High speed HV mosfet switch

This forum is for specialized infomation important to the construction and safe operation of the high voltage electrical supplies and related circuitry needed for fusor operation.
Post Reply
Andrew Seltzman
Posts: 815
Joined: Sun Feb 01, 2004 8:02 pm
Real name: Andrew Seltzman
Contact:

High speed HV mosfet switch

Post by Andrew Seltzman »

I recently bought a high speed 2kV mosfet switch on ebay capable of 10ns rise and fall time and down to 50ns pulse width:
http://hvswitch.com/HVSwitch_Products_Standard.html
SAM_5488.JPG
I x-rayed it to see the inner workings of the switch. There are 3x to220 mosfets in series, each with an independent gate drive. The mosfet gate drive is floated with respect to ground and found that it appears to be inductively driven via 2 transformers. Each mosfet has an independent gate drive. The interesting part is that the switch can remain in the closed state indefinitely. Does anyone have an Idea how they are accomplishing this?

My thoughts so far are:
1: One transformer generates an on trigger pulse, the other an off trigger pulse and the mosfets latch in the on/off state.
2: One transformer supplies power to a gate driver via an AC signal the other controls the driver via a on/off trigger pulse
3: The gate on/off is controlled by a phase shift between an AC signal on the two transformers
CIMG4237contrast.jpg
CIMG4235processed.jpg
CIMG4238processed.jpg
Andrew Seltzman
www.rtftechnologies.org
User avatar
Richard Hull
Moderator
Posts: 14991
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: High speed HV mosfet switch

Post by Richard Hull »

Low RGSon values can be found in a number of moderate voltage, fast mosfets. Series connections can be utilized, but issues abound and perhaps this was designed and tested at some level that warrants the claims made. Did they offer an effective realized RGSon? What is the rated power disappation? Inquiring minds want to know.

As most electronics folks know, low RGSon values are obtained at the expense of increased capacitance and sluggish switching issues. I can't imagine they have licked this in some significant way. The net RGSon will tell the story. This will also impact the dispation in a sealed epoxy block unless the bottom of the block is an exposed metal sink that can be bolted to an external sink.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
Dan Knapp
Posts: 402
Joined: Wed Aug 06, 2008 8:34 am
Real name: Dan Knapp

Re: High speed HV mosfet switch

Post by Dan Knapp »

This type of switch is used extensively in time of flight mass spectrometers. The established manufacturer is Behlke (http://www.behlke.com/). They have application notes and other information on the web site discussing the switches, which may be of interest to you. Interestingly, they also have another web site (http://www.hv-switch.com/) where they talk about a knockoff product that uses one of their old designs and warn about potential problems with this product. From the photo, it looks like you have purchased one of these.
Last edited by Dan Knapp on Fri Dec 02, 2016 2:47 pm, edited 1 time in total.
Andrew Haynes
Posts: 90
Joined: Sat Dec 01, 2012 1:25 am
Real name: Andrew Haynes

Re: High speed HV mosfet switch

Post by Andrew Haynes »

You can easily series mosfet,with a parallel forward voltage drop of diodes, they have to be less than Bv and together, based on npn enhanced type, and nh5001 diodes, 8_14v drop

http://nz.mouser.com/Search/m_ProductDe ... EYAg%3d%3d
Last edited by Andrew Haynes on Fri Dec 02, 2016 8:14 pm, edited 2 times in total.
Andrew Haynes
Andrew Seltzman
Posts: 815
Joined: Sun Feb 01, 2004 8:02 pm
Real name: Andrew Seltzman
Contact:

Re: High speed HV mosfet switch

Post by Andrew Seltzman »

Datasheet says http://hvswitch.com/StandardSpecs.pdf
15A peak current
0.79A steady state current
8.1ohm total resistance for the 3 mosfets in series
5W Rated dissipation

I found 2 relevant patents
https://www.google.com/patents/US8536929
https://www.google.com/patents/DE3630775A1?cl=en

And the comparison page on Behlke:
http://www.behlke.com/comparative%20test%20hts81.htm

It does seem that these are the knockoff, though I got it at a very good price and will be using it at much less then it's rated capacity(fast switching on/off of my ion sources, 1-2kV at a few 10s of mA to 100mA for 100us-1ms pulses). I'm actually trying to make a knockoff of the knockoff :) and am interested in learning how the gate drivers work. These are inductively driven gate drivers with minimal components that allow continuous on time of the switch(not just pulses).
Andrew Seltzman
www.rtftechnologies.org
Peter Schmelcher
Posts: 228
Joined: Sat Nov 13, 2010 1:56 am
Real name: Peter Schmelcher

Re: High speed HV mosfet switch

Post by Peter Schmelcher »

Andrew have a look at Analog Devices iCoupler family of parts. It is not a perfect fit but prototyping would be simple. The dual magnet gate driver is rated at 5kV for 1 minute however for a 50 year life it is only 1100V. At 2kV it won't be anything like 50 years. You might look for an application note on the topic or just ask them.
-PETER
Jerry Biehler
Posts: 975
Joined: Tue Nov 24, 2009 3:08 am
Real name:
Location: Beaverton, OR

Re: High speed HV mosfet switch

Post by Jerry Biehler »

Andrew Haynes wrote: based on non enchanted type
So they set up a semiconductor fab at Hogwarts?
Andrew Seltzman
Posts: 815
Joined: Sun Feb 01, 2004 8:02 pm
Real name: Andrew Seltzman
Contact:

Re: High speed HV mosfet switch

Post by Andrew Seltzman »

I've looked at a number of mosfet drivers with high isolation, 5kv for 1min is pretty typical, and many are not rated over 1kV isolation for any long period of time.

