@neutronenstern.]

could you go without arduino if you just did some easy hardware programming? should be faster tho

[@davidarapkilel4942]

hallo brother...which software is that you used for designing the electronics

[@dpidcoe]

This is a pretty nice gun you have, and a very good job on the explanation and video editing. Last year I built exactly this but scaled up to half inch projectiles, 48v, and a teensy 3.2 microcontroller (there's some video of it on my channel if you want to see it). Some thoughts on what you have here: - I recommend putting the IR sensors in front of the coils instead of behind them. That gives you more time to respond to the input and switch the coil on or off. This becomes especially important as you reach higher muzzle velocities. - I saw you calculated peak current from the batteries, but I didn't hear anything about your coil resistance. I'm sure you know this already, but the maximum the gun will demand from the batteries will be I = V/R, with R being the resistance through your coil. Make sure that your batteries are specced higher than that number (and leave some copious safety margin) to avoid issues like premature (and possibly spectacular) battery death. - A flyback diode across the terminals of the batteries may also be a good idea. I know with my gun I was seeing a fairly substantial negative voltage spike hit the batteries after a shot. It was extremely brief so I have no idea if it was damaging them, but putting a diode across that cut down on it substantially. I don't know about yours, but the batteries are the most expensive part of my gun so anything that will make them last longer seems like a good idea. - As your mosfets get bigger, a simple pulldown resistor isn't going to be enough to turn them off quickly. I learned this the hard way after I set a $30 mosfet on fire because it remained in a half-on state too long as it was turning off. What you want is a gate driver. You could look up gate driver circuits and build your own, or you could buy some off the shelf ICs that meet your spec (I highly recommend the latter option). One of the nice things about gate drivers is that you can buy ones that are optically isolated from the low side, which means your microcontroller is safe in the event of something going horribly wrong on the high voltage side of things.

[@Azagro]

Defintely right on the gate drivers. You really require dedicated drivers in order to get a good switching transition time. Now you're charging the gate at a very low current, therefore a very long turn on time. The IRF3205 has a total gate charge of 146nC and if you want to turn on your MOSFET from infinite Rds to 8mOhm Rds at e.g. 50nS you need a driver that does 146nC/50nS = 2.92A. Your OP AMP isn't going to deliver this amount of current without a collapsing waveform, transients, etc. Definitely without a gate resistor. Another thing that makes me shit myself, is that you apply 24V to the gate of the MOSFET. This MOSFET is definitely going to blow up one time. The very maximum you should apply is given in the datasheet as 20V. Even then, in the graph you can see 15V is as good as 12V and 15V gives you probably the best on-state.

[@richardtaylor9227]

Very nice. I'm looking forward to the next build video for the coil gun.

[@dalenassar9152]

To make a sharp point on the projectile, put it in a drill,as you did, but use a metal file--not paper. This always works great for me, as does a bench grinder with a 45 degree holder. ONE QUESTION: How do you cut of the magnetic field when the optical sensors say it is time to do so??? THANKS FOR THE -GREAT- VIDEOS!

[@timovneerden]

Try cooling the coils, with dry ice for instance: this will reduce the resistance in the coils and create a significantly stronger propulsion.

[@knoxx5553]

Prototype form looks cool dude

[@davypeleman3672]

about the mosfet 20V is the maximum voltage. you should look more below on the datasheet for the threshold voltage that is the voltage needed to put the mosfet on I would say 15V to be on the safe side and not blow the gate, but 10V will suffice

[@vijayendran912]

Hi gyro.. i am not able to get any output after assembly.. can you please share the circuit diagram for the full circuit.. my presentation at college is the day after tomorrow.. would be really grateful if you can share ??

[@Gyro1]

There is a link in the description with instructions, schematics and the code

[@michaelcomte2598]

I would like to see more, sorry if you posted more about your work. but i have some quick thoughts that well i would first like to see if you worked them out yourself yet. but over all loved the coverage showing everything that worked or didn't. is how life is and some many of tubers hate showing error(s) that they edit them out and forget to talk about somewhere other then in the description, then wonder why people think its fake. so hats off and I am happy to have subscribed.. I look forward in watch other posting later..

[@DualDesertEagle]

What's the music that starts around 4:20 and why do I have to ask instead of finding that info somewhere in the video?

[@Magneticitist]

This should actually be a good relative measure of how to build one using caps. If the high current cap discharge is not sending the projectile farther than a pulse straight from some higher resistance batteries then something is wrong. The energy it seems to put into the bullet seems to be pretty good though. I guess enough lipos and there would be significant bang to reduce the need for a cap charging circuit.

