High Voltage Inductive Load Switching with MOSFET

You may want to try

http://www.eevblog.com/forum/index.php

This link is for EEs, electronic technicians, hobbyists types.

Thank you. Your comment and advice appreciated.

You are free to use the CR4 search feature to try and find information already existing here. Since nobody performs a tight control of the topic titles or in which forum a topic will be discussed (you asked your question in the General category instead of Electrical Engineering) you may not find this discussed here.

Using a MOSFET to drive an inductor is very rudimentary electronics. My preference is to also use a Schottky diode to allow the fly-back current a discharge path. Depending on the amount of core material that will become magnetically charged, you might wish to put a loading resistor in series with this diode to speed up the quench cycle. You'll have to make sure that you do not exceed the breakdown voltage during this time shortened fly-back period.

Working at 1000VDC is not rudimentary work. Since you ask for assistance that includes rudimentary electronic concepts I am apprehensive to guide you to work at these voltages. If you're lucky a mistake in the design or implementation of a circuit at such a voltage will only damage some expensive components. If you are unlucky then death from electrocution or fire is very possible. It may not be only you that dies, too.

Good advice. OP should check that the diodes have sufficient voltage and current ratings, stand well back when switching on, and expect a few bangs, which are almost inevitable, and very bad for the "nerves".

Thank you for your comments.

I get chills...though I know my distances are good...whenever I test.

It's a good chill...

Ok. Thank you for your feedback.

At this voltage (you didn't mention the current), an IGBT is better suited. You will need the freewheeling diode in parallel with the load. I suggest using a half bridge module with gate driver already produced by a reputable manufacturer such as Infineon, Powerex...

A good candidate is Infineon http://www.infineon.com/cms/en/product/power-modules-discs-and-systems/igbt-modules/igbt-modules-up-to-1600v/1700v/igbt-modules-up-to-1600v/1700v-dual/FF225R17ME4/productType.html?productType=db3a304419c6f396011a00ad12bd286f

This isolated module will give you about 250A of capacity with a proper grounded heatsink. That is about as safe as it gets. You will also need a main capacitor and a snubber capacitor mounted on the IGBT. This is to absorb the inductive kickbacks from the wiring inductance. Look at Cornell Dublier or Illinois capacitors for candidates.

With the gate driver http://igbt-driver.com/products/scale-2-plug-and-play-drivers/2sp0115t

For about $600, you will have a complete assembly that only requires 15V to operate. No isolated supply needed. Keep all wiring as short as possible. This is as simple as it gets but there are still many details that will bite you. Buy a few spares..

That being said, be careful. This is dangerous stuff. Cover everything with a protective shield as these do explode when used improperly. At these voltage, the explosion is similar to shooting a shotgun. Protect your ears and eyes. Never work alone on high voltage. Make sure a competent person look over your circuit before powering up.

Thanks for the feedback. The operating voltage needs to be in the range of 1000 Vdc to 1200 Vdc at currents in the range of 10-20 mA - for a max of about 24 Watts (1200 at 20 mA). If the supply is at 1000 Vdc and 10 mA the dissipation will be in the order of 10 Watts. Obviously, here power dissipation is not the issue but the applied voltage. Need to switch at low speed - 100 Hz maximum, so very, very slow. The trigger should be optical. The coil will be air cored so the inductance is not that large.

I am aware of the protection required and the use of free-wheel diode (better to use Schottky if possible) to guard against back emf issues. Was also considering an available gate driver. Could you perhaps refer me to usable application notes for some practical driver circuits for high voltage switching to inductive load?

Forger about Schottky at these voltages. Read about what is available on the manufacturers sites. They have all the applications notes that you will ever need. I will not do the work for you.

I don't have much experience at such a low power. You should look at the manufacturers I suggested above for much smaller devices. They do have optically coupled gate drivers.

Regards

Never expected you to do any work for me !! Just thought you could refer me to a website or some application notes. Thanks for the help.

You mentioned that this is an inductive load. If it was a purely inductive load (in theory) there would be no real power dissipated by the inductor. Your inductor will store and release energy. Your fly-back load will be dissipating energy. With an air core inductor, all of this stored energy gets placed into free space. This does have the advantage that core saturation and hysteresis non-linearities are impossible. However, unless the coil windings are superconducting there will be a noticeable parasitic series resistance that will be robbing power from entering and leaving free space through the inductor. I suspect that once you calculate your anticipated low inductance value and include the series resistance from the wire gauge and length you use and then look at the phaser impedance you'll find that you have made a wire wound resistor and not an inductor.

You should also remember that if you are trying to impose a 100 HZ square wave that this has many odd harmonic frequencies higher than your fundamental frequency. A rule of thumb for producing a square wave is that the first five odd harmonics are the critical frequencies.

I could be wrong here but you give us so little information to work with here. Since you are new to the forum, I have nothing to go on your abilities but the sparse information you've provided in this thread. The voltages and currents you are pondering are lethal and easy to cause things to ignite when a design fails. Be careful.

Thank you for your reply and interest in my problem and questions.