The advantage of an inductive gate drive like this is that a high level of common mode isolation(from ground) can be sustained(15kV in this case) and that the mosfet switch can be floated with respect to the control circuit ground. In addition, a large stack of mosfets in series can be controlled with a single driver(on the controller side) driving a wire passing through the transformers for all the mosfets on the switch side. This switch design is very versatile, since the floating switch allows the system to be neither constrained to a high side or low side application.

Any ideas on how the gate drive setup that they have here works and how it is able to drive the mosfets in a closed state indefinitely?
Andrew Seltzman
www.rtftechnologies.org
Andrew Haynes
Posts: 90
Joined: Sat Dec 01, 2012 1:25 am
Real name: Andrew Haynes

Re: High speed HV mosfet switch

Post by Andrew Haynes »

Just curious Andrew, but can't pull up or down resistor be replaced with secondary of a transformer, with a fixed reference to collector by resistor or zener make the desired voltages,
Etc a Npne the zener on the gate and emitter, and the transformer sec gate to collector.
You still have to protect against series mosfet not turning on at the same time and diodes would work better than resistor

DAMN SPELL CORRECTER
Andrew Haynes
Andrew Seltzman
Posts: 815
Joined: Sun Feb 01, 2004 8:02 pm
Real name: Andrew Seltzman
Contact:

Re: High speed HV mosfet switch

Post by Andrew Seltzman »

Sure, a fixed protection diode could be placed between the gate and source terminals to limit Vgs to acceptable limits, as in figure 35.
fig35.jpg
This setup is commonly used to isolate high voltage mosfet bridges as in figure 39, infact the same setup that is used in the spellman power supplies in another thread
fig39.jpg
The issue with these circuits is that such a circuit would be inherently AC in nature; the mosfet could only be turned on for short pulses, typically in even duty cycles. The interesting part of this switch circuit is that it achieves both rapid rise/fall times, high common mode isolation, and indefinite switch closure(a characteristic not typically common of inductively driven mosfet circuits)

Circuits with inductive isolation and active gate drivers on the mosfet side
fig37.jpg
and driving mosfets inductively for infinite on time do exist:
fig38.jpg
However from the xrays of the switch unit, it appears to have a very simple gate drive circuit and uses a pair of pulse transformers. Further the ability to synchronously switch a stack of mosfets on/off with 10 ns switching time is very impressive, and I'm trying to figure out what circuit they used on the mosfet side.

Figures from:
http://www.radio-sensors.se/download/gate-driver2.pdf
Andrew Seltzman
www.rtftechnologies.org
Peter Schmelcher
Posts: 228
Joined: Sat Nov 13, 2010 1:56 am
Real name: Peter Schmelcher

Re: High speed HV mosfet switch

Post by Peter Schmelcher »

The FET gate looks like a capacitor and if the circuit driving it has high impedance when not changing state the charge will stay put. One transformer puts the charge on the gate the other removes it. I think it can even be done with zeners and switching diodes and a single transformer. Create a 10 volt drop either polarity in series with the transformer and drive the transformer to have a +-20V output pulse. I would put a zener across the gate to protect it from over voltage damage.

I think the iCoupler has 3 pulse transformers per gate driver inside the ic. One just powers the output circuit and the other two are set reset flip-flop signals.

-Peter
Andrew Haynes
Posts: 90
Joined: Sat Dec 01, 2012 1:25 am
Real name: Andrew Haynes

Re: High speed HV mosfet switch

Post by Andrew Haynes »

This is what I meant if wanted, once the zener conducted it changes the voltage divider to always stay on, the only way to stop it conducting is a negative back elf which is made at the speed the zener switch off from Alachua other it stays on.
Attachments
tmp_22378-IMG_20161204_082225-1955745111.jpg
Andrew Haynes
Jerry Biehler
Posts: 975
Joined: Tue Nov 24, 2009 3:08 am
Real name:
Location: Beaverton, OR

Re: High speed HV mosfet switch

Post by Jerry Biehler »

Pretty hard to tell whats going on in the driver circuit. Willing to bet the mosfets are all matched before they are assembled into a unit. Pretty neat little unit. Sure beats using a giant vacuum tube like my e-beam power supply uses.
Post Reply

Return to “High Voltage - Fusor Input Power (& FAQs)”