How to remove spikes from IGBT drived by TLP250?

Thread Starter

Saeed Yousofi-Darmian

Joined May 7, 2016
8
Hi everyone,
I want to drive an IGBT with TLP250 optocoupler. The schematic of my circuit is attached. I use the output of signal generator as input of the TLP and a potentiometer between output of TLP and Gate of IGBT. The problem is, there is spike at rising edge of PWM signal on load. this issue gets better when I increase the gate resistance, but the time constant is increased in this case and its not good for me. I have attached some pics from oscilloscope. (blue color is output of TLP and yellow one is load voltage). How can I remove these spikes?
Any help is really appreciated.

Saeed
 

Attachments

Last edited:

Thread Starter

Saeed Yousofi-Darmian

Joined May 7, 2016
8
Actually I did not connect the collector of IGBT to load, but I will connect it to a RL load. Currently I want to have a pure square wave voltage between collector and emitter of the IGBT. The implemented circuit is exactly the same as schematic shown in Circuit.jpg with an exception and it is power supply of TLP. I used a transformer (220v/15v) and full-wave rectifier to supply the TLP.
 

Bordodynov

Joined May 20, 2015
3,025
1. Use a Schottky diode (as a limiter).
2. Use a transistor 2N7000.
In order that would be more correct to give you advice, we need more information!
 

Thread Starter

Saeed Yousofi-Darmian

Joined May 7, 2016
8
Dear Bordodynov,
I am using GP8NC60KD IGBT which is 600V/8A. currently I just want to know that my driver operates correctly or not?
what is the other information you need to know?
 

Kjeldgaard

Joined Apr 7, 2016
475
Hi,

A portion of the overshoot you are seeing, can be caused by unfavorable measurement points.

It is just right to use the short ground wires directly from the probes, but try to move all four measuring points right down to the transistor, possibly a pieces of component pins close to the transistor.

1N4148 is probably fast enough for a freewheeling diode, but I do not think it has big enough chip to clamp the freewheeling current.
 

Thread Starter

Saeed Yousofi-Darmian

Joined May 7, 2016
8
Hi,

A portion of the overshoot you are seeing, can be caused by unfavorable measurement points.

It is just right to use the short ground wires directly from the probes, but try to move all four measuring points right down to the transistor, possibly a pieces of component pins close to the transistor.

1N4148 is probably fast enough for a freewheeling diode, but I do not think it has big enough chip to clamp the freewheeling current.
Good suggestion,
Thank you very much.
 

Bordodynov

Joined May 20, 2015
3,025
I advised you to use a Schottky diode. Even fast diodes of conventional design do not work well as limiters. The parasitic inductance also influences. It is necessary to minimize the length of the wires. In my example, I added a parasitic inductance of 2.5 uH. Do you have a metal, wire resistor load? Why do you use such a powerful, high-voltage transistor? In the electronic circuit shown by you, this transistor does not work in a typical mode, without a tail (slow dissipation of carriers). At a higher current, the overvoltage will decrease.
 

Thread Starter

Saeed Yousofi-Darmian

Joined May 7, 2016
8
I advised you to use a Schottky diode. Even fast diodes of conventional design do not work well as limiters. The parasitic inductance also influences. It is necessary to minimize the length of the wires. In my example, I added a parasitic inductance of 2.5 uH. Do you have a metal, wire resistor load? Why do you use such a powerful, high-voltage transistor? In the electronic circuit shown by you, this transistor does not work in a typical mode, without a tail (slow dissipation of carriers). At a higher current, the overvoltage will decrease.
Dear Bordodynov,
The reason of using this IGBT is that I want to use it in a 1kW Full-Bridge inverter. So, at the first step I am trying to drive it. If it works, I can go to next step, designing a suitable PWM waveform and etc. I will consider your valuable suggestions and I hope it work well. I'll post results of my work soon.
Many thanks for your kindness.
 

Veenahr

Joined Aug 16, 2021
1
Hii , I am simulating the 15 level inverter in matlab simulik by IGBT switches , but I am facing a spikes in output waveforms please any solutions ?
 
Top