IGBT Switching Not Working

[mikepicco]

I am building a switching circuit using an IGBT. What I have is a 120VAC circuit that goes through a full-wave bridge rectifier. An IGBT (collector - emitter) is placed between the + and - terminals of the rectifier. I am using a 10 Ohm gate resistor with a 12VDC source for the gate. The turn off voltage is 0V. To test the circuit I am simply tying the gate (through the gate resistor) to either +12VDC or gnd (0V) with a double throw switch. The IGBT always is in a conductive state regardless of the gate input. Later it is my intention to switch the gate at 120Hz (very slow). Any ideas why the IGBT will not "turn off"?

Thanks,

Mike

 

[Mr.Cool]

umm... are you trying to short out the 120V bus?

 

[nashwan]

hi
wat atype of load u have its resistive or inductive
i think u use circuit like this tray to add 10k resistore
between gate and emitter take care ur work with
not isolated circuit

 

[nashwan]

hi
wat atype of load u have its resistive or inductive
i think u use circuit like this tray to add 10k resistore
between gate and emitter take care ur work with
not isolated circuit

 

[mikepicco]

nashwan,

The circuit that you are showing is exactly what I have except I do not have the 10K resistor. I will try that this morning. Thanks for the help. I will let you know if it works. Why is the 10K resistor needed?

Mike

 

[nashwan]

hi
when u toggle switch from 1 to 0 or 0 to 1 there is small time in ms
for that small time the gate will be float its better to pull down with resistor
and if ur switch damage the gate still in off position with that resistor

 

[mikepicco]

Well I'm back with an issue for the next step. I am now trying to control the aforementioned circuit with a PIC. I have been trying to use a mosfet dirver to do the task. I am using a Microchip MCP1405 driver to be the "switch" for the IGBT. The circuit is the same as originally described but instead of a double throw switch, 10K pull down resistor and a 10Ohm gate resistor; I now have the mosfet driver with the output tied in series with a 500 ohm gate resistor (want to drive the IGBT slower). The 500 Ohm resistor is then connected to the gate of the IGBT. The driver uses a 12VDC source and the input to the driver is a 3.3V output from the PIC. All of the grounds (3.3VDC, 12VDC, and the rectified 120VAC) are tied together with the IGBT's Emitter, the driver's ground and the PIC's ground. All seems to go well except there is now a mysterious sinewave on the 3.3VDC bus. It goes away when I disconnect the IGBT and the recitified 120V ground. The sine wave is centered at 3.3V (max 3.9V / min 2.8V). When I put a scope probe on the 3.3V bus and reference the probe to ground I am able to see the sinewave. It is almost a perfect sinewave that is obviously dervied from the 120VAC source. What I don't get is why the sine wave is there and how I go about getting rid of it. It is causing the PIC to have fits.

thanks in advance,

Mike

 

[FvM]

I wonder, what the GND symbol at the IGBT E terminal means? The control circuit must be floating, no connected to an
external circuit except to an isolated power supply. Surely, you can't connect an oscilloscope probe to the circuit, if
the oscilloscope isn't supplied by a safety transformer (or you have a high voltage differential probe).

 

[mikepicco]

You are correct about the probe. I am using a high potential differential probe to view the waveform. Any ideas what my problem is? This small segment of the circuit has been giving me fits from day one.

Mike

 

[FvM]

In this case, I see two possible explanations (there may be more)
- the 3.3V power supply isn't working correctly. Where get you these voltage from?
- there is no sine wave. It's a measurement artefact generated by your differential probe, that has insufficient common mode rejection.
Do you see true zero, when measuring Gnd to Gnd?

By the way, did you solve the "shorted IGBT" problem from your original post?

 

[mikepicco]

First of all sorry for taking so long to respond. I had to go to the office (lab) to test your sugestions on the circuit. Anyways, I think you are on to something with the probe issue. I was NOT using a diff. probe for the 3.3V portion of the circuit. I was just using the diff. probe for the line voltage protion. After your comment I connected both the (non diff.) probe and the ground lead of the probe to ground and what do you know the sinusoidal signal was still there (with both leads essentially tied together). When I use the diff. probe on the 3.3V portion of the circuit the sine wave goes away. It must have something to do with the grounds of the dc sources and the rectified ground of the AC source and the way my scope references things. The good news is the circuit works with 3.3V as the input control. Tomorrow I will try and control the circuit with the PIC!!

thanks again!!!!

Mike

 

[FvM]

As I mentioned earlier, a safety transformer would be needed to connect the circuit to a standard single ended probe.
The safety transformer should supply the load rather than the oscilloscope for safe operation. But it's still possible
to observe interferences from capacitively coupled currents in the measurement.