voltage measurement across IGBT

[khalid4145]

I want to measure the voltage across the IGBT to moniter the voltage if there is short circuit or not and then feed it back to microcontroller? what is the best way of doing that. Do i have to use a differential amplifier, isolation amplifier or voltage transducer. i am attaching the circuit. if you have any circuit to measure the voltage please try to post. the maximum voltage when the device is on is equal to Vce (sat) is 1.8V according to the data sheet. the maximum voltage when device is off 120V.

 

[crutschow]

That's a tricky measurement. You need some form of isolation from the 120V.

What short are you trying to monitor, the transistor shorting?

If you don't need high analog accuracy, you could use an opto isolator. It could perform a rough voltage check across the transistor to tell you if the voltage is high or low.

 

[Leftyretro]

I want to measure the voltage across the IGBT to moniter the voltage if there is short circuit or not and then feed it back to microcontroller? what is the best way of doing that. Do i have to use a differential amplifier, isolation amplifier or voltage transducer. i am attaching the circuit. if you have any circuit to measure the voltage please try to post. the maximum voltage when the device is on is equal to Vce (sat) is 1.8V according to the data sheet. the maximum voltage when device is off 120V.

Humm, first you should varify by measurement the worst case voltage drop across the transistor, i.e. what is voltage drop when supplying maximum design current to the load, and what is the voltage drop when turned off, a short of course would develop 0v. with that in mind I suppose you could run two voltage divider strings (such that the A/D sees a maximum of 5v = 120v) driven from the emitter and collector leads and then run them through two A/D channels in the micro and compare them in software and take whatever action you have in mind.

Lefty

 

[Ubergeek63]

That's a tricky measurement. You need some form of isolation from the 120V.

What short are you trying to monitor, the transistor shorting?

If you don't need high analog accuracy, you could use an opto isolator. It could perform a rough voltage check across the transistor to tell you if the voltage is high or low.

pretty simple actually you put a mosfet to a cap as a sample circuit with the gate fed by a one shot triggered by the IGBT gate drive signal for a high voltage sample and hold.

 

[crutschow]

pretty simple actually you put a mosfet to a cap as a sample circuit with the gate fed by a one shot triggered by the IGBT gate drive signal for a high voltage sample and hold.

For that to measure the drain-source voltage on a floating transistor, you would need a flying capacitor sample & hold controlled by four high voltage (>120V) switches in a bridge configuration. Is that what you mean? If so that's not exactly simple to design or build.

 

[khalid4145]

That's a tricky measurement. You need some form of isolation from the 120V.

What short are you trying to monitor, the transistor shorting?

If you don't need high analog accuracy, you could use an opto isolator. It could perform a rough voltage check across the transistor to tell you if the voltage is high or low.

crutschow, Yes i would like to monitor the transistor shorting. i dont want high anologu accurcy, i just want to tell if the voltage is high or low. Actually my idea is to intrupt a fault in the circuit by giving the IGBT driver a dc signal of 5V from microcontroller (it is not a real fault), when it is on all the time , the voltage between the collector and emitter should be the saturation voltage Vce (sat) which is 1.8V. so if the microcontroller sees a continous of 1.8V then a short circuit occurs and start to activite a relay to disconnect the IGBt from the circuit. Does opto isolater do the work. if so, would you please provide me with data sheet suitable to my application?? do i have to add some circuits between the opto isolater and the uC? Any body have another ideas??

 

[crutschow]

The are many opto devices that should work for you. An example is the Fairchild FOD 852 photodarlington opto coupler. You would connect the opto control diode anode through a resistor to the IGBT collector and the diode cathode to the IGBT emitter (see attached View attachment circuit w opto.doc). The value of the resistor will depend upon the sensitivity you need but start with a value of about 30k ohm. That value will start to turn on the opto isolator when the IGBT voltage goes above ≈3V.

The opto isolator transistor output is turned on when the control diode is conducting and thus the transistor output to the µC is low when the IGBT is off. A short transistor fault condition wil be indicated by the opto transistor output being high when the IGBT should be off.

Edit: I forgot to ask what the operating frequency of your circuit is. That could affect the choice of opt couplers, since a photodarlington is rather slow.