On-Grid Solar inverter - power convertion and MPPT issue

[Prince Vegeta]

I want to design a solar inverter that outputs 220v AC voltage @ 50Hz freq. without batteries.

Does anyone have any idea how that's done?

Our design was like 40v input boost that outputs 220v DC voltage then 50Hz inverter which makes it 220v/50Hz signal.

The inverter doesn't need control, just 50% duty PWM by Compare module in PIC18 was enough. Boost, however, needs to be config and control to maintain 220v output...

Now my question, how can I implement MPPT in this design? using P&O algorithm in MPPT will make Boost's output voltage varies and by this the inverter will never output 220v AC... How can I generate 220v AC in this situation?

for stuff like push-pull, flyback... I don't know about them so I ask u.

I need help to decide which thing to do to have 220v/50Hz AC with MPPT from a solar panel. ( I need it to be 220v cuz the project should run an airconditioner that works at 220v/50Hz AC)

 

[chuckey]

I think you need batteries or mega caps to start an air con unit. Look at its motor label, you will need 5 times that current for 2-3 seconds to start the compressor. Also with out battery you will NOT be using a lot of available solar energy.
Frank

 

[Prince Vegeta]

I think you need batteries or mega caps to start an air con unit. Look at its motor label, you will need 5 times that current for 2-3 seconds to start the compressor. Also with out battery you will NOT be using a lot of available solar energy.
Frank

OK, regardless of air conditioner, if I want to make a small design just for prove of consept, can you answer my questions?

I want to get some ideas, ideas about MPPT and stuff like these... I explained all.

thanks

 

[chuckey]

The proof of concept you have to prove is what a mains powered piece of kits going to do when the mains is varying all over the place. The proof of the battery support is demonstrated by 10s of thousands of such installations working world wide.
Frank

 

[Prince Vegeta]

I didn't quite understand what you meant.

I asked about ways to do such design... how to design the power circuitry? how to implement MPPT especially in code? << stuff like these...

I posted what me and my group did so that you can help us find what's wrong and what to replace...etc...

so, can you help?

 

[picgak]

If you want 220 VAc you will require 300VDc

 

[mtwieg]

So basically you want to make a grid tie inverter with MPPT? There's nothing preventing MPPT from working in a grid tie inverter. But since your output power will be AC, that means that there will be ripple somewhere in the system, either in the DC bus voltage or in the PV current. Ideally there is some component(s) in the system which can store a good deal of energy (batteries or capacitors) in order to filter off that ripple, but there's a limit to how much is feasible in order to keep things cheap and lightweight. So in practice you are going to end up having to relax your control loop somewhere (probably the DC bus).

The DC-DC converter has the task of implementing MPPT, so it will just be spitting out as much power as possible regardless of its output voltage (except for overvoltage conditions). The inverter has the task of delivering the proper amount of power to the load in order to keep that bus voltage within some limited range. So in fact the inverter regulates the DC bus voltage, not the DC-DC converter.

 

[Prince Vegeta]

So basically you want to make a grid tie inverter with MPPT? There's nothing preventing MPPT from working in a grid tie inverter. But since your output power will be AC, that means that there will be ripple somewhere in the system, either in the DC bus voltage or in the PV current. Ideally there is some component(s) in the system which can store a good deal of energy (batteries or capacitors) in order to filter off that ripple, but there's a limit to how much is feasible in order to keep things cheap and lightweight. So in practice you are going to end up having to relax your control loop somewhere (probably the DC bus).

The DC-DC converter has the task of implementing MPPT, so it will just be spitting out as much power as possible regardless of its output voltage (except for overvoltage conditions). The inverter has the task of delivering the proper amount of power to the load in order to keep that bus voltage within some limited range. So in fact the inverter regulates the DC bus voltage, not the DC-DC converter.

OK I get what you said. I saw a similar design that have a boost converter with MPPT that have Vin=36v and outputs 400v DC voltage to DC bus (why naming it like that?) then the inverter makes it 220vAC/50Hz...

so how can I design such circuit? I mean, boost converter and inverter... especially the inverter because I don't know how it can reduce (regulate) the output voltage of the boost from 400vmax to 220v?

and... the boost's voltage will never be 400v always right? i.e it will be less and more... ?

I hope you help me more in this because I have a very limited timeline... currently we are designing an ATS (automatic transfer switch) to implement it with the solar inverter.

looking forward to your post

thanks

 

[mtwieg]

OK I get what you said. I saw a similar design that have a boost converter with MPPT that have Vin=36v and outputs 400v DC voltage to DC bus (why naming it like that?) then the inverter makes it 220vAC/50Hz...

so how can I design such circuit? I mean, boost converter and inverter... especially the inverter because I don't know how it can reduce (regulate) the output voltage of the boost from 400vmax to 220v?

