New to Coilguns - Coilgun Buildlog (Charger apparently fixed) / Electromagnetic Projectile Accelerators / Forums

Forums

4hv.org :: Forums :: Electromagnetic Projectile Accelerators

« Previous topic | Next topic »

New to Coilguns - Coilgun Buildlog (Charger apparently fixed)

first previous 4 5 6 7 next last

Move Thread

LAN_403

Artikbot

Wed Sept 29 2010, 08:54PM

Registered Member #3247 Joined: Mon Sept 27 2010, 09:42AM
Location: Spain
Posts: 137

Wow thanks!

But that is still more hard to wire and more expensive than the oscillator solution... So far my only issue is the drain>gate voltage flow killing the MOSFET, the rest of the circuit happens to work correctly :P

I appreciate your effort, if the oscillator project happens to turn out to be hard, I'll count with yours :)

Let's see what we got in my chart:

This is so far the simplest yet the best working circuit I've done:

Btw, I'm manually shutting down the charger once voltmeter reaches 330V or either the LED lights up.

DerAlbi

Wed Sept 29 2010, 09:07PM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

Cancle out the 47R. The NE555 can drive a gate directly.. OR at least make it smaller.
It can be that the capacitance between Drain and Gate couples the high voltage of the Drain into the gate (during fast rising edges) and the gate voltage is only "slowly" discharged though the 47R. In this time the Gate may blow due to overvoltage. An other solution could be a Zener Diode at the gate.

Btw: manual shutdown is a source of failure since it involves a human beeing. Keep in mind that your decission for the stepupconverter limits your outputpower extremly (Stepupconverters do not work verry effivient if the dutycycle is too high (but it must be high for your outputvltage) If it takes a minute to charge your cap, you definitivley start doing something different after a while

And.... boom

..only maybe...
My 300W charger is controlled and secured analog and digitally (overcurrent, overvoltage and inputvoltage failure AND it is capable of discharging the caps in short time when an error occurs) and it makes terrible noise when my 54mF are charged.. THIS is safe

Artikbot

Wed Sept 29 2010, 10:43PM

Registered Member #3247 Joined: Mon Sept 27 2010, 09:42AM
Location: Spain
Posts: 137

Your charger has a bunch of advantages over my design, but due to inavailability of parts I can't build it right now, a big pity :S (trust me, parts would have to be ordered from different stores from all over the world bumping the cost to the order of 60ish euros, kinda expensive :/ )

Yep I'll be ommitting that "manual control" part. I'll just put two 330V Zeners as a breakdown circuit as GhostNull sugested a few posts ago. But I'll keep the manual switch to serve the on/off purpose.

The MOSFET thing... Well now the overvoltage thing has been solved with a Zener, but now it is an overcurrent issue what kills it. This N-MOSFET is killing me, everything was working so fine with a P-MOSFET >.< but well, since N type actually leaks current from drain and gate, we got a problem there. P was like "isolated" not allowing current go back.

If I place a reverse biased Zener on the cap, the voltage gets properly limited. one issue solved.
if I place a reverse biased Zener on the drain>source of my MOSFET, it switches but oh surprise, no current goes through the inductor.

Oh god. The stupid mosfet driver is killing me. Okay time to go and have some sleep.

Thanks a lot :)

DerAlbi

Wed Sept 29 2010, 11:49PM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

Add a 1Ohm resistor in series to the Inductor. The thing is that current throught an inductor follows Imax*(1-e^(-t/T)) where Imax is U/R in your case 9V/RSDon (should be a large value) and T = R/L. R is again your RSDOn currently. And L is your given inductance. With the small L AND the small R you have a verry fast current rising.

Solve it:
1) Add a resitor in series as mentioned. This is in general recommended when simulating, because also your later inductor will have resistance that is measurable. For small coils it would be a value from 0.05 to 1Ohm.
2) Use a bigger inductor. This slows down the current so it will not reach the critical maximum within one complete cycle.
3) Increase switiching frequency - this shortens the time the current can rise.
4) well 4 is not 3+2+1, but try it.

5) It depends on your mosfet, but it _may_ not be good enough to drive the gate with 9V. Maybe the losses are to big, switch to 12V.

5.1) PLEASE use google. PLEASE find out that N and P-Channel Mosfets means and how they work. Its absolute senseless to write everything down for you. It is writen thousenads of times in the internet. PLEASE.
Of course a P-Channel was working different. In fact it was not really working (at least not as supposed).
PLEASE stop reading first when you understood why P-Channel it not usable in this application.

