Im using a N-Channel High Voltage Mosfet IXTL2N450 to charge a High Voltage capacitor at 2kV between the Source and the ground however it is not working. Im trying to charge it on the high side so after I switch the mosfet off I can take a charge reading from the capacitor. Is this possible? What do i need to do this?

VDSS: 4500V
VGSS: +/-20V
VGS(th): Min 3.5v / Max 6.0v
VGS: 10v
RDS(on): 20Ω

 

To use an N-FET on the high side you need to generate a gate voltage higher than the voltage you are switching- by about 10V

There are several ways this is done- but much simpler to use a P-FET

 

What circuit do you have that's "not working"?

You need need an isolated gate driver that can tolerate 2000V.
One example is an isolated photovoltaic optocoupler such as this.
They can be connected to generate a voltage at the MOSFET gate-source to turn it on in response to the isolated input control current.
That device is a dual unit and you may have to connect the two inputs and outputs in series to get sufficient gate voltage to fully turn on the MOSFET (10V minimum desired).

Don't confuse this with the typical optoisolator which simply causes an output diode or transistor to conduct when it is turned on.

If you wanted to do this quick and dirty, you could use a small 10-12V solar cell to provide the isolated gate-source voltage.
Just make sure the cell is well isolated from ground since it will be sitting at 2kV of potential.

 

There are several ways this is done- but much simpler to use a P-FET

But you then also need a ground side switch that can tolerate 2kV.

 

There are several ways this is done- but much simpler to use a P-FET

@Sensacell
I wish I could find one, I was not able to find a P-Fet rated =>2kV

If you wanted to do this quick and dirty, you could use a small 10-12V solar cell to provide the isolated gate-source voltage.
Just make sure the cell is well isolated from ground since it will be sitting at 2kV of potential.

@crutschow
I tried to use a 9v isolated battery however i didnt do it right. II ran the 9v+ and Negative to the gate. (btw, I am testing things at only 12v for now. My other powersupply is 2.5kV at only 5mA)

Power Supply #1: DC, 0-30V, 0-15A
Power Supply #2: DC, 0-2.5KV, 0-15MA

 

@Sensacell
I wish I could find one, I was not able to find a P-Fet rated =>2kV



@crutschow
I tried to use a 9v isolated battery however i didnt do it right. II ran the 9v+ and Negative to the gate. (btw, I am testing things at only 12v for now. My other powersupply is 2.5kV at only 5mA)

Power Supply #1: DC, 0-30V, 0-15A
Power Supply #2: DC, 0-2.5KV, 0-15MA

If I understand, you connected positive of the battery to the source and the negative to the gate. That is backwards. Connect the positive to the gate and negative to the source. Of course, when you go to >1000V remember that the battery will be at that voltage too relative to whatever you are calling 0 V. How will you switch it?

 

Im using a N-Channel High Voltage Mosfet IXTL2N450 to charge a High Voltage capacitor at 2kV between the Source and the ground however it is not working. Im trying to charge it on the high side so after I switch the mosfet off I can take a charge reading from the capacitor. Is this possible? What do i need to do this?

VDSS: 4500V
VGSS: +/-20V
VGS(th): Min 3.5v / Max 6.0v
VGS: 10v
RDS(on): 20Ω

@Ezeikiel Rich
What value capacitor are you trying to charge? How quickly must you charge it? Will your application allow charging via the negative side (ground?) of the cap, while the positive side remains connected to the 2.5KV power supply?
View attachment 186242

 

If I understand, you connected positive of the battery to the source and the negative to the gate. That is backwards. Connect the positive to the gate and negative to the source. Of course, when you go to >1000V remember that the battery will be at that voltage too relative to whatever you are calling 0 V. How will you switch it?

What if i build a 4kV capable volatage divider to bring the 2kV voltage down to 9v. Then I use a opto isolator to triger the 9v on the gate? Would something like that work?

 

@Ezeikiel Rich
What value capacitor are you trying to charge? How quickly must you charge it?

