[SOLVED] Need help for simulation of 3 pulse three phase to single phase cycloconverter in Pspice

[Hamz28]

Need help

 

[BradtheRad]

This looks like you want 20 Hz sinewave to the load at the center of the schematic? Seeing the definition of cycloconverter, it creates this by chopping off pieces of 3-phase 60 Hz waveforms at the right moments for sending to the load. The final result is a single AC waveform to the load at the desired frequency. (It's not a perfect sinewave but it needs no DC conversion in the process.)

The control logic (I suppose) chooses from either of 3 phases at either positive or negative excursion. Tutorial articles describe algorithms how to combine the portions. The big hurdle must be sending gate signals to triacs or mosfets at the right moments. Is that where you're asking for help?

 

[Hamz28]

This looks like you want 20 Hz sinewave to the load at the center of the schematic? Seeing the definition of cycloconverter, it creates this by chopping off pieces of 3-phase 60 Hz waveforms at the right moments for sending to the load. The final result is a single AC waveform to the load at the desired frequency. (It's not a perfect sinewave but it needs no DC conversion in the process.)

The control logic (I suppose) chooses from either of 3 phases at either positive or negative excursion. Tutorial articles describe algorithms how to combine the portions. The big hurdle must be sending gate signals to triacs or mosfets at the right moments. Is that where you're asking for help?

I have already made the control stage and now I am making the power stage which is the cycloconverter and well when I presented the simulation I was told that it differs greatly from what I developed in Mathcad. I used the method of THE COSINE WAVE CROSSING PULSE TIMING METHOD. the question would be if the circuit diagram is right, in any case I do not know how to connect the reference voltage. or what else is missing in my circuit.
Thanks for your answer.

 

[BradtheRad]

For the first part of the cycle you want the load to go positive for 50mS. Obviously your load should draw from a series of positive-going waveforms. It should receive parts of (pulses, I suppose?) of those waveforms. (About 9 of them.)
You must locate mosfets/triacs which deliver positive waveforms. This requires examining all 12 of those components. Turn on those but shut Off any which deliver negative during the first half of your cycle.
The other end of the load should be permanently attached to 0V ground.

For the next 50mS you want the load to receive negative polarity. Give it parts of the next 9 negative-going waveforms by turning on mosfets/triacs which deliver negative.
Turn off each mosfet/triac during the time it delivers a positive waveform.

Again keep the other end of the load attached to 0V ground. Your schematic looks like it will have to be revised.

There's something you'll have to decide about. Is an inductor contained in your load? If you try to shut it off abruptly it continues to yank current briefly through components that are suppose to be off. Some kind of snubber network might be needed.

 

[BradtheRad]

As a matter of fact, in order to create a positive waveform lasting 1/40 sec...
You can select pieces from all 18 waveforms which occur in that time, positive and negative. (Likewise to create a negative waveform lasting 1/40 sec.)

The reason you can do this is because your 12 switching devices give you the ability to send any waveform occurring in any direction through the load at any given time.

This is like saying your load is between 3 full-H-bridges at the same time, each of which can be independently controlled. There are also many tendencies to shoot-through yet this must be avoided. The switching scheme becomes a fiendishly challenging effort.

I guess I was mistaken to say your load must have one end connected to 0V ground.

 

[FvM]

A SCR based cycloconverter has limitations regarding feasible switching sequences. SCR can be only turned off by current zero crossing respectectively switching over to another SCR of same polarity. Switching to opposite polarity isn't possible before other SCR commutates or phase short is caused. The waveform in post #3 is most likely not compatible with SCR. But it's not possible to decide without knowing load phase angle (relation of R and L component).

 

[BradtheRad]

Seeing if this simulation imitates a cyclo-converter....
During 1/20th of a second it optimally utilizes each waveform coming out of the 3 generators. As 18 pulses of current arrive (9 positive, 9 negative), it changes their direction as needed. The SPDT switches resemble the switching action through H-bridges.

By making the generator voltage 2V it is sent through 6 led's which reduces Ampere pulse width. Together with SPDT switches this simplifies the overall picture (even though it's done with the help of several 'cheats' playing a part). The rightmost scope trace is simplified for a resistive load. You may choose to shape it into a sine by filtering it or apply Spwm.

--- Updated Dec 1, 2024 ---

Link below runs my schematic in Falstad's animated interactive simulator:

tinyurl.com/2aly2tsh