My body became the gate IGBT resistor

Thread Starter

pastinsain

Joined Aug 25, 2012
163
OK I just tested my mg400q1us41 igbts. I used the 12v battery method

they work . connected as follows for the test

1) connected the collector to a 12v bulb load. then to battery 12v battery pos.
2) connected the emitter to the battery neg terminal.
3) then I touched the 12v pos battery post to gate turn on IGBT "light on"
4) then touch battery negative post to gate turn off "light off"

OK perfect. now the question. How long will the igbt stay on
without any volt connected to gate?

is there any decay in the gate curcuit with out any volts applied?
or will it stay on indefinitly?
another issue , I noticed if I did not touch the gate to 12v pos for a long enough time, the bulb was only partially on.
Thus I became the resistor for the gate "my body " no direct connection eccept though me.

thank you for the answer
 

shortbus

Joined Sep 30, 2009
10,045
Can't tell you how long it would stay on, but the reason it stayed on for a short while is that the gate and collector are in effect a capacitor. As the 'capacitor' leaks off voltage it turns off the IGBT. Some circuit use a high ohm resistor from gate to collector to make sure the IGBT turns off and stays off when it is not supposed to turn on. Stray inductance in the circuit can turn the IGBT on without this resistor.
 

Thread Starter

pastinsain

Joined Aug 25, 2012
163
So it would be ok to leave the 12v connected to the gate with no side effects?
In other words, The gate will only draw as much current from the 12v Battery as it needs to stay on? Or should a current limit resistor be in series from the Battery to gate?

Battery pos or neg ...>>> resistor...>>> gate "if so what ohm resistor is required?"
 
Last edited:

t_n_k

Joined Mar 6, 2009
5,455
Can't tell you how long it would stay on, but the reason it stayed on for a short while is that the gate and collector are in effect a capacitor. As the 'capacitor' leaks off voltage it turns off the IGBT. Some circuit use a high ohm resistor from gate to collector to make sure the IGBT turns off and stays off when it is not supposed to turn on. Stray inductance in the circuit can turn the IGBT on without this resistor.
Gate to collector resistor or gate to emitter? I would think gate to collector would tend to turn the IGBT on.
 

Thread Starter

pastinsain

Joined Aug 25, 2012
163
Lets look at three scenerios .

1)battery>resistor>gate When I used this method my body was the resistor 5meg dry hands give or take. Of course I would put a 5meg here
2)gate>resistor>collector "would a 5meg here keep the igbt on ?"
3)gate>resistor>emitter " same here "
why is one preferred over the other?

do all three keep the igbt continually on ?
then when I decide to turn it off ,touch the gate with a neg 5 or 12 volts
 

ErnieM

Joined Apr 24, 2011
8,377
1) Always on, the resistor is feeding voltage to the gate.
2) Always off, the resistor is feeding a low or voltage to the gate.
3) Always off, the resistor is feeding zero voltage to the gate.

In case 1, if you "touch the gate with a neg 5 or 12 volts" it will turn off for the duration of the touch, then come back on.

You need some sort of flip flop to preserve the state and actually drive the gate.
 

Thread Starter

pastinsain

Joined Aug 25, 2012
163
thank you, I always have to over simplify to understand .

So in case#1 I could toggle switch on / off this with 12v battery all day long and it would not harm the igbt? or overheat the igbt?

pos 12v on gate , neg 12v off "with no overheating"

Next step.. flip-flop or gate controller.
 

t_n_k

Joined Mar 6, 2009
5,455
In the case (2) instance - with a 5M resistor between collector and gate (with collector load powered from the positive supply and emitter grounded) one may get an unstable behavior. The gate capacitance would initially charge from the collector circuit [initially collector virtually at the supply voltage with IGBT off] enabling the IGBT to commence turn-on once the gate "VGE on" condition is reached. As the collector / load current then rises, the collector voltage would tend to fall, thereby reducing (reversing) the charging of the gate capacitance with a consequent reduction in gate voltage. One could imagine the collector voltage switching erratically near the VGE on condition, due to small voltage perturbations at the gate. Or perhaps a quasi-stable on condition might ensue, which biases the IGBT such that it is neither in saturation nor cut-off.

In any case it is not a method one should adopt in practice.

When setting up suitable drive circuits for an IGBT in the past, I have used a couple of techniques including employing dedicated MOS driver devices such as the TSC426-7-8-9 series IC devices. These are designed to source the initial high capacitive current impulse occurring at the rising edge of the gate-to-emitter drive voltage. If one wishes to obtain optimum switch off times, it may be necessary to provide some negative voltage from gate to emitter at turn-off, to more quickly sweep out the on-state gate charge.
 

shortbus

Joined Sep 30, 2009
10,045
Sorry, my bad, had "collector" on my mind from original question when meant emitter.
Here a is good link to IGBT driving - http://www.hyiq.org/Library/IGBT-Driver%20design.pdf

@pastinsain, Make things simpler for your self and use a IGBT/mosfet driver chip. You don't say if you are using the IGBT as a low or high side switch. A low side is easier in that you don't need a charge pump circuit to keep the IGBT on continuous.
 

shortbus

Joined Sep 30, 2009
10,045
If one wishes to obtain optimum switch off times, it may be necessary to provide some negative voltage from gate to emitter at turn-off, to more quickly sweep out the on-state gate charge.

I was under the impression that the latest type IGBT's don't have the need for the gate reverse voltage anymore. I know the early ones had a problem with "tail currents" and needed a reverse bias to shut off fast, but thought/read some where that problem was cured.
 

Thread Starter

pastinsain

Joined Aug 25, 2012
163
Im sure this is not the newer 5th of 6th generation IGBTs It an older MG series switch that could benefit from neg bias to trim shutoff times

the newer igbts are trench gate design that have the faster shutoff low trail current ect..

The silicon keeps getting smarter , soon they will have programable logic
or incorporate logic chips
 
Top