[SOLVED] Pic 16f876a Pure Sinewave inverter

Status
Not open for further replies.
Hello thuhtay i thank you so much for your pcb, what i dont understand is there are connector's on the pcb you did'nt explain , the connection to current-h and current-l,v-out,also mux-a,and mux-b,pwm,etc all these area are terminated by a connectors please explain,also originally it was compiled with pic16f876a now you use pic16f877a can you send to me the source file on pic16f877a or the hexfile.
Best regards.
Obialor
 

hmmmmm....thuhtay can you post the output waveform with capacitive load, so I can see whats the problem, I will test my circuit again within a few days and post my detailed results.
 
I converted thuhtay's pcb design file to pdf version in case someone needs to make pcb by toner
transfer method and doesn't have circuitmaker.
 

Attachments

  • tinv.pdf
    270.4 KB · Views: 941
Pure Sinewave inverter

Hello all, quite a good design for an inverter, I see you are using the 50Hz transformer as part of the o/p filter, and that the low voltage switching is applied directly to the low voltage winding of the transformer - I would be interested to hear if the transformer steel heats up noticeably (after 3 hrs say) with a light load on the output (say 10W) due to the high dv/dt applied to the LV winding. Looking at the circuit diagram it is hard to see how you have kept the flux swing of the transformer core more or less symmetric about zero? Are there calculated dead times to help reset the BH loop or are you relying on the resistance of the LV winding of the transformer? Is there a primary (LV) side current limit in case the flux staircases and sends the transformer into saturation? If too high a load is put on the o/p , or a short, does the inverter current limit or stop? Some appliances put non synchronous loads on the mains and may cause flux staircasing due to the nature of the load. Is there a soft start or ramp up (from zero) of the Vout at power up? I look forward to hearing more about the testing and results - very well done so far!!! Regards, Orson Cart.
 
Last edited:

One important thing about getting corect value of the output voltage, which should be 230VAC. First of all, Vac_1 and Vac_2 should be conected like this: you need osciloscop with two channels, you must observe two signals with it. The first one is A/B (50 Hz square) and the second on V_OUT (sinusoidal). When A/B is 5V then V_OUT is positive, when A/B is 0V then V_OUT is negative. It means both signals are in fase. In this case you will get less then 200 VAC between TR1 and TR2. So next step is remooving JP1 and JP2 and put there two resistors (just for test) 2MOhms or more. Run the invertor and you should have few volts more then before. So you have to find correct value of those resistors to get correct voltage on output of the inverter.

One more thing, I noticed overheating of the tantal capacitors C23 and C24. Anybody had situation like this?

Regards
 
Ok, additional capacitors (same values) connected parallel to C23 and C24 helped.

Regards
 
thuhtay said:
hi Fragrance
I posted the PCB design for the Pure Sine Wave Inverter. Good Luck for your experiment. Sorry for my delay to post this file.
with regards
thuhtay
Click to expand...

hi thuhtay, let me tell that in which software did you make this file. i tried to open in express pcb but it does not open there. kindly tell me the software name............Regards Ahmad
 

soulful.ahmad said:
hi thuhtay, let me tell that in which software did you make this file. i tried to open in express pcb but it does not open there. kindly tell me the software name............Regards Ahmad
Click to expand...
The file is related to "Eagle PCB". Try opening with Eagle.
Cheers
 

hi all
here is an attached file , this a microchip project developed in 1997, about a sinewave ups using ferrite core transformer, check that out and tell me , what do you think about?
 

Attachments

  • UPS-picref01.pdf
    991.2 KB · Views: 1,096
the problem is the software is very simple, and i doubt it can work,it is compiled with mplab c. has two versions: open loop( without regulation) and feedback (with regulation) and each one for 50 hz and 60 hz, this project could be converted to work with an iron core transformer(50 or 60 hz), i would like some expert analyze this microchip project and tell me what he thinks.
 

soulful.ahmad said:
hi thuhtay, let me tell that in which software did you make this file. i tried to open in express pcb but it does not open there. kindly tell me the software name............Regards Ahmad
Click to expand...

this is a circuit maker2000 file
 

Dear Taner , need to work on this PIC16 inverter design to gather confidence in development of 10kw unit with 300VDC supply voltage from Solar panels as my field is far away and IR losses are heavy.
Could you mail me the schematics and BOM /PCB/software writing ..
Any ideas for a Transformerless output stage unit for 230VAC/50Hz delivery ?
send mail on ID:: vimal@cosmotecindia.com .
thank you
 

Hi Guys I am interested in building this inverter for my own home use. This thread is very old now so I imagine there are some very experienced members on here, so I have a few simple questions.

To upgrade this unit to handle about 1000VA I would need to modify a few things right. Namely the output mosfets, heatsink, lower rds on etc and the PCB design to handle the extra current. I think the firmware should be OK, except for the PWMmax settings. The current sensing will also need to be upgraded. I already have a suitable transformer. Anyone like to comment on this please.

I am aiming to run an inductive load, namely a small fridge from it. Anybody tried to start a fridge compressor with this inverter please comment.

Graham
 

Fridges draw about 5 x more than rated current at start up (first 1-2 seconds) and need 3 x rated VA for 5 seconds to start properly (unless it is a very modern fridge with a variable compressor technology) - so this is a big consideration for an inverter designed to run such loads - most inverters don't measure up to fridge starting. Regards, Orson Cart.
 

For highly inductive ratings you will need to add a delay in the current sensing for about 5secs.
I had to declamp the current limit window and revert back after this delay .
The Mosfet ratings were more than three times the present .
After a whole lot of efforts , the ref. started with hiccups ....
It is advisable to choose a higher wattage inversion only for the starting kicks
 

Hi Orson, thanks for your comments...

I have modified my fridge to reduce this surge at start up by adding an extra run capacitor. This reduced the start up current enormously and enabled me to start it from an old modified square wave 600 watt inverter. But now I feel the need to replace it with a better designed sine wave inverter.

---------- Post added at 08:52 ---------- Previous post was at 08:49 ----------
Thanks for your comments vimalkhanna...
Did you implement this delay in the firmware, or by analogue techniques.
Graham
 
How can I make a pure sine wave inverter without using high frequency career wave and ferrite-core transformer. I want to chop the 12V DC (from accumulator) with sine modulated PWM (SPWM) and feed this to the primary side of a step-up transformer; the output of the transformer should be filtered with a low pass filter (a capacitor in parallel with the secondary of the transformer) ultimately rendering a pure sine wave! If my notion is correct then please guide me in this direction. Please tell me if there are problems. What values should I use in the look-up table to construct the SPWM - if I want to construct each half cycle of the sine wave with 16 nos of sine modulated pulse width?

Thanks p55xp
 

This type of inverter - called MPW - multiple pulse width - was common in the 80's using transistors and iron/Cu transfomers, before the advent of high power mosfets / igbt's. Often built for 48V telecom systems - using a full bridge of high current transistors, the xfmr o/p usually had harmonic traps for the 7th, 11th, the MPW scheme was designed to reduce the lower order harmonics leaving only the higher ones to be filtered, usually 3 or 5 pulses per half cycle - their width modulated to regulate the o/p volts to 230rms.
 

Status
Not open for further replies.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…