6) If you sleep at night an live over the day you will never become a good freak

Btw: since you are counting in euro, i think you will have no problems to get the parts. you are just too lazy to look for distributors in your count(r)y, USE GOOOOGLE!. Trust me.. if you are not from North Corea it should be easy

E.g. for 40eur you may buy everything you need at digikey, twice.

Thousands of people have built something like this.. it is well documented in the internet... really.. look for a schematic that is easy to build and then just do it. A "self" development without understanding is impossible.. and its even more impossible if you are not willing to spend time to understand.. AND that means by learning by _your self_.

Dont blame me, but this thread will end in 100+ posts and you will still try to use a P-Channel, as long as no one is telling you the stuff that could be found with 1min googleing.

No, that is not mean

This is true

GhostNull

Thu Sept 30 2010, 12:28AM

Registered Member #2648 Joined: Sun Jan 24 2010, 12:45PM
Location: Australia
Posts: 291

I think you other guys are being a bit misleading =S

I also really think you are relying on the yenka sims too much. You must remember it is more school science class education aid, not a proper electronics simulator.

And remember that the duty cycle is what most importantly determines the output voltage in a boost converter not so much the input voltage and frequency.

Another better idea for target voltage shut down would be to use a comparator or op-amp.

1285804941 2648 FT97387 Boost Converteri

Here is a the boost converter from rwilsford07 on instructables.
He has incorporated a target voltage shut down using a comparator (U1) and some voltage dividing resistors (VR1, Rdiv1, Rdiv2).

How is works: The inverting input, the input which turns the output low if it higher, is feed voltage from the voltage dividers (Rdiv1 and Rdiv2). These resistors "divide" the voltage to a lower potential/voltage is it can be compared by the U1 to the non-inverting input. The non inverting input is connected to the voltage regulator through VR1, this acts as an adjustable reference voltage. When the voltage across the capacitor, divided by the voltage dividing resistors, is greater than the reference voltage the output of the comparator goes low, disabling the 555.

In operation: Power is turned on, the 555 starts up and works as the driver for a boost converter. The boost converter charges up the capacitor, as it does so the voltage supplied to the inverting input rises. The circuit keeps charging up the capacitor until the voltage to the inverting input is greater than the non-inverting input. The comparator then outputs low, this activates the reset pin on the 555, stopping the charging and lighting up LEDREG

BTW: This a a good thread, concentrating the alot of knowledge and information. In future we could also refer future newbies to this thread but let's not get off topic

Edit, corrections:
It will NOT be like shorting through a transformer. The inductive reactance of the inductor (not transformer, two different things) and the high frequency means that the impedance will limit the current. A cap would cause it to not work AT ALL because the this outputting DC not AC. DON'T omit the gate resistor! This will cause ringing, which will mean inefficiencies and the MOSFET heat up and as for slow:
T=RÃ—C
Where:
T = time in seconds
R=Resistance in Ohms
C= Capacitance in Farads

assume gate capacitance is 10nF (about 10 times the normal)

T= 47 * 10^-8
T= 4.7 Ã— 10^-7
T= 0.000000047 Seconds
=47 Nano Seconds
Not slow, the rise time of the MOSFET is probably greater.

There is probably more misleading information here which I didn't pick up on.

Edit: ScotchTapeLord found another one!

Edit 4: Also remember that you power source needs to be capable of supplying enough current.

Edit 5: Sorry DerAlbi

ScotchTapeLord

Thu Sept 30 2010, 12:43AM

Registered Member #1875 Joined: Sun Dec 21 2008, 06:36PM
Location:
Posts: 635

The gate will not be charged to 300V. That would also destroy the 555 timer. The output of the 555 is either HI (9V) or LO (0V). If the gate was so easily charged by capacitive coupling to the drain, then MOSFETs would be *useless*!

9V is enough to turn a small FET on. You don't need a 40 amp FET for this as your average power cannot exceed your battery's maximum current output!

Damn the torpedoes; you're ready to build.

jaysun92

Thu Sept 30 2010, 01:23AM

Registered Member #2114 Joined: Fri May 08 2009, 11:58AM
Location: London, Ontario
Posts: 36

Here is the complete circuit i used for my coilgun. It works pretty well, charges my caps in about 15 secs.
]schematic.pdf[/file]

DerAlbi

Thu Sept 30 2010, 02:01AM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

You guys are compleately right. I will shut up

Sorry for being constructive by showing an other approach.

