Yet another coil gun prototype / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

Yet another coil gun prototype

first previous 6 7 8 9

Move Thread

LAN_403

DerAlbi

Thu Dec 27 2018, 06:29PM

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

StepUp: a step-up converter will be short circuited by any load that demands a lower voltage than Vin. So if you supply 12V to your Step-Up, your capacitor cannot be discharged below 12V or excessive current will flow from your 12V supply.

If your board goes bad from the shots, try to make a better build. You are lucky that the components are not blown - respect your luck and give your components a suitable work environment; they have earned their place with their spirit.

What you delivered was a layout, not a schematic. Schematics are the language to describe the circuit intention. Your schematic looks like this:

(drawn with LTSpice)
A layout is a physical representation of the circuit. As you can see, some stuff from your layout didnt make it into the schematic: Control and the -12V make no sense; -12V is in fact Gnd. :-/

If you see a short circuit between Source an Drain of the mosfet, first make sure that there is no gate voltage. A pull down resistor [2.2kR between gate and source (the outer pins of the mosfet)] would be good.
Also note, that you cant measure as long as the capacitor or coil are connected, because they force a voltage between Source and Ground that annoys your resistance meter.
Check your diodes too. Maybe they are already blown. If the Zener is shorted, it would create weird effects for sure. (depending on how high resistive your layout is)
If you measure, try to reverse the polarity of your measurement. Only if both readings are consistent the effect you see is actually a resistive short circuit. Anything else may be a junction.

Zuckerstange47

Sat Dec 29 2018, 12:40AM

Registered Member #61926 Joined: Wed Nov 22 2017, 04:08PM
Location:
Posts: 34

DerAlbi wrote ...

StepUp: a step-up converter will be short circuited by any load that demands a lower voltage than Vin. So if you supply 12V to your Step-Up, your capacitor cannot be discharged below 12V or excessive current will flow from your 12V supply.

I do not exactly understand this, is this connected to the reforming of the capacitors? When I did this, I directly connected the supply that can be tuned from about 0 up to 18V, and only used the step-up for the higher voltages. When charging to shoot the coilgun I of course used the step-up, but I stopped charging, when the multimeter showed the desired capacitor-voltage. Does this seem possible? (For that reason there is a push-button between step-up and capacitors, that does not seem to be included on the picture you made.)

Concerning the damaged circuit I realized that I forgot to remove the MOSFET when measuring. Now I know that the MOSFET is destroyed, again (but there is no short circuit on the boards soldering this time). Do you think this is due to the heat, or because the circuit is not good enough? I thought that the heat transfer would not be quick enough when the MOSFET is only switched on for 5 to 8ms at a time, so that a radiator would not matter. But after thinking about it, I got to the conclusion that I might be wrong and I have no clue about the speed of heat dispersal in a MOSFET.
I did measure the Zener, too - it seems to be okay. It starts conducting at about 6.5 V, just like a new one does. The diode is good, also.

The resistor beween gate and source is installed now, and I will care about the cooling next.

DerAlbi

Sat Dec 29 2018, 02:37AM

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

Your intuition concerning the cooling is correct. Any power component that actually gets hot in a coilgun is basically dead. The thermal capacity of the mosfet is enough to swallow the heat. If you need a heatsink, then something is wrong. (Except if you shoot in rapid succession, then you need some thermal considerations)

If the mosfet is dead, there are multiple reasons:
1) too much current (who would guess)
2) excessive avalance energy: high voltage peak after mosfet shutdown. You think that you have the free-wheeling diodes, but they do not work instantaneous. It takes time for a diode to become conductive and in that time the mosfet burns all the energy in the coil, until the current path is finally redirected. Slow diodes (and the P600 is a dog slow diode!) are not the only reason - the diodes, depending on how they are soldered/layouted form an inductor (with a single turn). Any inductive current path resist current change - so the sudden expectation that there should be flowing current through the diode path, is a bit too simple - the current needs to build up slowly.. and during this time, your mosfet avalanches (drain source breakthrough due to high voltage spike).
3) bad gate drive: if the gate drive is inductive (long wires to the gate), the gate may overshoot and undershoot during turn on and turn of. This can be so bad that the gate is destroyed (due to excessive voltage)! After all, the wires are an inductor, the gate is a capacitor - what you have built is an oscillating circuit. To suppress this, people use a gate resistor. Such resistor lowers the quality factor of the oscillator so it wont oscillate anymore. Your mosfets internal gate resistance is about 2 Ohm already, if you have really bad layout, you might add 4.7Ohm in series with the gate.
Gate ringing can also cause parasitic turn-ons after turn of (if the gate voltage oscillates so much that the peaks of the voltage waveforms are high enough to turn the mosfet on again). If that happens, the mosfet is extremely under stress.

Regarding the step-up: dont get confused. My main point was that you need to take into account the circuitry that is connected to your capacitor. If there is a switch that is always open when you shoot, then this is ok. If the capacitor charger is always directly connected, then there can be issues. (but not necessarily)

Zuckerstange47

Sat Dec 29 2018, 04:34PM

Registered Member #61926 Joined: Wed Nov 22 2017, 04:08PM
Location:
Posts: 34

I calculated the current and discovered that it is too high, indeed. Strangely the measured resistance of the coil is only about 50% of the one that I calculated before winding it, so that I underestimated the flowing current.
Now I shortened the wire that connects the gate driver to the microcontroller and added a resistance, but I will not shoot in this configuration anymore. I want to research about snubber circuits and rethink the whole thing first.

first previous 6 7 8 9

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