HV Boost Converter Blew Up My Stuff...

I simply want to make a boost converter that took a 12V input and output 400V to charge a capacitor bank for a coil gun. I successfully made it, but after a while it broke suddenly. The 12 volt linear regulator I was using to power my Arduino Micro shorted the circuit and blew my main filter cap (which was too small anyway at only 1uF) and my Arduino was also fried. Then the voltage regulator just stopped doing anything after that, no voltage output at all... I tried replacing the voltage regulator just to see what would happen. It worked for a little while (I was using a separate square wave generator now) until the voltage regulator did something else weird. It now wasn't regulating at all, just letting the voltage pass through :o ... I don't understand what is happening. Maybe a low pass filter and a filter diode is needed?

The circuitry is pretty close to a normal boost converter using an N-Channel Mosfet, except I added in a 12V regulator before the inductor with the Arduino on the output. The Arduino was being used to generate the square wave to switch the transistor.

Thanks for reading and trying to help :smiley:
Tell me if you want more information.

Tell me if you want more information.

Schematic please.

Seconded.

Rail gun implies pulse discharge.

You may need a diode across the storage capacitor to prevent a reverse voltage being impressed after discharge.

This can destroy a polarised capacitor.

High voltage and current circuits can do unexpected things.

I don’t have fritzing installed… So here is a professional “Photoshop” (Paint) version of the schematic…

That only shows the upconvertor, it does not show the rest of the circuit you have connected to it.

Can you post a photo of your setup.

I don't have fritzing installed

Good keep it that way.

So here is a professional "Photoshop" (Paint) version of the schematic.

But no labeling, component values and the Arduino powering circuit. Also no supply decoupling, how was it laid out not bread board I hope.

Boardburner2:
That only shows the upconvertor, it does not show the rest of the circuit you have connected to it.

Can you post a photo of your setup.

I don't have it connected to anything. I was charging up the storage cap, then discharging it on random pieces of metal. It was fun while it lasted lol.

Grumpy_Mike:
But no labeling, component values and the Arduino powering circuit. Also no supply decoupling, how was it laid out not bread board I hope.

Inductor is about 720uH.
Storage capacitor is 180uF rated at 450 volts
Mosfet is a C3856
Diode is a PR3005

It was laid out on proto board (Radioshack, Had been pre-traced)

TehProgrammer:
Storage capacitor is 180uF rated at 450 volts

It was laid out on proto board (Radioshack, Had been pre-traced)

Proto board ,- bad news for switchmode supplies generally.

They need PCB with short traces.

Suggest you put a diode across the output like you would a relay.

Wires have inductance and a high current pulse (which you have) will generate back emf.

180 uF at 450 V would suggest an electrolytic.

A photoflash cap is needed, standard electrolytics are not rated for pulse discharge like this and are easily destroyed.

EDIT

Can you supply a link for your inspiration for this please ?

Standard covering my *** warning.

You are playing with LETHAL circuits here but i suspect you already know that.

Storage capacitor is 180uF rated at 450 volts

And do you know the ripple current rating?

I meant to say I am using perfboard that has been pre traced… Here is a photo of my setup:

Grumpy_Mike:
And do you know the ripple current rating?

I do not... I just pulled it from something.
All I know is what I told you, and that it's a Rubycon brand cap.
All other markings on the cap: "MXG (PET)" and "CE 105 degrees C" and finally "7948AW".

I looked up the last number thinking it was a model number and came up with nothing but some basketball thing :confused:

Rubycon cap , they are normally photoflash rated caps.

Depends what you pulled them out of.

Thy make normal stuff too so look up the part no.

Looking at your pix it appears that you have a socket for a nano.

I would suggest using separate boards for high and low voltage.

Sorry for the poor posts , im using a phone which is not that friendly to this forum.

Like I said you had no decoupling on that regulator. That makes it unstable and the layout doesn't help. But what finally triggered the capacitor to go was probable exceeding the ripple current.

Grumpy_Mike:
But what finally triggered the capacitor to go was probable exceeding the ripple current.

I think we've misunderstood each other. The cap that exploded was a tantalum 1uF cap I had across the power rails. The Rubycon is the storage cap and it's fine. I took the other one off the board..

The cap that exploded was a tantalum 1uF cap I had across the power rails.

Which wasn’t on your schematic, so how the hell are we supposed to know that?

There is also no ground on the Arduino in the schematic.

Schematics are the way you convey information, missing stuff out is just like lying.

Never ever put a tantalum capacitor across the power rails, it is a big NO NO.

Posting delays are annoying.

Tantalum caps are very unfriendly when it comes to voltage spikes and i strongly suspect this is the problem.

Grumpy_Mike: Sorry man. I forgot it because I took it out before I posted this. Plus I’m just forgetful in general sometimes. I didn’t mean it… Here is the corrected schematic…

On your corrected schemaic there is a diode before the storage cap.

What is its spec. ?

Can you remove a leg and test ifit is still functioning as a diode ?

Here is the data sheet on the diode: http://www.diodes.com/_files/datasheets/ds26014.pdf