Hi

I have a design problem that I cant solve. I have a negative supply (-1000VDC) that I need to switch on and off through a TTL level PWM signal. The power MOSFET I have picked is this STW9N150.
I have been experimenting with the power mosfet I have (which is n channel mosfet) and noticed the following:
1- I have to connect the gate of the mosfet to the source (the negative supply), otherwise the mosfet will never switch.
2- when I connect the gate to ground or positive voltage, the FET turns on, the only way I found to switch the mosfet off is by leaving the gate FLOATING (not connected to anything) which is not a good practice..
How do I turn this mosfet switch off?

(the load I am driving will draw about 2A, the current will be limited by the supply).

I attached an image of the seed circuit that I need to drive.

Thanks in advance.

 

Hi

I have a design problem that I cant solve. I have a negative supply (-1000VDC) that I need to switch on and off through a TTL level PWM signal. The power MOSFET I have picked is this STW9N150.
I have been experimenting with the power mosfet I have (which is n channel mosfet) and noticed the following:
1- I have to connect the gate of the mosfet to the source (the negative supply), otherwise the mosfet will never switch.
2- when I connect the gate to ground or positive voltage, the FET turns on, the only way I found to switch the mosfet off is by leaving the gate FLOATING (not connected to anything) which is not a good practice..
How do I turn this mosfet switch off?

(the load I am driving will draw about 2A, the current will be limited by the supply).

I attached an image of the seed circuit that I need to drive.

Thanks in advance.

Why do you open a second thread ?

 

If the PWM signal and the -1000V share the same reference it cannot work this way, independently where you put the load resistance.

However, if they don't share the same ground the PWM reference will be either connected to the -1000V and source terminal (in your schematic) or it will be simply connected to the source terminal if you put the load resistance from source to -1000V.

 

You connected the gate to ground? You smoked it. The gate can not survive 1000 volts Vgs.
Next, this transistor has 8 ohms of resistance when on. That's twice the resistance of the load. Bad choice.
Next, 2 amps through 12 ohms from 1000 volts. Something wrong with the E=IR equasion here.
Next, with that circuit, you must have the PWM generator referenced to the -1000 volts because the transistor only cares about volts from gate to source.
And finally, you used the wrong symbol for the transistor. I had to look it up and here's the datasheet.

 

You connected the gate to ground? You smoked it. The gate can not survive 1000 volts Vgs.
Next, this transistor has 8 ohms of resistance when on. That's twice the resistance of the load. Bad choice.
Next, 2 amps through 12 ohms from 1000 volts. Something wrong with the E=IR equasion here.
Next, with that circuit, you must have the PWM generator referenced to the -1000 volts because the transistor only cares about volts from gate to source.
And finally, you used the wrong symbol for the transistor. I had to look it up and here's the datasheet.

The transistor has 1.8 ohms on resistance, and the total current draw will be set by the power supply. And the pwm is referenced to the same ground as the power supply.
The thing is that I don't know how to turn this transistor off, with the source connected to the negative supply..

 

Why do you open a second thread ?

Oh, just because I thought this is a different issue that I have, and has different schematics and description.

 

Ay

You connected the gate to ground? You smoked it. The gate can not survive 1000 volts Vgs.
Next, this transistor has 8 ohms of resistance when on. That's twice the resistance of the load. Bad choice.
Next, 2 amps through 12 ohms from 1000 volts. Something wrong with the E=IR equasion here.
Next, with that circuit, you must have the PWM generator referenced to the -1000 volts because the transistor only cares about volts from gate to source.
And finally, you used the wrong symbol for the transistor. I had to look it up and here's the datasheet.

About the symbol, I didn't have the symbol with the drain source diode, so I just chose the closest nmos that I have available in my library..

 

And the pwm is referenced to the same ground as the power supply.

However yo cannot the load it cannot work if the PWM and the -1000V power supply have the same reference (the ground symbol in your schematic).
The power supply you use for the PWM generation and the -1000V power supply would need to be isolated (separated) power supplies.

 

However yo cannot the load it cannot work if the PWM and the -1000V power supply have the same reference (the ground symbol in your schematic).
The power supply you use for the PWM generation and the -1000V power supply would need to be isolated (separated) power supplies.

Ok, so if I achieve that (isolate the power supply and pwm generator) how can I drive the mosfet to do the switching?
Can you please give me a simple schematic to s me how the switch will be controlled?
Thanks

 

Well if the PWM and the -1000V are not related then you tell me why you need -1000V.

"-" with regard to what voltage/reference?

I suspect your load is not just a simple resistor and your -1000V somehow needs to be related to the PWM generator.

Why don't you tell us what your project is about, what else will be connected to the load side etc.?

If your load is a simple resistor and the -1000V are not related to anything else I don't see why you need -1000V in the first place. Just measuring the other way around also gives you a negative voltage if for example you apply +1000V to your resistor.

 

Well if the PWM and the -1000V are not related then you tell me why you need -1000V.

"-" with regard to what voltage/reference?

I suspect your load is not just a simple resistor and your -1000V somehow needs to be related to the PWM generator.

Why don't you tell us what your project is about, what else will be connected to the load side etc.?

If your load is a simple resistor and the -1000V are not related to anything else I don't see why you need -1000V in the first place. Just measuring the other way around also gives you a negative voltage if for example you apply +1000V to your resistor.

Well, as I indicated in my previous thread, I need to switch between a positive and negative power supplies. The negative and positive supplies are separate, so is the pwm generator.
The idea is to deliver +10 V for 20% duty cycle, and -1000V for 80% duty cycle, at 100Khz frequency. That's why I need two individual supplies.

Thanks

 

See if this circuit works for you. Q2 and Q3 have to be 1000V rated.

 

Like Praondevou, I have also wondered about what is being done in your circuit. Nothing I can come up with makes any sense. To have two widely different voltages on a resistor is very puzzling. Why the secrecy?

 

See if this circuit works for you. Q2 and Q3 have to be 1000V rated.

Thanks for the circuit.
Can you please tell me where I connect the load?
I assume R3 (1kOhm) is the load for that circuit..

Also, V1 is shown as a -1000V DC source, does that mean the positive side of the symbol is actually -1000V and the negative is the ground? (I attached my interpretation of the circuit)
(does that mean the drain of Q3 is -1000V in respect to Q2 source?, or is it the other way around?)
Please let me know if my assumptions are correct.


Thank you again

 

Like Praondevou, I have also wondered about what is being done in your circuit. Nothing I can come up with makes any sense. To have two widely different voltages on a resistor is very puzzling. Why the secrecy?

well, it is not about secrecy, its just that this circuit will be part of a bigger device, and I dont entirely know how it is used, except the specs that I mentioned before (Resistive load with resistance of 4ohms, plus the availability of the two power supplies that I mentioned)

 

R3 is the intended load. Since there is no common ground between the control circuit and the load, you can attach your high voltage supply anyway you want, negative or positive side to the Q3 drain. It will change only the direction of the current flow.
Your assumptions are correct. It will work.

 

R3 is the intended load. Since there is no common ground between the control circuit and the load, you can attach your high voltage supply anyway you want, negative or positive side to the Q3 drain. It will change only the direction of the current flow.
Your assumptions are correct. It will work.

Does it hurt the functionality of the circuit if I have a common ground between the logic that drives the circuit and the supply being switched?

 

Public apology: I missed the 1 on the 1.8 ohm label.
Sorry about that.

 

Public apology: I missed the 1 on the 1.8 ohm label.
Sorry about that.

That's alright #12..

 

Yes it will, the only way this circuit will work if your HV supply has separate ground.