Variable frequency drive IGBT Sequence (paint picture)

[ghoetic]

Hello everyone!

Short introduction:

Im an Automation technician with a burning desire for electronics and power electronics and pretty much everything connected with electricity.
And i have very limited electronics knowledge, but i am slowly getting a hang for things as time allows, watching countless youtube videos, reading books and searching the internet! :thumbsup:

recently there is something that has been on my mind alot, i want to design a very basic and barebone variable frequency drive. like a proof of concept build.

I understand the very basic idea behind rectifying, storing and inverting. but i also like to think that some of my ideas related to concept are far ahead of some of my practical understanding ^^.

So lets get to it then! One fundamental point about the three phase system is this "There will always be a point in time when two of the phases are returning current through one" if i got that right.

And one fundamental thing about the igbt bridge is this "The two gates per phase lead (e.x 1 & 4) may never be open at the same time, this will short circuit the dc bus" much like if the igbt chopper would get damaged and stay closed.

So i started thinking and drawing symbols on paper much like the one will provide, but only just recently got a hang of "HOW I THINK" it should work or best resemble true three phase sequence. Now i have tried to search for hours something resembling this but i have had no luck. I might be dead wrong, so i will just upload this paint picture and you can tell me what you think. the boxes with number represents the pulse in time that the igbt should stay closed.

/Alexander

 

[FvM]

See https://en.wikipedia.org/wiki/Space_vector_modulation

A VFD needs to adjust the output voltage proportional to the motor speed to achieve a constant flux, it interleaves the six active vector output with zero vector intervals (PWM).

Instead of setting it up as a direct space vector modulation, you can also generate the VFD modulation scheme as three-phase sine pwm.

 

[chinuhark]

an984 microchip
an887 microchip
Search and download these two app notes. They should be sufficient to get you started.

 

[ghoetic]

Finally tested my own VFD (video & pictures)

Hello power electronics enthusiasts!
How are you today?

Like the title says this post is about my VFD project. And what a better place to start then to refrence my first post here at edaboard.
https://www.edaboard.com/threads/326338/#post1395524

lol :laugh:.

Well anyway i have come quite far since then, i have actually built and testet my own VFD though a very crude one!
The amount of time i spent researching this and the amount of hipfiring things.. im suprised it turned out this well!

And id like to add that the biggest programming ive done before is something like an led routine on an arduino with breadboard.
And in soldering i think i soldered a transistor with a potentiometer for a 24v dc fan with a veroboard.

I have taken big leaps with this project, and i might do a better write up later of my whole experience.

In a nutshell i soldered a big ass gate control card with a PIC18F4431 and three IRS2113 drivers.
And a seperate 6 discrete IGBT card. Everything on verobard.
Used a 60v 3A supply as my first test on an 230/400 50Hz 2.3A Induction motor, delta connected.
The potentiometer was not implemented for this run, because trying to calculate the frequency input and scale for volt/hertz and changing timer0 reload value slowed the whole thing down to mush with XC8.

Im currently taking my time to remake the program and will hopefully someday fix this issue, might be my lack of understanding proper C programming that made it all slow down.

Anyways I used 1 slide switch for "run or stop" command, and one slide switch for "forward 50hz or reverse 10hz" with precalculated values.

Here is the video: https://www.youtube.com/watch?v=djPJI0xilC4

And wow what an satisfaction when i did the first run at 10hz and heard a slight coil whine and i watched the current go up. man if i only could describe the feeling!

 

[FvM]

Impressive!

To address a possible problem nevertheless. With 60 V bus voltage, there's hopefully sufficient margin to get away without any bus capacitors near the IGBT bridge. When increasing the bus voltage to a regular 310 or 560V level, this might bring up serious issues.

 

[ghoetic]

Thank you, thats why im stacked with 4 5700uF 400v capacitors in the background, i was initially going to used rectified 230, but satisfied myself with this until i get the potentiometer working

 

[FvM]

The large electrolytic capacitors are fine, if you don't manage to bring them very near to the IGBTs, there should be X µF of low ESL+ low ESR film capacitors directly across the half-bridges.