SPWM inverter

Thread Starter

Akki7213

Joined Apr 26, 2021
4
Is it possible to design a 1-phase or 3-phase where we have used SPWM technique to control output using opamp and other analog IC (without using microcontroller .)?
 

LowQCab

Joined Nov 6, 2012
323
Yes,
but its much easier, and cheaper, to learn how to design an adequate
Output-Filter for a standard single Supply.

For a 3-Phase setup, you are basically creating 3- complete, independent, SMPSs,
PLUS,
Circuitry to accurately keep them all running at the same Frequency,
and keep them perfectly out of phase with each other.
It's not worth the required cost and complexity.

6-Phase SMPSs are common on Computer Mother-Boards,
this allows much smaller Filter Elements, but requires lots of them.
.
.
.
 

Thread Starter

Akki7213

Joined Apr 26, 2021
4
Thank you for your response , it mean that without microcontroller it will be vary much difficult for 3 phase inverter designing using SPWM but 1-phase it can be done with proper filter designing?
I tried to design a spam using op-amp mostly, a square wave to triangular wave generator and comparing a sine were using comparator but my output of inverter was of a half were inverter. Were am I going wrong.? Please suggest6757C1CC-4414-41A2-B63C-EA885DC9568B.png564803C0-2262-4621-898F-75DFFA536FCD.png
 

LowQCab

Joined Nov 6, 2012
323
1) What problem are you trying to solve by using a 3-Phase Topology ?

2) What is the purpose of having a 3-Phase SMPS ?

3) Are you trying to control a 3-Phase Motor ?

4) Are you trying to reduce Power-Supply-Ripple-Voltage/Current ?

5) Are you trying to reduce the size of the Filter-Components ?









 

Ian0

Joined Aug 7, 2020
1,888
Thank you for your response , it mean that without microcontroller it will be vary much difficult for 3 phase inverter designing using SPWM but 1-phase it can be done with proper filter designing?
I tried to design a spam using op-amp mostly, a square wave to triangular wave generator and comparing a sine were using comparator but my output of inverter was of a half were inverter. Were am I going wrong.? Please suggestView attachment 237179View attachment 237178
Your main problem here is the supply for the signal processing - the sinewave goes below zero and you have no negative supply.
The design is no different to designing any other buck regulator, except that the reference is a sinewave not a steady voltage.
All the same designs for stability apply (see Texas Instruments design note SLUP340) - stability is easier to achieve in the continuous-time domain than digitally.
Probably the only thing you really need a MCU for would be generating the sinewaves in accurate phase and amplitude for the 3-phase version. (You can generate two sinewaves 120 degrees apart and get the third by adding the two and subtracting from zero)
It's sine-wave reference -- error amplifier -- compare with triangle wave -- drive output MOSFETs.
 

Thread Starter

Akki7213

Joined Apr 26, 2021
4
1) What problem are you trying to solve by using a 3-Phase Topology ?

2) What is the purpose of having a 3-Phase SMPS ?

3) Are you trying to control a 3-Phase Motor ?

4) Are you trying to reduce Power-Supply-Ripple-Voltage/Current ?

5) Are you trying to reduce the size of the Filter-Components ?


1. I am not working on 3 phase inverter for now… firstly i was asked in an internship interview to design an inverter whose supply is DC 400V and fundamental 50hz output has a rms voltage output is 90 per, and I don’t have to use any transformer or filter at end.
2. Single phase motor.
3. For now 1-phase motor.
4. Power supply ripple control was not asked so better to avoid it here .
5. Asked in interview is 90Per fundamental…so definitely a case of overmodulation and square wave inverter…i think so.
 

LowQCab

Joined Nov 6, 2012
323
My mistake. .........
I have too many Acronyms to keep track of in ~7 completely different fields.
I did not realize you were talking about a Sine-Wave Switch-Mode Inverter.
I assumed, incorrectly,
that there was simply a Language-Barrier thing going on,
and that you were talking about a 3-Phase Switch-Mode Voltage-Regulator.

Anyway .............

To design a Variable-Frequency / Variable-Voltage, AC Inverter,
much more information is required than simply "400-Volts", and 90%.

Where did the 400-Volts DC come from, and is it well Filtered and Regulated ???
Are any other AC or DC Voltage Sources available for use ???

90% of what ???
Efficiency ???
90% of 400-Volts Peak-Output-Voltage ???

How much Current is expected ???
Is Current-Limiting required ???

Is the AC-Voltage expected to be "fixed" at a standard Common Voltage, or adjustable ???

Is the expected Frequency "fixed" or adjustable ???
What is the expected Frequency-Range that must be accommodated ???

Is "Soft-Start", Ramp-Up / Ramp-Down Required ???

What is the allowable level of output Ripple ???

Here's a Block-Diagram ...........
.
.
.
AC Inverter FLAT .png
 

LowQCab

Joined Nov 6, 2012
323
As I was making this Schematic in my head, it evolved, several times.
Several basic changes were made, and I was to lazy to spend the time to change them.
Now that all the SiC-FETs are the same part, it makes much better sense to change it.
It would further reduce the total RDS-ON,
but it would also require a second High-Current FET-Driver,
and a slightly different part number, one with an "Enable-Pin".
The Schematic is just to show the concept,
it hasn't had all the details and all the Math worked-out.
Maybe I'll fix it if I have the time, and if there is some interest in it.
Finding problems in other peoples work is a good thinking exercise.
Thanks for pointing it out.
.
.
.
 

Ian0

Joined Aug 7, 2020
1,888
. . and I thought that it was something really smart that I was missing! I've not as yet had a go with SiC FETs.
TC4469 is a quad low-side driver with AND gate structure. I don't know of a dual.
 

LowQCab

Joined Nov 6, 2012
323
SiC FETs aren't to difficult to deal with,
but GaN-FETs have very special Gate-Drive requirements.

For High-Voltage, Sic-FETs are the easy compromise.

The "Dead-Time" needs to be carefully calculated.
.
.
.
 

Thread Starter

Akki7213

Joined Apr 26, 2021
4
As I was making this Schematic in my head, it evolved, several times.
Several basic changes were made, and I was to lazy to spend the time to change them.
Now that all the SiC-FETs are the same part, it makes much better sense to change it.
It would further reduce the total RDS-ON,
but it would also require a second High-Current FET-Driver,
and a slightly different part number, one with an "Enable-Pin".
The Schematic is just to show the concept,
it hasn't had all the details and all the Math worked-out.
Maybe I'll fix it if I have the time, and if there is some interest in it.
Finding problems in other peoples work is a good thinking exercise.
Thanks for pointing it out.
.
.
.
Where I can find all the maths and details ( all theory ) any reference will also be helpful.
 
Top