[@sspence65]

Why not use a logic level mosfet like the IRL540? They use a 5v gate voltage.

[@fredlllll]

as another commenter already said, you dont have to (and shouldnt) put 20V on the gate for the mosfet to fully turn on. notice it says absolute maximum ratings above the section where you have the 20V from. if you scroll down in the datasheet you will likely see some curves, and some describe the used Vgs for the test. in most cases that is 10V. so you should go with 10V for the gate. also for mosfet switching, the more current you can pump into the gate, the better. the gate is like a capacitor, so the faster it is charged, the quicker the switching. for the isolation issue there are thermal pads specifically for use with mosfets.

[@stevehawley5618]

Why not use a ULN2003 (or something in that family) to drive the mosfets? They take logic level triggers. I'm currently using them to drive small solenoids directly, but nothing as beefy as your coils.

[@Gyro1]

+Steve Hawley I'll be swapping the MOSFETs in the next one.

[@Derek_Read]

It might seem obvious (so maybe you've done it already) but if you move the screw mounts on the 3d printed frame for your coils to the outside of the frame they'd be out of the way and it would be easier to do the winding because they would not be in the way.

[@Gyro1]

+Derek Read You're right. However there needs to be a flat surface to 3d print it from

[@Derek_Read]

Ah, OK. What if you made the sides longer by adding about 1cm (so the entire thing is initially taller by 1cm), and exclude your molded screw mounts, but include a horizonal "scored" area along the outside of each side (maybe 1mm tall that's about 1/2 thickness of each side -- this "score" would then be about 1cm up from the new bottom). Then you could heat that slightly and bend it 90 degrees outward? This extra 1cm of material would then have the two holes molded into it and they'd be in the right location after you do the bending.

[@KevinSidwar]

This is amazing. Thanks so much for sharing. Really love the homemade tool solutions. I find my self doing that stuff all the time.

[@jcschadt]

you shouldn't need a zip tie to have the drill running. that's what the orange button on the side of the grip and the dial on the trigger are for. pull the trigger, press the orange button to lock it on then adjust the speed with the dial on the trigger.

[@Gyro1]

+jcschadt I feel so dumb. Thanks for letting me know.

[@cypherf0x]

The coils are done quite nicely. The electronics need to be cleaned up and I didn't notice a flyback diode on the coils themselves. The MOSFETs need their own, and most have built in ones, but taking care of the surge at the inductive source is a best practice. As for your projectiles try and find some ball bearings that'll fit your barrel. They're easier to build a gravity fed magazine for to give you multiple shots as fast as you can pulse the coils and feed it.

[@Gyro1]

+cypherf0x the ball bearings would be really easy solution but the way infrared sensor is placed I must be using rods.

[@AlexandrWord]

Good job! Not bad gauss

[@ashvinla]

What is the kind of currents that you would see on the MOSFET? How long is the pulse on the coil?

[@Gyro1]

Around 50A for couple of milliseconds. In case something goes wrong the coil won't be on for more than 500ms.

[@streamware7746]

Why dont you just make a cone coil instead you wouldnt need to turn it off

[@mpvader1]

instead of using two 12V batteries is it possible to power the coil with a 9V battery connected to a step up converter that increases the volts to 24?

[@Gyro1]

No. 9v battery isn't able to provide enough current.

[@mpvader1]

Gyro I want to make a version of this and put it on a quadcopter... the batteries seems to be the heaviest part... do you have any suggestions on how to decrease the weight?

[@Gyro1]

Well you'll need huge quadcopter 😀. In this case using traditional design with capacitors would probably be better. You can use small battery to charge them up to high voltage and then fire. But still the coils themselves are quite heavy

[@mpvader1]

I calculated that a 0.8mm copper wire with a length of 7 meters (what I think is enough for one coil) will have a resistance of about 0.23 ohm, with this resistance one 12V battery will supply 52A which sounds a lot to me... yours had two of these batteries and the result was OK but not amazing like you would think from 100+ Amps. I'm trying to think how to improve performance without dangerous 400V capacitors. Thank you for all your help you are amazing! :)

[@gorillaau]

Nadav Friedman You can get Lipo batteries which have surge discharge of 30C or higher... where C is the capacity of the battery pack. So for a 4AHR battery you get a discharge of 120 Amps. Try Hooby King for what's out there. Best of luck with your research.

