Hey guys! For a little project of mine I need to control high voltages (400V, DC) with an arduino, preferably the Nano. I need to have a current stored in capacitors dumped into a coil somewhat precisely (time-wise), so my first thought went to a MOSFET. After looking at the price ratings for 500V 40A (expected maximum current) MOSFETS, I was looking at thyristors. But "switching" those off seems to be a little difficult.
So I decided to go for IGBTs, 600V, 80A rating. Should be enough. It "switches on" by applying positive voltage to the gate-pin and by therefore "charging up" the capacitive gate, allowing electricity to pass from source to drain. But before hooking anything up, especially with such high voltages (obviously safety precautions will be met before the first run) I thought about the electronics of actually switching that thing in general.
So obviously you also have Gate, Source and Drain with an IGBT. As far as I understood you can "switch off" the IGBT by connecting Gate and Drain with a resistor that lets the capacitive Gate discharge relatively quickly by pulling gate negative (if you can say it like that).
So far, so good.
But: My thought wouldve been, that I could take the Arduino and let it switch a second, smaller transistor. This one will control a 12V power line. That 12V power line will need to be connected to the GATE from the IGBT. So if the Arduino "activates" the small transistor , that one will let those 12V flow to IGBT-Gate and switch it on, too. Obviously, the Arduino needs to be connected to the small-transistors drain aswell, since you could not switch the transistor without it. And there's the issue.
MY WORRY: I need to connect the 12V-Gate to the IGBTs Drain aswell - and that thing has 400V sitting there. That must be a problem, right? Or have I gone absolutely mad right about now? If it is indeed problematic, how could I overcome that? I would like to stay off of relais since theyre a little slow and bulky for my application. I am normally familiar with electronics, Ive just had way too many projects for some time and start to mix up knowledge as it seems.
So I tried making a quick diagram to help showing my issue. Please ignore polarity and the fact that this may be the worst circuit diagram (if you dare even calling it that) anyone ever made. I'm just a little in a hurry right now.
A MOSFET is switched by its gate-source voltage. It has nothing to do with the drain.
Please show a regular schematic, showing actual component symbols and their I/O labels and all the wires, so we can see what you are actually considering.
Playing with high current 400V, possibly not being knowledgeable enough to do that safely for yourself and other people who may interact with your device.
I'm going to ignore the obvious safety aspects here and merely point out that if you are really driving a transistor at those power levels, you need to use a proper gate driver and not just an Arduino pin.
Of no interest to anyone, but I just noticed that the last name of the author of that pdf is the same as the person I got most of my high-power drive knowledge from! I wonder if they're related.
As much as I appreciate your concerns, let me just say that this is an experiment. Noone will likely interact with it, its just to try something out. Everything will be isolated and properly handled, discharged correctly and monitored. Nothing more but also nothing less. Let's please focus on the technical aspect, shall we?
So resistors and diodes will be evaluated and placed, but this should basically cover the principle of how I imagine this to work.
Trying to formulate my question a little different:
Whenever I want to switch a transistor via gate, I also need to connect the switching circuitry to the (lets keep it simple) emitter. But wouldnt it be an issue if the controlling circuit is of much lower voltage (lets say 5V with an Arduino) than the controlled circuit?
EDIT: I know, the symbol for the IGBT is not correct and the transistors polarity seems wrong-way around. Software issue with Fritzing.
EDIT2: And no, there won't be an actual switch in a 400V circuit. It's a step up module, and the switch will turn that power supply on or off as needed - and that would be 24V.
I think I would use some form of opto isolation to define a line between HV and LV bits and then physically separate the parts , for safety if nothing else .
400 v DC can easily kill .
You can use a thyristor, it will switch off when the capacitor has discharged and the coil attempts to dump the energy stored in it back to the capacitor.
It used to be common to build electronic ignition systems that way.
Actually, this shows such little understanding of electronics, that I have to tell you to please stop now before you harm yourself. If that doesn't sit well with you, I will address it to the audience at large, don't play with high voltages unless you fully understand electricity and its associated risks.
One of the techniques I try to practice (not always successfully!) is minimal involvement. Make a simple suggestion and if OP seems interested in learning more, I'll elaborate.
That said, I find that no one ever takes me up on my suggestions to use proper gate drivers. Everyone finds them unnecessary until they blow up a few expensive FETs.
I had an open mind until the last schematic posted. That just shows that the necessary skill is absent. If it was a 12V circuit that would be fine, this one can easily be lethal. Also when I see this kind of pretense, I doubt any pleadings about how careful and safe they will be.
Sorry, but you obviously do not know enough about circuit design to be doing this project, and the voltages involved are absolutely deadly. The schematic in post #8 is hopeless.
Have a professional design and build the circuit. No free rides on this one.
Absolutely! An opportunity to learn more is always greatly appreciated. Unfortunately, some people rather spend their time diminishing OPs that clearly don't know every little fact about the matter - why else would they/ I come to a forum like this? And I dont mean you by this - thank you a lot for your suggestion, I will take a look into FET-Drivers in the next few days!
It's always easy to tell someone how little they know, but explaining it to them to make it work seems like too much of an effort. Sad.
Hey, thanks for the info. Probably will. But since this is a forum and I wasn't planning on doing anything before I have a solid plan of what I am doing in the end, I just wanted to clarify: The diagram was, is and will not be the "final" structure - it was to show how I wouldve hooked up the IGBT, Transistor, Fet, Mosfet or whatever else one might hook up there. Hence no diodes etc. It was not a question about how you like the project or if you find it too dangerous for a person you dont know the slightest bit - it was about transistors. About the mechanics behind it, and how I can get it to work safely. I want to understand what I am doing to the fullest, which is why I come here asking in the first place. So thanks really, for nothing.