The inverter part uses SPWM modulation, which allows the amplitude of the AC output to be varied (so long as the DC bus voltage is high enough). For grid tie inverters, I believe current sensing on the inverter output is sensed and regulated.

and... the boost's voltage will never be 400v always right? i.e it will be less and more... ?

Yes, the DC bus will vary during line load transients, and it will have ripple at twice the line frequency. But overall it should be regulated around some nominal DC voltage.

From a circuit perspective, there's nothing really special about the boost converter or inverter. The tricky part is the system control.

 

[Prince Vegeta]

SPWM? I made a program that outputs 50Hz/50% duty cycle PWM.. PIC18 can't output this PWM signal using PWM functionality while having 16MHz clock cycle... so I used Compare function to achieve it. What do you suggest here? what's the difference between PWM and SPWM? I thought that PWM only activates/deactivates stuff but control the amplitude???

OK, now current sensing is done by any current sensor, right? but how can it be regulated?

DC bus voltage will be varying as you said and that's what MPPT is about right? varying voltage so that current can be varied and power is calculated... is that what you meant?

ripple stuff I don't quite understand how to handle them... putting capacitors maybe?

_______


for circuit, can you make me a proteus (or anything) file that demonstrates it? I know I asked a lot but I want to learn a lot of stuff in a very short time.. ^_^

any recommendations for parts & components to buy?

 

[mtwieg]

SPWM? I made a program that outputs 50Hz/50% duty cycle PWM.. PIC18 can't output this PWM signal using PWM functionality while having 16MHz clock cycle... so I used Compare function to achieve it. What do you suggest here? what's the difference between PWM and SPWM? I thought that PWM only activates/deactivates stuff but control the amplitude???

SPWM means sine wave PWM. You use a PWM frequency much higher than the sine wave frequency you want to produce (for example 20KHz PWM frequency to make a 50/60Hz sine wave), and you modulate the duty cycle sinusoidally. You apply a lowpass filter on the output of the inverter and end up with a low distortion sine wave, as opposed to a poor square wave you get from using a simple 50% fixed duty cycle PWM.

OK, now current sensing is done by any current sensor, right? but how can it be regulated?

When using SPWM, you can modulate the amplitude of the output voltage simply by scaling your sine wave duty cycle.

DC bus voltage will be varying as you said and that's what MPPT is about right? varying voltage so that current can be varied and power is calculated... is that what you meant?

MPPT is about optimizing the operating conditions of the PV source for max power output. Yes, that means the PV voltage will change, but I was referring to the output of the DC-DC converter. The HVDC bus will fluctuate because the instantaneous input and output powers will not always be exactly equal.

I know I asked a lot but I want to learn a lot of stuff in a very short time.. ^_^

any recommendations for parts & components to buy?

You had best start by gathering and reading lots of literature on power conversion and control before actually thinking about real hardware implementations. Unfortunately, detailed info on grid tie inverters is hard to come by, but reading up on similar topics such as PFC and VFD will be helpful.

 

[Prince Vegeta]

SPWM means sine wave PWM. You use a PWM frequency much higher than the sine wave frequency you want to produce (for example 20KHz PWM frequency to make a 50/60Hz sine wave), and you modulate the duty cycle sinusoidally. You apply a lowpass filter on the output of the inverter and end up with a low distortion sine wave, as opposed to a poor square wave you get from using a simple 50% fixed duty cycle PWM.

now this SPWM can be learned by reading the data sheet, right? OK, let's say that the CCP pin is RE0 for example. This RE0 will produce a 50Hz sine wave? then how can I connect it to the full-bridge inverter thyristors (via IGBT/MOSFETs..)? I know that inverter thyristors should take gate pulses and then the DC-to-AC conversion starts... but having a sine wave...?

When using SPWM, you can modulate the amplitude of the output voltage simply by scaling your sine wave duty cycle.

Still I don't understand how to change the amplitude of the voltage. Let's say that the boost is designed to output 400v DC voltage (DC bus they call it), how can the inverter change it to 220 or 230v AC @ 50Hz? the 50Hz is OK, but the amplitude...?

you say scaling up my duty cycle... How is that done here at this SPWM?

MPPT is about optimizing the operating conditions of the PV source for max power output. Yes, that means the PV voltage will change, but I was referring to the output of the DC-DC converter. The HVDC bus will fluctuate because the instantaneous input and output powers will not always be exactly equal.