The cap is ceramic
200nf//220nf limit
Freq 1 kHz
Charging Time: @300ms

Will your application allow charging via the negative side (ground?) of the cap, while the positive side remains connected to the 2.5KV power supply?

It needs to charge via the 2.5kV then disconnect and leave connected to ground to drain. I have already built the charging system at 12v and it works good, however i need to convert it now to 2.5kv using the Fet and now also a voltage divider with opto to trigger the gate at 9v. I hope this will allow me to to switch the power on the high side and the cap remain connected to ground.

 

Not sure the HV divider is necessary. A lot of optoisolators are rated at thousands of volts isolation (e.g., 20kV). Just be sure the "9V" is isolated and use such an isolator between your logic and it.

 

What if i build a 4kV capable volatage divider to bring the 2kV voltage down to 9v. Then I use a opto isolator to triger the 9v on the gate? Would something like that work?

No, because the gate floats with the source output voltage, so the output of the 9V divider will be reduced a low voltage as the MOSFET charges the capacitor.

But, you could use an opto isolator with an isolated 9V battery at the output of the isolator to drive the gate-source voltage (below).
Just be aware that the battery will be at 2kV above ground when the MOSFET is on.
The control In voltage can be from 5V to 20V to turn the MOSFET on.
The battery current only 90μA when the MOSFET is on, so a 9V alkaline transistor battery should last for at least a couple thousand hours of on time.

 

Thanks, Curmudgeon. I just didn't have time then to draw it.
John

 

No, because the gate floats with the source output voltage, so the output of the 9V divider will be reduced a low voltage as the MOSFET charges the capacitor.

But, you could use an opto isolator with an isolated 9V battery at the output of the isolator to drive the gate-source voltage (below).
Just be aware that the battery will be at 2kV above ground when the MOSFET is on.
The control In voltage can be from 5V to 20V to turn the MOSFET on.
The battery current only 90μA when the MOSFET is on, so a 9V alkaline transistor battery should last for at least a couple thousand hours of on time.
View attachment 186259

Crutschow, thank you kindly. You save me. I followed your schematic.

It is working! On the high side!

I used opto pc817 which I was able to get easily.
I also ordered one more Step Down Voltage Regulator so I will replace the 9v battery with it. I hope it will hold up against 2kV

Thank you Jpanhalt , Sensacell and TeeKay6!

When I wake up I will start finishing the rest of project thanks to your help!

 

Glad it worked for you.

Before you substitute the drop-down regulator for the battery, please post a schematic of that plan. The battery is an isolated power source. It sounds like your voltage regulator might not be so isolated.

 

I hope it will hold up against 2kV

Hope, seldom works in a technical design.
Even if your regulator is isolated, it likely is not rated to withstand 2kV.
If you can't stay with the battery as a permanent design (why?), then I suggest you use a photovoltaic isolator as I mentioned in post #3.

An alternate is to use a small pulse transformer that has insulation rated for a least 2kV.
You put an AC pulse into the input (say at 1-10kHz) and rectify-filter the isolated output to provide the DC gate voltage for the MOSFET.

 

Or you can use the solar panel as an insulated power source. Switch on the light and charge your capacitor.

 

Or you can use the solar panel as an insulated power source. Switch on the light and charge your capacitor.

As I mentioned at the end post #3, that's an option with a solar panel that will output around 10V or more.
The cheapest way may be to connect several small solar cells in series to get the needed voltage, such as these.

 

As I mentioned at the end post #3, that's an option with a solar panel that will output around 10V or more.
The cheapest way may be to connect several small solar cells in series to get the needed voltage, such as these.

I don't think I can argue with you. From #3 I understood that you were offering an optocoupler. And I meant the solar panel, which has a lot of elements and they are already connected in series.

 

I don't think I can argue with you. From #3 I understood that you were offering an optocoupler. And I meant the solar panel, which has a lot of elements and they are already connected in series.

This is for military application. I dont think they will accept the solar option. Industrial standard solution would be better.

Capacitor will need to be charged/discharged. 60 cycles per minute.

 

This is for military application.

Then I suggest you look at the voltaic output optocoupler I mentioned.
And I know that hope doesn't work when doing a military design.