[@fredlllll]

also, could you measure the current spikes with one of those big 100A current shunts? i would love to see how much they pull. and test if adding a normal electrolytic cap can increase performance

[@Gyro1]

+fredlllll thanks I'll have to fix that part. I'll also try measuring the current

[@ScottPlude]

does it make noise when it fires, other than the projectile moving quickly? It seems like it would be silent...

[@Gyro1]

+Scott Plude No it's very silent. it's just the projectile hitting things that makes noise

[@Gyro1]

+Scott Plude no it's very silent. The projectile makes noise flying all over the room but that's ok

[@dingsens2810]

Hey great project, but you got something wrong about the mosfet gate voltage. the +-20v is the absolut maximum this mosfet can handle before it gets a short. the layer between the gate and the terminals is only a few atoms (overdramatisation) thick and will be destroyed at voltages above 20v. you need to look for the gate threshhold voltage, it is usually much lower commonly between 1v and 10v and a dcdc converter is actually a bit op. there are many many possibillities to accomplish the goal with less expensive parts. for example use a pnp to vcc with two resistors in series at the collector (remember pnp goes to + with the emitter) to the gate and then a z-diode between gate and source to limit the voltage. now you take the first resistor and calculate it to limit the current to a reasonable value (say 75% of the pnp's max pulse current) and the second to set the current to a few mA. then you put a ceramic capacitor parallel to the second larger resistor. (between 10 and 100nF) with this you can deliver a really high current to charge the gate very fast, but the current doesnt have to flow the whole time. then you need an npn so you can switch the pnp. additionally aou can take another npn that shorts the gate (with the same double resistor and one cap circuit) to discharge it quick. be careful with the timing, so you dont have both transistors switched at the same time. you better insert a dead time.

[@Azend_1]

Music at 14:30 ?

[@Gyro1]

Vicetone - I hear you

[@Azend_1]

@@Gyro1 Thankyou

[@vijayendran912]

hi.. is there any possibility to replace the arduino with any other device like atmel or something???

[@Gyro1]

Sure. Any microcontroller can work

[@vijayendran912]

@@Gyro1 thank you gyro.. i am making this for my 3 rd year mini project.. it's great that you reply to all your admirers who seek help regarding the project.. god bless you.. if you ever visit india please drop a message..

[@divita5717]

how about using relays for controlling the coils with arduino?

[@Gyro1]

With such a high current you'll have trouble finding suitable relay. The contacts will just weld together

[@Rubysh88]

It gave me an idea for video, since steps seem to matter more than just raw magnetic power, why not do a performance comparison between more steps with a shorter coils vs fewer steps with longer coils, both with the same length of wire (the shorter ones will simply have a larger diameter). An example could be a test with 2 guns, one with a 3 steps with those coils in the video and the other 6 steps with coils that are half the width, keeping both guns with about the same "barrel" length.

[@Gyro1]

+Ruben Fernandes that would be interesting to see

[@NikyNine]

Like the Pylot track behind ;)

[@nikhilmadhusudananpillai4659]

can u give a circuit diagram for this

[@louis2760]

Very good Video, when are you planing to upload the video with the finished Project? Very good work :)

[@Mb0ts]

in 2 weeks maybe. I'll see how fast I can finish it.

[@bloodaid]

Where did you get that music from at 9:40

[@Gyro1]

+LongJohn Vllasaliu Lakey inspired - fast lane

[@jotham123]

What C rating do your batteries have? I should think 4 or 6 low mAH high C rated batteries in parallel with a few more coils could make this shoot a lot faster.

[@jotham123]

I found a 850mah, 130C = 110Amp peak current battery on hobby king for 12 euro. If you have 6 .... 660amps. This should work quite well!! EDIT: looking further I found a 3S 5000mah cell 130C burst that'd deliver 650amps peak.

[@Gyro1]

My batteries are 65C So that is around 90A and they have even higher peak current rating. It's actually not a great idea to put the batteries in parallel since the coil will draw 40-50 amp(Based on my calculations). You're better of putting the batteries in series for more voltage.

[@WildfoxFabrication]

*TRIGGERED*

[@manickn6819]

Not bad. Ramp it up and see if you can get some more power in. The other thing is checking the power in and energy of the projectile.

[@monicagarcia3721]

Hey! I tried to download the 3D printed part but i coudn't would you mind sharing it again? please!

[@Gyro1]

The link is working for me. Did you try it from here www.thingiverse.com/thing:2481600 ?

[@monicagarcia3721]

Yeah thank you! turns out it was just my computer 😋

[@vasimleegpc8983]

Is it in mm or inch?