I was referring to DC-to-DC Boost converter's voltage too. it's output is the DC bus voltage right? so the nominal voltage of the panel is about 40v... the boost design will have a nominal output of 400v... this 400v is increased/decreased depending on current readings and power calculation. <<<< that's what I understood so far.

and if the voltage decreased to 300v rather than 400v, how does that affect the amplitude of AC sine wave?

what kind of control needed?

 

[mtwieg]

now this SPWM can be learned by reading the data sheet, right? OK, let's say that the CCP pin is RE0 for example. This RE0 will produce a 50Hz sine wave? then how can I connect it to the full-bridge inverter thyristors (via IGBT/MOSFETs..)? I know that inverter thyristors should take gate pulses and then the DC-to-AC conversion starts... but having a sine wave...?

You can't use thyristors for the inverter, you need devices that can be turned off, like IGBTs or MOSFETs.

Here's a picture of what SPWM looks like on a scope:


Top trace is the output of the inverter. Its duty cycle is varied continuously according to a sinusoid function. After passing through a filter, you get a clean since wave (bottom trace).

Still I don't understand how to change the amplitude of the voltage. Let's say that the boost is designed to output 400v DC voltage (DC bus they call it), how can the inverter change it to 220 or 230v AC @ 50Hz? the 50Hz is OK, but the amplitude...?

you say scaling up my duty cycle... How is that done here at this SPWM?

So look at the waveform above again. The duty cycle varies between 0 and 1, so the output voltage will vary to the full extents of the DC bus voltage. However, if you were to change the SPWM to only vary the duty cycle between 0.25 and 0.75, then the resulting since wave would only have half its original amplitude. So by varying the amplitude of the duty cycle modulation, you vary the amplitude of the sine wave.

I was referring to DC-to-DC Boost converter's voltage too. it's output is the DC bus voltage right? so the nominal voltage of the panel is about 40v... the boost design will have a nominal output of 400v... this 400v is increased/decreased depending on current readings and power calculation. <<<< that's what I understood so far.

It's a bit more complicated than that.... the control loop will do its best to keep the DC bus voltage close to some specified level (for example 400V). But it will be allowed to vary when required to (when the input and output power levels can't be exactly the same).

and if the voltage decreased to 300v rather than 400v, how does that affect the amplitude of AC sine wave?

Then it will limit the max amplitude achievable by the inverter.

At this point you really need to go out and do some background reading on inverters and control. Here are a few links to get started:
The best way to get acquainted with SPWM is to learn about VFDs: https://en.wikipedia.org/wiki/Variable-frequency_drive
There a lot of good literature available on active PFC. Your project won't involve this directly, but understanding active PFC systems makes understanding grid tie inverters much easier (it's mostly the same thing except in reverse): https://www.fairchildsemi.com/an/AN/AN-42047.pdf
And here are a couple links with some info on grid tie inverters:
**broken link removed**
https://solar.smps.us/grid-tie-inverter-schematic.html

 

[Prince Vegeta]

Thanks, now I understand SPWM. But you didn't answer my question about what is the signal to be fed into inverter switching devices (MOSFETs for ex.)?

^
the 1st signal u put is a PWM signal with varied duty cycle, is this the signal that should be fed into the H-bridge inverter? I know that H-bridge takes 2 control signal one for each 2 MOSFETs. So, what is the control signals from microcontroller that should feed the H-bridge inverter MOSFETs?

___

Is there any code example of SPWM?

____

It's a bit more complicated than that.... the control loop will do its best to keep the DC bus voltage close to some specified level (for example 400V). But it will be allowed to vary when required to (when the input and output power levels can't be exactly the same).

Hmmm.... keep the DC Bus close to the nominal 400v? but MPPT calculations can lower or rise the voltage to achieve the required max point.

now:

(when the input and output power levels can't be exactly the same)

this part I didn't quite understand it.

 

[mtwieg]

Thanks, now I understand SPWM. But you didn't answer my question about what is the signal to be fed into inverter switching devices (MOSFETs for ex.)?

^
the 1st signal u put is a PWM signal with varied duty cycle, is this the signal that should be fed into the H-bridge inverter? I know that H-bridge takes 2 control signal one for each 2 MOSFETs. So, what is the control signals from microcontroller that should feed the H-bridge inverter MOSFETs?

Right, you use complementary PWM signals to drive each opposing pair of FETs. Often dead time is inserterted into these complementary PWM signals to prevent shoot through on the inverter. Many modern MCUs (including many PIC families) can implement all this in software.

Is there any code example of SPWM?

