Explanation of IGBT 'Rectifier' Operation and the function of Freewheeling Diode

mbro12

Joined Dec 21, 2016
5
Evening Gents,

First pots here, I work in rail industry and am currently doing my HNC in electrics/electronics so please excuse if I get tripped up by some basic principles but there's quite a lot of simple principles that can sometimes fall out of my brain.

I currently am trying to wrap my head around a HVAC to HVDC Converter circuit which utilises IGBT's for rectification purposes.
I am trying to understand the 'firing sequence' of these IGBT's and how the Freewheeling Diode functions and more importantly, why it does not cause a short circuit back to the HVAC supply

Attached is the basic circuit, no inductor is used on the supply line and as such most of the online articles I've read don't really explain the function.

If anyone can shed some light even on to the firing sequence and potentials on the 'Load' side during these firing phases, it would be much appreciated.

Thanks

Attachments

• 53.5 KB Views: 54

Dodgydave

Joined Jun 22, 2012
10,087
Looking at that circuit, i would say its an inverter oscillator pulsing a transformer, two transistors are fired at the same time,

if you look at the circuit, the transistors at top left and bottom right are fired together on one pulse, and the other two are fired together on the other pulse, producing a Seesaw Push Pull action, pulsing the Transformer across the supply. The diodes are to suppress the Back Emf.

NPN-1

Joined Mar 11, 2016
16
As Dodgydave said, the diagonal IGBTs are fired at the same time. Also the load (which must get only dc) is only shorted via the diodes when the ac supply is of the same polarity of the dc load. If you remove the IGBT from the circuit but keep the diodes (mentally I mean, in reality these diodes are part of the construction of the IGBT), you will still receive a pulsating dc voltage at the load.

The diodes as mentioned above are to suppress the back emf of the inductance in the circuit.

Hope it helps,
Student like you

shortbus

Joined Sep 30, 2009
8,956
May be wrong here but, aren't the diodes in the IGBT like the diodes present in a mosfet? Just an "artifact" of the manufacturing process? For the most part I've read and been told, not to 'count' on any specific value attributed to them. And definitely never count on them as a flywheel diode.

NPN-1

Joined Mar 11, 2016
16
... For the most part I've read and been told, not to 'count' on any specific value attributed to them. And definitely never count on them as a flywheel diode.
What is the problem with these inbuilt diodes used as flywheel/freewheeling diodes ?

BR-549

Joined Sep 22, 2013
4,938
A pwm full wave bridge?

mbro12

Joined Dec 21, 2016
5
Thoughts are the diodes operate to provide a full wave bridge rectifier and as mentioned above, diagonal igbts are fired to provide a short circuit and generate an emf in the transformer coil (or a coil not listed on the diagram). When these igbts are turned off the field collapses and provides current to the circuit, have yet to work out how this current will affect the output waveform.

Is this whatyou were getting at npn and dave?

BR-549

Joined Sep 22, 2013
4,938
Show us the whole circuit and show us the load. What power source are you starting out with?

mbro12

Joined Dec 21, 2016
5
Show us the whole circuit and show us the load. What power source are you starting out with?
Source is AC 50hz at around 1500v.
Load is variable as it has two potential loads. One is for an igbt inverter for induction motors and the other iside for igbt inverters for utility power such as power sockets lighting and compressors.

NPN-1

Joined Mar 11, 2016
16
Source is AC 50hz at around 1500v.
Load is variable as it has two potential loads. One is for an igbt inverter for induction motors and the other iside for igbt inverters for utility power such as power sockets lighting and compressors.
Is the DC output voltage fixed at a certain value?

mbro12

Joined Dec 21, 2016
5
Is the DC output voltage fixed at a certain value?
Yes

Fixed at 1300v line to earth. However the line to line voltage is 2600v. As such onevents of the powering lines must be negative dc

NPN-1

Joined Mar 11, 2016
16
Thoughts are the diodes operate to provide a full wave bridge rectifier and as mentioned above, diagonal igbts are fired to provide a short circuit and generate an emf in the transformer coil (or a coil not listed on the diagram). When these igbts are turned off the field collapses and provides current to the circuit, have yet to work out how this current will affect the output waveform.