Maybe its also destructive.

Just a few things:
1) Gates indeed charge to damagable voltages. It depends on the internal capacitive voltagedivider in the mosfet. Just so some simulation with fast risetimes at the drain

It really works, and high gate resistors are not good for this situation. But sorry for that. Maybe it was a bit too deep. I was just looking for possible sources of the booom.
2) Of cource a transformer does not really short ciruit. But the current on the primay and secondary side is high.. and high over normal operation currents. Sorry it was mistaken considdering that as shorted.
3) GateResistor: keep the risetime of the NE555 in mind. its sooo slow.. sorry for that. in general a resistor is of course good. But since the software is nonsens it could have been a source of an error.
4) Current source: it belongs to the simulation problem. If there is no parasitic resistance.. oh guys.. do i really need to explain that..
5) A Cap in series with the transformer will of cource limit the current since it limits the energytransfer per cycle. In case of an error, no short ciruit would be possible. Sorry for that anyway.

I was just thinking that someone who will experiment with high voltage just should get some idea what he is doing - getting an idea of the simulation seemed to be a good way. If i am the person who is wrong..

hmmh

I am a bit disapointed that my help for the simulation was so misunderstood. Of course not all relates to reality..

But GhostNull is right! "His" (better jaysun92) kind of voltage limitation should be prefered!! I was hesitating to involve another IC

A optocoupler seems to be easier to understand (but indeed crap).

jaysun92: your schematics look cool.. but i see the questions comming about the values of the used elements. It also would be great to see a picture of the finished device so that starters get an imagination of what they are going to build

Artikbot: keep on trying! Its hard the first time, i guess its commonly known

I am sorry to shout at you so loudly that google also helps

Grettings
The googeling Albi

Artikbot

Thu Sept 30 2010, 11:00AM

Registered Member #3247 Joined: Mon Sept 27 2010, 09:42AM
Location: Spain
Posts: 137

DerAlbi wrote ...

5) It depends on your mosfet, but it _may_ not be good enough to drive the gate with 9V. Maybe the losses are to big, switch to 12V.

5.1) PLEASE use google. PLEASE find out that N and P-Channel Mosfets means and how they work. Its absolute senseless to write everything down for you. It is writen thousenads of times in the internet. PLEASE.
Of course a P-Channel was working different. In fact it was not really working (at least not as supposed).
PLEASE stop reading first when you understood why P-Channel it not usable in this application.

Trust me, I've red for HOURS about stupid mosfets, why P and why N but I still don't get the difference between carrying energy with holes or with electrons. I guess N channel is faster due to speed of transmission with electrons, but I don't get why one blows up and the other doesn't. Guess I gotta read more then :P Thanks for the advice :)

Well well, I got a lotta things to answer now... :P

Just a side note... I am willing to learn as much as I can, if you were on my side of the keyboard and you could see how many searches I did on Google and how many webs/wikis/schems/examples/whatever did I read... But hey, it's been three days now since I started reading all this stuff, three days before I did not know any noticeable difference between N and P, what a 555 timer was, why were Zeners useful, or ANYTHING related to understanding stuff! All I knew before was to calculate adequate resistance values for specific currents, to use several NAND to create an ultra-basic logical circuit and to use a transistor to amplify currents! I think I've made some nice effort on this, trying to understand all this stuff, trying to develop my own circuits (if I really wanted something easy I'd started by copying instructables' DCDC booster and case solved), trying to make MY OWN STUFF work! Y' know, I don't like to do what's already done, I want to develop it by myself, development for me means understanding (or at least, having an idea of what am I doing), and since I am starting the electronics degree at university, I need to start to understand what the hell is going on down on the boards because otherwise that is no place for me. Hey, don't misunderstand me, I'm not doing this for my university, all I do there is maths, physics, chemistry and computer structure :P
What did I want to say with all this? I'm a curious guy, a VERY curious guy. And this has caught my attention more than any other experiment I've ever done. That is why I am developing every single thing by my own (and ofc with you help guys :) ), that is why I am asking so much and redoing the same circuit all over again and again. Not because I am a lazy ass and I want everything done. If I wanted that, I've already said it, I'd just done the one on Instructables and issue solved.

About the distributors: I was only able to find three big stores on Spain (holy crap no one moves its ass here) and they have got a kinda limited inventory. I can go to the city 2h away from home and get the parts, yes, I can, but the trip is worth â‚¬40 and the lazy asses from that shop (I know they have the stuff) won't want to send me the pieces. just because it's too much hassle for them. On Spain they call "electronics store" to everything that sells audio, video and home appliances. Not to a pure electronic component shop.