I've seen many topics on these forums, some of which have full code and schematics given. Just search around.

Hmmm.... keep the DC Bus close to the nominal 400v? but MPPT calculations can lower or rise the voltage to achieve the required max point.

The MPPT controls the voltage on the PV source. The DC bus I'm talking about is the the output of the DC-DC converter, which can be independent of the source voltage.

 

[Prince Vegeta]

The MPPT controls the voltage on the PV source. The DC bus I'm talking about is the the output of the DC-DC converter, which can be independent of the source voltage.

aha... but how can u control the PV voltage? and if u control the PV voltage for the sake of MPPT, what is the purpose of controlling DC Bus' voltage?

Right, you use complementary PWM signals to drive each opposing pair of FETs. Often dead time is inserterted into these complementary PWM signals to prevent shoot through on the inverter. Many modern MCUs (including many PIC families) can implement all this in software.

Hmm... the complementary PWM signal is the 1st one right? so we have 2 pairs, one is on while the other is off and the other half cycle the opposite... is that correct?

the CCP pin will output that PWM signal in your scope pic (1st signal) so how to feed this into the 2 pairs? I have a suggestion please see it here:

if (CCP pin == 1 ) {

pair1_pin = 1;
pair2_pin = 0;
}

if (CCP pin == 0) {

pair1_pin = 0;
pair2_pin = 1;
}

^

this way pair 1 will have say the exact PWM signal and pair 2 will have the opposite one... this is close to what I made using Compare module to output 50Hz signal then feeding it like this. What do you think of this?

Often dead time is inserterted into these complementary PWM signals to prevent shoot through on the inverter

???

 

[mtwieg]

aha... but how can u control the PV voltage?

By varying the current drawn from the PV. You should read up on available literature on MPPT if this is unclear.

and if u control the PV voltage for the sake of MPPT, what is the purpose of controlling DC Bus' voltage?

The HVDC bus voltage must be large enough so that it can generate the required AC output voltage (so to generate 220VAC you need at least ~312VDC), but not so high that it destroys the inverter (depends on the FETs/IGBTs used).

Hmm... the complementary PWM signal is the 1st one right? so we have 2 pairs, one is on while the other is off and the other half cycle the opposite... is that correct?

Yes, that's basically correct.

the CCP pin will output that PWM signal in your scope pic (1st signal) so how to feed this into the 2 pairs?

Depends on the specifics of your inverter, especially the gate drivers you use. Again, you should read basic literature on the subject.

if (CCP pin == 1 ) {

pair1_pin = 1;
pair2_pin = 0;
}

if (CCP pin == 0) {

pair1_pin = 0;
pair2_pin = 1;
}

^

this way pair 1 will have say the exact PWM signal and pair 2 will have the opposite one... this is close to what I made using Compare module to output 50Hz signal then feeding it like this. What do you think of this?

Well, that's one way of doing complementary PWM. But most good MCUs will have PWM modules built in that minimize the amount of code necessary (for example, you can use two independent PWM modules in a synchronized way to make the complementary signals, with dead time.

???

Dead time.

Look, you're new to that stuff, and that's fine. But a GTI is not a good beginners project. You should start with a couple simpler projects, like a simple inverter or DC-DC converter before you tackle a GTI.

 

[FvM]

A SPWM inverter can use "complementary" PWM, also called bipolar modulation, but most industry standard inverters are rather using unipolar modulation, which requires to generate two different PWM signals plus deadtime generation.

I agree with mtwieg, that grid-tied inverters are considerably beyond the reach of a beginners project. Besides sine pwm generation, it also involves grid synchronization and control of sine currrent waveforms with unity power factor and protection features.

 

[Prince Vegeta]

so what is this unipolar method? Is it like what I suggested with my sample code or what?

You two are right, these kinda things are tough to me right now.

our project is about making a grid-tie solar inverter with an ATS (automatic transfer switch). Currently I'm working on designing this ATS and program a PIC16F877A MCU. It requires ADC and stuff like that... Do you have any idea about these ATS devices?

I'll be thankful to you if you give me a "roadmap" to learn all these stuff about inverters and so... i.e. from where should I start and what to learn next...

Thanks

 

[FvM]

A brief discussion of PWM modulation methods can be found in this thread. Please notice the literature links given by bigdogguru. Bipolar modulation is alos referred as 2-level PWM and unipolar as 3-level PWM.
https://www.edaboard.com/threads/274928/

- - - Updated - - -

ATS is a topic for emergency power supply rather than grid-tied inverters. I guess a confusion of terms regarding "on-grid inverter".