Is this whatyou were getting at npn and dave?
Kind of, actually. 4 actual diodes placed in a similar manner would be a full wave bridge rectifier. In that case the voltage in the DC output is only controlled by the AC input. You cannot control when the diodes are on or off, the ac voltage does. the 4 IGBTs connected in the same way also will do full wave rectification, only in this case firing can be controlled and the DC output voltage can be changed. The IGBTs may have been there for a different reason altogether too.

The inbuilt diodes of the IGBT are there to allow the inductive current to safely pass through, I doubt they will have the full strength to take the rated current. Most of the current will pass through the IGBTs I suppose.

mbro12

Joined Dec 21, 2016
5
Kind of, actually. 4 actual diodes placed in a similar manner would be a full wave bridge rectifier. In that case the voltage in the DC output is only controlled by the AC input. You cannot control when the diodes are on or off, the ac voltage does. the 4 IGBTs connected in the same way also will dowave rectification, only in this case firing can be controlled and the DC output voltage can be changed. The IGBTs may have been there for a different reason altogether too.

The inbuilt diodes of the IGBT are there to allow the inductive current to safely pass through, I doubt they will have the full strength to take the rated current. Most of the current will pass through the IGBTs I suppose.

Transformer secondary winding is fixed but the source can be known to fluctuate by up to 20%. As such is it possible the igbt are used to compensate for this discrepancy in the sources consistency

BR-549

Joined Sep 22, 2013
4,938
Do you have a link to the equipment that you are using? I can not make sense of your descriptions.

Is this part of the circuit just provide a dc bus? Or is this the output to a load.

Whether those diodes are needed or not depends on what you are doing. In which I can not figure out.

NPN-1

Joined Mar 11, 2016
16
Transformer secondary winding is fixed but the source can be known to fluctuate by up to 20%. As such is it possible the igbt are used to compensate for this discrepancy in the sources consistency
The source variation may be the reason why the IGBTs are used. Also one more reason they can be used is to reduce harmonics, which are unwanted high frequency signals going back to the ac source and cause interference to other electronic equipment among other problems. Those high frequency signals are present due to the conversion of ac to dc. The IGBT when fired in a particular way can help bring these unwanted components to a minimum.

Do you have any idea about the control being used for the IGBTs?

shortbus

Joined Sep 30, 2009
8,956
What is the problem with these inbuilt diodes used as flywheel/freewheeling diodes ?
The inbuilt diodes of the IGBT are there to allow the inductive current to safely pass through, I doubt they will have the full strength to take the rated current.
You kind of answered your own question. The "inbuilt diodes" are called 'intrinsic diodes' they are a result of the layering of the different silicon layers when making the device. For the most part they are of an unknown value and to be ignored and not counted on to meet the standards of the real part of the device. For reliability a real (known value) diode in parallel should be used, then you know it is correct value.

NPN-1

Joined Mar 11, 2016
16
You kind of answered your own question. The "inbuilt diodes" are called 'intrinsic diodes' they are a result of the layering of the different silicon layers when making the device. For the most part they are of an unknown value and to be ignored and not counted on to meet the standards of the real part of the device. For reliability a real (known value) diode in parallel should be used, then you know it is correct value.
Of course for more critical applications you need a proper diode, but can these intrinsic diodes not suffice as freewheeling diodes to suppress the inductive kickback? In real life inverters/rectifiers, do they rely on these intrinsic diodes to do the freewheeling business, or do they use additional proper diodes for the same?

tsan

Joined Sep 6, 2014
116
May be wrong here but, aren't the diodes in the IGBT like the diodes present in a mosfet? Just an "artifact" of the manufacturing process? For the most part I've read and been told, not to 'count' on any specific value attributed to them. And definitely never count on them as a flywheel diode.
To my knowledge, there is no intrinsic diode in IGBT. Sometimes diode is added on purpose to the component package as often freewheeling diode is required anyway. On this case the diode is selected to be fit for use as freewheeling diode.

tsan

Joined Sep 6, 2014
116
The inbuilt diodes of the IGBT are there to allow the inductive current to safely pass through, I doubt they will have the full strength to take the rated current. Most of the current will pass through the IGBTs I suppose.
On the few IGBT module datasheets for motor inverter application I have looked, diode current rating was the same than IGBT current rating. This makes sense as IGBT current is commutated to diode. However, thermal resistance to module base was higher on the diode. This is optimatization for typical power factor of intended motor load. IGBT losses are higher than diode losses.