Sinewave inverter prototype...

[DeepOne]

Re: Sinewave inverter prototype.

two opto-coupler is simple switch on/off charger, it is possible not to use it, use one of them or take any external charger.
DIP28 version of AtMega8 is not equal to TQFP32 because it has no adc channels 6&7.

 

[naseerak]

Re: Sinewave inverter prototype.

Can this programm be converted to DIP version of M8 and which compiler are you using

 

[DeepOne]

Re: Sinewave inverter prototype.

Yes, it can - it is necessary to free two ADC channels for this. I use CodeVisionAVR 2.05 .
_______________________________________________
testing more powerful version (SPWM generator based on Tiny43) -

 

[naseerak]

Re: Sinewave inverter prototype.

Can you convert the code as I work in BASCOM.

 

[DeepOne]

Re: Sinewave inverter prototype.

But testing is your deal, ok?

 

[naseerak]

Re: Sinewave inverter prototype.

Hi DeepOne!
Thanks for the effort and time, I have a few suggestion here:
1:- The current sensing should be done on the High side to reduce the no of resistors or we better use a C.T.
2:- As optocoupler are Non-Linear and mains voltage sensing would be in-accurate.
3:- An LCD in place of 7-Segment would require less processor sources.
4:- As the transformer is 8-8---> 230v It is difficult for use as a charger
5:- I may be wrong but the FET driving stage is a bit complex we should work on reducing the part counts to make it more economical and flexible.
6:- Have you tested this and previous designs in Push-pull mode.

Now I have started learning CAVR to further enhance the code for charging and other modifications according to my needs and availability of Parts.

Thanks and Regards

 

[DeepOne]

Re: Sinewave inverter prototype.

Hi, Naseerak.
2. mains voltage sensing is between absent or presence, so optocoupler give sufficient information.
3. 7-Segment is more reliable and cheap.
4. it may be used with simple step up converter.
6. this and previous schemes is not designed for push pull mode.

I think for non-professional using CVAVR is very good choice since it is suitable and reliable.

_________________________________
test version of DC_DC 12->360V converter

 

[naseerak]

Re: Sinewave inverter prototype.

Thanks but I have a few question in my mind.
1:- Where is the Mains/Inverter relay control pin on the micro-controller
2:- Can an IR2110 be used instead of IR2109 as this is readily available here.
3:- By turning off the upper mosfets the lower mosfets can be used for charging by utilizing boost PFC charging.

 

[DeepOne]

Re: Sinewave inverter prototype.

1. its removed, but relay is switched itself.
2. IR2110 may be used instead of IR2304 (may be with external RCD delay chains), not 2109.
3. For me something based on MC34063 looks as more simple decision.

___________________________________________
first test of 1kW DC_DC 12_360V at power 500W

 

[naseerak]

Re: Sinewave inverter prototype.

The delay could be done in software

- - - Updated - - -

Today i tried a simple atmega8 circuit with a pot connected 5v ground and the center pin to the ADC but no pulse width change by varying the voltage on ADC0.

 

[DeepOne]

Re: Sinewave inverter prototype.

but it work in Proteus
_________________________________________
version with tiny43 and conversion from 350VDC. Wave is little distorted, supposedly because filtering is unbeseeming.

 

[naseerak]

Re: Sinewave inverter prototype.

Thanks for your prompt reply. I simulated your design, as I was looking for ADC0 which is used for feedback but no change in wave form as I slide down and up the slider.
BTW how many samples are you using and what timer values are you using.
The timer value you given in of your previous post was not working in win7

 

[DeepOne]

Re: Sinewave inverter prototype.

but no change in wave form as I slide down and up the slider.

If i simulate this design - it work, if you - not work. Seems nothing will not to do with this.

how many samples are you using and what timer values are you using.

samples amount is thirteen as you may see in source code ("#define n_stable 13"), and about what timer you talk?

________________________________________________________________
next test of tiny43 last scheme

 

[Golden Electronics]

Re: Sinewave inverter prototype.

In inverter mode due to load current battery discharge so output may not remain constant. So we require change pwm duty cycle. In ur project how u do ?

 

[DeepOne]

Re: Sinewave inverter prototype.

for 350VDC->sine wave inverter pwm duty cycle changed in DCDC 12->350V part, for inverters with 50Hz transformer is used several PWM tables in mcu. Also possible correct tables in the fly, but in my opinion output voltage stabilization is not very important function.

 

[naseerak]

Re: Sinewave inverter prototype.

By samples I mean The timer values entries in the table.

- - - Updated - - -

I am not talking about the no of tables, I am talking about the elements in the table which is more than 32.

- - - Updated - - -

Ok Got it now, BTW can present the M8 version of 350v to 220v version I mean only AC section driven by M8 and chip drivers. I have found some IR2104 here in pakistan.

 

[DeepOne]

Re: Sinewave inverter prototype.

In such case amount of the elemens is three hundred fourteen for half-wave, spwm top value 250. But it is easy to see in the source code.
IR2109&2104 is not suitable for driving IRF740 (or 2SK3299, etc.) fets, because it's deadtime (around 0.6uS) is not sufficient.

 

[naseerak]

Re: Sinewave inverter prototype.

Thanks but I want to drive IRF3205 with IR2109 not IRF 740. another method would be to keep the last and start entries to 0 to generate the necessary dead time.

 

[DeepOne]

M8 version of 350v to 220v

VDS for IRF3205 is 55V, so it is impossible to use it at 350V

 

[naseerak]

I mean to use 3205 for 50 hz version