[@uniko78]

Where i can get the coil can i use coil from transformator?

[@Gyro1]

If you mean the wire then yes you can you a wire from a transformer. Otherwise it's called enameled wire. You'll have to wind it yourself.

[@stalker989]

will rotating the projectile at high rpm increase velocity

[@Gyro1]

+ENGLISH RESISTANCE I have no idea. Probably won't make much difference. It might decrease the friction it the barrel maybe.

[@gametabulas]

Nice i am currently building one without capacitors and will try to push further. Nice work man.

[@Gyro1]

Thanks man. good luck with your gun.

[@leif-erikhallmann]

when will you finish the enclosure? i'm waiting for it... my teacher asked me to build this as a school project and i want a nice enclosure too :)

[@Gyro1]

+FPVSprint Music I have it done. Electronics are giving me some problems at the moment. Once I have it working I'm uploading

[@leif-erikhallmann]

Nice ;) Thanks for the extremely quick response!

[@Gyro1]

+FPVSprint Music ;)

[@xbing61]

you got a new subscriber because of this video thanks for sharing

[@elijus9889]

Is this insulated copper wire ?

[@Gyro1]

If you mean the coil wire, yes it is. It's enameled wire.

[@Gyro1]

If you mean the coil wire, yes it is. It's enameled wire.

[@thecharlesfickle3647]

What program was that for the schematic?

[@thecharlesfickle3647]

Nvm i just cant read. Its EAGLE right?

[@Gyro1]

+TheCharlesFickle yeah

[@dablet]

how far can it shoots?

[@Gyro1]

+dablet00 not sure. I will test it.

[@tomasz5315]

Great project, but LM358 is one of the worst op amps which you could use in this project, it has poor slew rate so swithichg the mosfet is slow. Better but not the best would be choice of TL072 which has push-pull output stage and much better (13 V/μs) slew rate tham LM358 (0,3V/μs) so overall efficienty would be better.

[@trevorc.4350]

High voltage is anything over 1000V medium voltage is anything in-between 1000v-24v

[@НикитаЦыпанов]

Maybe, maybe. I also have a new video: coil gun new test kzbin.info/www/bejne/in7WfXx3otqohZI kzbin.info/www/bejne/Zp-wqqluaMSsma8

[@matatisan]

you can use a simple npn transistor. this is much easier. I did it in my coilgun too, look at my channel, on the video of my coilgun there is a diagram

[@kgtc]

I used magnet wire reel forectly as the coil :c

[@theflamecoreguy7929]

It's not powerful enough

[@НикитаЦыпанов]

Of course. Hi. I also have a new video: powerfull coil gun new test kzbin.info/www/bejne/in7WfXx3otqohZI kzbin.info/www/bejne/Zp-wqqluaMSsma8

[@krunomaricak2497]

pleas send me code for arduino and sheme

[@Gyro1]

It's all in the link in the description

[@peterhofmann9784]

I made a 4KJ Coilgun this is funny ;)

[@Gyro1]

pics or didn't happen :D

[@peterhofmann9784]

Ok! I'll propably upload a vid about this Coilgun. I built a two stage coilgun with 8x450v 4,5mF Capacitors. For the 450V I built a boost converter and it works just fine! (Sorry for my bad english im german) Actually my Coilgun is still in work because I need to put some safety circuits in it. Every time after two shots my mosfet blew up. So im adding a temparature sensor ;) And yeah. Your coilgun is pretty insane because you didn't use ANY capacitors nice work!

[@Gyro1]

Can't wait to see it :)

[@giannivergine1156]

Gyro Same!

[@killthesource4740]

Gyro this is my seconds channel i wanted to say that i almost finished it

[@JeremyCook]

Very cool!

[@markhodgson3045]

make a 240 x 3300 mf @100vdc bank and connect it to it

[@jcwdenton]

Cool project, Im up to make one myself but Im thinking more about coilgun-railgun marriage :3 (życz powodzenia) ;)

[@gameselectronics2560]

Ten deň ked si tu mal len 54 zhliadnutí a cca 30 odberatelov :D A video je veľmi dobre prepracované ;)

[@Gyro1]

+Andrewz! Electronics & More I got lucky. Thanks

[@klab3929]

are you swedish? xD

[@Gyro1]

+Lukas Bendiksen nope :)

[@klab3929]

damn it where are you from then?

[@Gyro1]

+Lukas Bendiksen Eastern Europe :D

[@klab3929]

nice im from northern norway :D

[@tariklive4233]

Очень не аккуратно и криворуко