Voltage single phase active front end rectifier - simulation problem

[FreeRiderCz]

well i get rid of some errors (obviously if i use hirearchical blocks i cant use PI regulator as name with "space" between PI and regulator) but now is there only problem in hirearchical block of PWM - nodes are floating....

here is curent version of my schematic
Moderator action: removed link to external file server

- - - Updated - - -

well i get rid of some errors (obviously if i use hirearchical blocks i cant use PI regulator as name with "space" between PI and regulator) but now is there only problem in hirearchical block of PWM - nodes are floating....

here is curent version of my schematic
singlephase voltage epsilon regulation.rar

[FvM]

As previously mentioned, I don't have Orcad PSpice 16.

Floating nodes, as indicated by the log file suggest basic errors in simulation circuit, not specific to AFE application.
I need to decode the SPICE netlist, getting schematic pdf would be helpful.

[FvM]

I reviewed the netlist and found that the error is rather trivial.

You have e.g. E_PWM_E3 driving E_PWM_E6, both E3 output nodes are floating. Solution is to ground one output node, or connect at least a high ohmic resistor to ground if you are not sure about the actual circuit operation.

[FreeRiderCz]

I reviewed the netlist and found that the error is rather trivial.

You have e.g. E_PWM_E3 driving E_PWM_E6, both E3 output nodes are floating. Solution is to ground one output node, or connect at least a high ohmic resistor to ground if you are not sure about the actual circuit operation.

I did what you suggest - make node for output on E3 and E3 (they was similar and got same problem) and it helps i can "run" simulation now, now i just cant get results duo to convergence problems

- - - Updated - - -

I reviewed the netlist and found that the error is rather trivial.

You have e.g. E_PWM_E3 driving E_PWM_E6, both E3 output nodes are floating. Solution is to ground one output node, or connect at least a high ohmic resistor to ground if you are not sure about the actual circuit operation.

I did what you suggest - make node for output on E3 and E2 (they was similar and got same problem) and it helps i can "run" simulation now, now i just cant get results duo to convergence problems

[CataM]

I managed to simulate it up to a point, where convergence problems appear again. However, I believe that using IGBTs is wrong because the control signals (PWM1+ and PWM1-) turns ON the IGBT which reverse biases the diode VR1, which means the current can not flow through it. Try maybe with only diodes first and IGBTs in the low side only.

Here is the simulation setup that allows you to simulate it somehow..


Likewise, the thetha=6.28 is wrong. You should try to implement it as thetha=6.28*50*TIME in an ABM/1 block (i.e. ABM with 1 input). The input to the ABM block is epsilon like follows:

ABM/1 block:
(6.28*50*TIME-V(%IN)) where V(%IN)=Epsilon. Then, take the output of this block to the "SINE" input.

Attached are also the pdf schematic files that represent post #21 + #24.

[FreeRiderCz]

I managed to simulate it up to a point, where convergence problems appear again. However, I believe that using IGBTs is wrong because the control signals (PWM1+ and PWM1-) turns ON the IGBT which reverse biases the diode VR1, which means the current can not flow through it. Try maybe with only diodes first and IGBTs in the low side only.

Here is the simulation setup that allows you to simulate it somehow..


Likewise, the thetha=6.28 is wrong. You should try to implement it as thetha=6.28*50*TIME in an ABM/1 block (i.e. ABM with 1 input). The input to the ABM block is epsilon like follows:

ABM/1 block:
(6.28*50*TIME-V(%IN)) where V(%IN)=Epsilon. Then, take the output of this block to the "SINE" input.

Attached are also the pdf schematic files that represent post #21 + #24.

Hello, thanks for the reply,

i just want to ask about IGBT - you said i should try it without them only with diods, im not sure what means "Try maybe with only diodes first and IGBTs in the low side only."

Also i thought that theta =omega*t=2*3,14*50*0,002 thus its 6,28 (i take 2pif so 2*3,14*50 and 0,002 as time for frequency of 50 is 20ms this isnt correct?

[CataM]

i just want to ask about IGBT - you said i should try it without them only with diods, im not sure what means "Try maybe with only diodes first and IGBTs in the low side only."

Sorry for the mistake. IGBT only makes sense for V1 and V44. Why do your reference suggest for V3 and V2 as well ?

[FreeRiderCz]

Sorry for the mistake. IGBT only makes sense for V1 and V44. Why do your reference suggest for V3 and V2 as well ?

If i understand you right well i thought that V1 and V2 should be ON and OFF at same time, and in opposite V44 and V3 should be ON and OFF at same time but inverted in compare of V1-V2, as the proper function of AFE rectifier. I read in theory i used that IGBT is best for AFE rectifiers thx to their parameters, i just pick some part to try it

[FvM]

However, I believe that using IGBTs is wrong because the control signals (PWM1+ and PWM1-) turns ON the IGBT which reverse biases the diode VR1, which means the current can not flow through it. Try maybe with only diodes first and IGBTs in the low side only.

In usual terminology, "active front end" refers to a bidirectional AC/DC converter which is also able to recuperate power to the grid and sink/source reactive power to act as active power factor correction device. To do so, it has to be implemented as four quadrant converter, e.g. 4 switch H bridge in case of the single phase circuit.

- - - Updated - - -

Likewise, the thetha=6.28 is wrong. You should try to implement it as thetha=6.28*50*TIME in an ABM/1 block (i.e. ABM with 1 input). The input to the ABM block is epsilon like follows:

ABM/1 block:
(6.28*50*TIME-V(%IN)) where V(%IN)=Epsilon. Then, take the output of this block to the "SINE" input.

Attached are also the pdf schematic files that represent post #21 + #24.

Same point was also addressed in post #12 and #14. There may be additional convergence problem (not unusual with power electronics circuits), but quite surely the circuit can't work as intended without supplying a time variant phase. I'm not sure over how many periods will ωt argument to sine function give sufficient accurate results, typically a sawtooth source will be used in the simulation circuit.