^^^^
That's not a compliment, its just a side note for you to understand my situation better, don't get me wrong oki? ;)

@GhostNull:

Yep, I am basing 100% on Yenka, I know it's not the best (maybe it's not even close to good) sim around but it does the trick... If it says 100A are going thru my gate, I rise the top value for my MOSFET and then it says 3kV are going thru it, something's definitely not working. I'll try DerAlbi's solution and see what happens.
Okay, I tried what derAlbi said, the MOSFET no longer blows up and I think I've found the problem: 203W are trying to make it through the coil, and of course, the resistor and the diode blow up. Also seems the coil is driving excessive voltage to the capacitors (406V), maybe the calculator was wrong? Gonna try to do it again.

About the comparator: When I looked at uzzors2k's design I saw one, so I searched a bit how did it work. Seems it will work better than the Zeners and will be easier to use, since I can't get 330V zeners at a good price in my country. But I can get the comparator :)
It will also be more precise, and the best of it, adjustable. That means I can increase the overall voltage limits for my components and get a regulable output instead of a fixed one, which is indeed good :)
I'll try to strap it up on my circuit and see what it does :) I think I've udnerstood everything you've said!!

ScotchTapeLord wrote ...

The gate will not be charged to 300V. That would also destroy the 555 timer. The output of the 555 is either HI (9V) or LO (0V). If the gate was so easily charged by capacitive coupling to the drain, then MOSFETs would be *useless*!

9V is enough to turn a small FET on. You don't need a 40 amp FET for this as your average power cannot exceed your battery's maximum current output!

Damn the torpedoes; you're ready to build.

NOOOO DON'T SAY YENKA IS JOKING ME! O.O
So... No 400V will leak through the gate? Yenka is exaggerating (oh waht a surprise)? Oh well that relieves A LOT of suffering.

Concerning the FET: I'm just using things like 40Amp, 400V gate voltage and those crazes because yenka is trying to blow everything up!

Okay, making components indestructible the booster works. My cap is charged with 3kV on 3 seconds of simulation. Due to the 230W output, I guess, and the fact I'm not limiting the voltage input on the caps. Let's call the op-amp to come to the aprty and we'll see what happens :)

I got a problem. My power supply CAN exceed 40Amp... In fact it is capable to deliver sustained 52Amp and spikes (10 sec maximum) of close to 80Amp at 11.1V! (3cell 2650mAh 20C li-po battery) But I guess that with proper limiting it will be easy to control.

@jaysun: Thanks for the schematic, surely there'll be something useful from there :)

Ohay, hands to work. Let's put the comparator there and try not to blow it on my face XD

Thanks to everyone for your support!

Edit: @GhostNull: YAY! Comparator addition makes everything work as it should, now I gotta get the right resistor values and voilÃ :)

So far with a resistor in series with the inductor (Yenka doesn't count the component's internal resistance, so I manually added the 0.5ohm to the circuit) nothing does blow up now except the capacitor, but well, that's due to wrong comparator voltage divider values. Update: with the right values it still doesn't trigger the reset pin. Seems I'm doing something wrong.
Edit2: Yenka doesn't count in the timer duty cicle.. For this circuit it should be about 97% according to calculator, that's a pretty low efficiency, but meh. It's still okay, I don't care much about efficiency since I'm using low powers.

ScotchTapeLord

Thu Sept 30 2010, 08:28PM

Registered Member #1875 Joined: Sun Dec 21 2008, 06:36PM
Location:
Posts: 635

Parasitic resistance is pretty important. Your inductor is going to have a few ohms of resistance! Perhaps you should add that to the simulation... because even 1 ohm of resistance is going to limit your output current in the worst case scenario (the MOSFET shorts) to 9A.

With regard to the gate, if you want protection you can use a zener diode or TVS (uni or bidirectional). I have never used protection on this single transistor type of topology. My guess the capacitive dividing effect is dwarfed by other parasitic losses.

I've never seen a capacitor used in series with the inductor of a boost circuit. I don't think it's worth it as it would have to be rated for 300V (though under normal circumstances it would see very little voltage, but it was described as protection against shorts, so it would need the full voltage rating to fulfill that purpose) and have a fair bit of capacitance so as not to impede the current's rise time through the inductor.

first previous 4 5 6 7 next last

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob