I have a general question about the inverter used in electric vehicles.

In an electric vehicle (EV), an inverter is a critical component that converts direct current (DC) from the vehicle's battery into alternating current (AC) to power the electric motor.

I’ve read that a microcontroller controls IGBTs using PWM to turn the switches on and off.

However, I’m unclear on how this switching process translates into an AC waveform. Could you tell me in how the PWM modulation of the DC battery voltage through the IGBTs generates the AC output?

 

I have a general question about the inverter used in electric vehicles.

In an electric vehicle (EV), an inverter is a critical component that converts direct current (DC) from the vehicle's battery into alternating current (AC) to power the electric motor.

I’ve read that a microcontroller controls IGBTs using PWM to turn the switches on and off.

However, I’m unclear on how this switching process translates into an AC waveform. Could you tell me in how the PWM modulation of the DC battery voltage through the IGBTs generates the AC output?

It is fairly simple. A reference sine wave is used in conjunction with a sawtooth oscillator to create a PWM version of the sinewave amplitude. This PWM representation switches between +Vdc, GROUND and -Vdc and is a Fourier representation of the fundamental and many harmonics. A low pass filter is used to recover the original sine wave at a higher voltage. Like this:

 

To further expand on Papa’s explanation, the bridge on a motor inverter is a three-phase one, and uses 6 IGBTs.
Firing in the correct sequence generates a three phase sinewaves, each offset 120 degrees.

 

It is fairly simple. A reference sine wave is used in conjunction with a sawtooth oscillator to create a PWM version of the sinewave amplitude. This PWM representation switches between +Vdc, GROUND and -Vdc and is a Fourier representation of the fundamental and many harmonics. A low pass filter is used to recover the original sine wave at a higher voltage. Like this:
View attachment 330836

That's perfect as a explainer (that's how it's still taught and modeled) but nobody actually uses analog PWM generation today, its all digital but the filters in anything modern.
https://developerhelp.microchip.com...ng-analog-behavior/transition-to-digital-pwm/


Figure 1 shows that the PWM signal's duty cycle is proportional to the value of the VTH. This technique of generating a PWM output waveform is called "edge-aligned” because the generated waveform always becomes high at the beginning of the rising ramp, i.e. the beginning of the PWM period. Figure 2 shows a similar PWM signal using triangular waves as the oscillator signal. In the analog circuit implementation, the oscillator frequency is either fixed or selected through the selection of the R and C connected externally. Highly optimized circuits may require an iterative, time-consuming process to determine the best R and C values. Any change in the PWM frequency may require replacing the parts and possibly redesigning the entire system. In the digital PWM implementation, the frequency can be changed in wide ranges by changing the values of internal registers. This technique will be covered in the next few sections.

 

That's perfect as a explainer (that's how it's still taught and modeled) but nobody uses analog PWM generation today, its all digital but the filters in anything modern.

If the sawtooth represents the number in a counter and the comparator is a numerical comparator, then it is still correct.

 

If the sawtooth represents the number in a counter and the comparator is a numerical comparator, then it is still correct.

I did say the "edge aligned" model is correct even if we no longer have analog wave forms and components. The main advantage with digital with that you can easily have other PWM alignments that work better with sinusoidal modulation, space vector modulation or even LED drivers.

1. Center-aligned PWM:
   • Symmetric PWM
   • Center-aligned PWMs are most often used in AC motor control to maintain phase alignment
2. Left-aligned PWM
   • Asymmetric PWM
   • Left-aligned PWMs are used for most general-purpose PWM uses
3. Right-aligned PWM
   • Asymmetric PWM
   • Right-aligned PWMs are typically only used in special cases that require alignment opposite of left-aligned PWMs
4. Dual-edge PWM
   • Dual-edged PWMs are optimized for power conversion where phase alignment must be adjusted

https://developerhelp.microchip.com...nalog-behavior/pwm-edge-center-aligned-modes/

https://forum.allaboutcircuits.com/threads/pic32mk-mc-qei-example.150351/post-1532387

https://forum.allaboutcircuits.com/...r-for-3-phase-115v-supply.181424/post-1667719
SAPWM (Saddle PWM) using centered Symmetric PWM cycles for the 3-phase 400Hz waveforms.

SAPWM (Saddle PWM) waveform theory and application.
https://forum.allaboutcircuits.com/threads/pic32mk-mc-qei-example.150351/post-1550621

 

Certainly it CAN be done just as shown in those explanations. BUT in the EVs produced by the major suppliers the inverters are much more complex. And the bad news is that is about the extent of the explanation they will provide. "The top secret circuits are rather more complex." True indeed, but not helpful.

 

Perhaps interesting here:
https://engineering.wisc.edu/news/p...-motors-land-them-on-forbes-30-under-30-list/
https://interestingengineering.com/transportation/h3x-raises-20m-for-high-power-motors

I can't believe I'm providing a link to "Interesting Engineering", but there it is. It looks like there is a real story there.

 

Certainly it CAN be done just as shown in those explanations. BUT in the EVs produced by the major suppliers the inverters are much more complex. And the bad news is that is about the extent of the explanation they will provide. "The top secret circuits are rather more complex." True indeed, but not helpful.

BS, that's marketing. Electrical engineering is the same everywhere, there is nothing special or secret about their inverter circuits (double true if it's manufactured in China). Most of the complexity is to squeeze the last 1 percent in a design.

 

BS, that's marketing. Electrical engineering is the same everywhere, there is nothing special or secret about their inverter circuits (double true if it's manufactured in China). Most of the complexity is to squeeze the last 1 percent in a design.

And you think squeezing that last extra bit out, without spending more, is simple so that just anybody can do it "just like that" ??? Implying that all engineering is the same, and that it takes no skill or talent to create a great design???
And thinking that nothing is special enough to be a company secret??

 

And you think squeezing that last extra bit out, without spending more, is simple so that just anybody can do it "just like that" ??? Implying that all engineering is the same, and that it takes no skill or talent to create a great design???
And thinking that nothing is special enough to be a company secret??

'tis the iconoclast's lament.

 

And you think squeezing that last extra bit out, without spending more, is simple so that just anybody can do it "just like that" ??? Implying that all engineering is the same, and that it takes no skill or talent to create a great design???
And thinking that nothing is special enough to be a company secret??

Please, we're talking about EV motor drive systems, not fusion reactors. Show me the great design in the vast majority of Chinese EV cars. They sell because they are cheap, not because of great design or superior engineering.

The KFC recipe is a company secret, in that context it means little.

 

Here is a fun video that demonstrates an H bridge driving a transformer

 

'tis the iconoclast's lament.

I'm just saying there's nothing really special about the current electronics in a current EV. I see lots of claims but when you watch the teardown it's the same variation of the old AC drivers.

This is as it should be, conservative engineering for EV drive electronics is the correct call for mass production.

 

I can't imagine that the motor drive circuits in an EV are much different from any other variable speed drive. The clever bit it the drivability: controlling the motor speed from the throttle pedal, knowing when to engage regenerative braking, co-ordinating more than one motor etc.
The average driver can waste far more than the last 1% by constantly accelerating and braking. Regenerative braking is far more efficient than friction braking, but it's by no means 100%.

 

Ian just stated what makes thing different: " The clever bit it the drivability: controlling the motor speed from the throttle pedal, " The better drivers are controlling the VEHICLE SPEED by controlling the drive torque. It is all of those ROTTEN drivers who are constantly running up on the car ahead and then mashing down on the brakes, that cause the problems.
That is why the foolish concept of having a single-pedal speed control is so very defective.
Of course there certainly are a lot who can not possibly focus their attention on driving well enough. They should all take a bus and leave the driving to others.

 

Ian just stated what makes thing different: " The clever bit it the drivability: controlling the motor speed from the throttle pedal, " The better drivers are controlling the VEHICLE SPEED by controlling the drive torque. It is all of those ROTTEN drivers who are constantly running up on the car ahead and then mashing down on the brakes, that cause the problems.
That is why the foolish concept of having a single-pedal speed control is so very defective.
Of course there certainly are a lot who can not possibly focus their attention on driving well enough. They should all take a bus and leave the driving to others.

I completely agree with the difference but it's not at the DC to AC inverter level. It's at the resource management level where clever software decides and controls operator demands with system optimization of resources within the physical limitations of the EV.

 

The whole package is a system and given the conflicts between cost, functions, and production yield requirements, a great deal of "engineering wizardry", unique to the application, is required. And certainly the details of that creation will be considered secret so as not to be shared with competitors of any kind.

So while "the basics" may be common knowledge, there is a whole lot more involved than just those basics.

THAT is far from obvious to the M.B.A. mentality that believes that all engineers are the same, and so are totally replaceable and interchangeable. That was the stated belief of the new division manager at the company I retired from. Who would choose to be at a place run by such a fool???

 

That is why the foolish concept of having a single-pedal speed control is so very defective.

I don't follow how you inferred from your previous comment.
How is it "defective"?
A single-pedal speed control with regenerative braking can certainly recover a fair amount of the kinetic energy when slowing down.

 

The whole package is a system and given the conflicts between cost, functions, and production yield requirements, a great deal of "engineering wizardry", unique to the application, is required. And certainly the details of that creation will be considered secret so as not to be shared with competitors of any kind.

So while "the basics" may be common knowledge, there is a whole lot more involved than just those basics.

THAT is far from obvious to the M.B.A. mentality that believes that all engineers are the same, and so are totally replaceable and interchangeable. That was the stated belief of the new division manager at the company I retired from. Who would choose to be at a place run by such a fool???

What are you talking about?

Any competent EE power engineer or student can (should be able to) design these systems. .
https://web.mit.edu/first/kart/
https://web.mit.edu/evt/vehicles.html

It takes a lot more than pure engineering to make saleable products. The suits are a necessary evil...