Has anybody done this ? (AC >DC >AC converter)

[Okada]

Has anybody done this ?

I am planning to make a AC to DC and DC to AC that is AC-DC-AC constant voltage stabilizer for refrigirator.

It will take 230V AC input and convert it to 325V DC using bridge and resorvoir capacitor then this will be used as VBUS voltage to DC-AC converter using 4 mosfets.

Of, course there will be parallel mosfets.

Output will be 230V constant AC 20A type.

The stabilizer can be used for multiple devices.

PIC18F will be used for the project.


Air Conditioner is 1000W

Refridgirator is 150W

Irms = P/V = 1000/230 = 4.35A

Ipk = 4.35 * 1.4142 = 6.15A

So, if I design a 20A output type AC-DC-AC stabilizer then it can easily power many devices.

 

[KlausST]

re: Has anybody done this ? (AC >DC >AC converter)

Hi,

This is how almost any frequency inverter works.
Some UPS work this way, too.

I don't recommend it to build it yourself.
Dangerous voltages, sefety regulations, hardware design, calculations....there are a lot of pifalls.

Don't hurt yourself or others, don't risk fire.

Klaus

 

[Okada]

re: Has anybody done this ? (AC >DC >AC converter)

No problem. I can handle it. First I need to make a 230V to 12V buck converter to get 12V DC for IR21xx devices used in the constant voltage stabilizer. From 12V I will use 7805 to get 5V for PIC but the problem is how to get 12V because initially PIC will be off and to get 12V from buck converter PIC should be running and it needs 5V for that.

This is the issue I am having.

Max ratings of voltage and current will be noted.

All calculations will be done and there will be proper cooling for the mosfet.

Initially 500W AC-DC-AC constant voltage stabilizer will be designed and then hgher wattage stabilizer will be designed after gaining experience and knowledge from lower wattage stabilizer.

Pure Sine Wave Stabilizer will be this.

SPWM generation and feedback have been tested and they are working fine.

Now the problem is to get 12V and 5V from 230V AC without using transformer.


Yes, KlausST.

I know about Inverters. I need to study about buck converters, cuk converters and boost converters.

My Rigol Oscilloscope is on its way.

I will first design 12V to 5V buck converter
12V to 24V boost converter
12V to -12V cuk converter

After gaining practical knowledge I will proceed further.


If I use transformer for power supply.

18V transformer

N = Vprim/Vsec = 230V/18V = 12.78

For Vpri = 180V

Vsec will be

Vpri/N = 180/12.78 = 14.08 (worst case)

This after bridge will be

14.08 * 1.4142 - 2.0 = 17.9V DC

This will go through 7812 and from 7812 to 7805 to get 12V for IR21xx devices and 5V for PIC.


For 6A output 1x Mosfet H-Bridge will be used.

So,

Mosfet is IPW60R041C6

Rds(on) = 41 mOhms

P = I^2 Rds(on) = 6 * 6 * 0.041 = 1.5W

So, with cooling this is good

for 20A output

3x Mosfet H-Bridge is used so that 20/3 = 6.67A divides into each mosfet.

- - - Updated - - -

Here is the circuit I designed. Now it is fine. It is a Pure Sine Wave Constant Voltage Stabilizer.

Regarding PCB track width.

I am using this.

https://www.4pcb.com/trace-width-calculator.html

2 oz / ft^2

Temperature rise is 20 degree C

Current max is 30A

It shows 28 mm that is 2.8 Cm track width for drain and source.

Is this track width enough for high current patch ? All components will be placed on PCB. FR4 grade Glass Epoxy PCB.




Stabilizer range 180V to 270V AC

180 * 1.4142 - 2.0 = 252.556V

270 * 1.4142 - 2.0 = 379.834V

I have to get constant 230V for these VBUS voltage range.


When the device is used for only refrigirator (150W) then power dissipation in mosfet pair is

Irms = P/V = 200/230 = 0.87A

Ipk = 0.87 * 1.4142 = 1.23A

Power in mosfet = 1.23 * 1.23 * 0.041 = 0.062W

Power in mosfet pair = 2 * 0.062 = 0.124W

So, no cooling is needed.

- - - Updated - - -

This much I have calculated for the power of stabilizer.

So, assuming refrigirator power will be max 150W I can easily build a stabilizer cum Inverter of 500W.

So, I can use the H-Bridge circuit as common and have two modes of VBUS voltage generation. One is from mains AC and another from 12V 200Ah battery.

When there is no mains then the stabilizer will work as constant voltage Inverter. In mains mode it will work as a constant voltage stabilizer.

I have one question just for saftey precaution.

If I accidentally short the terminals of the 12V 200Ah battery will the battery explode ?


This device can be used where power variations and power cuts are frequent.

Using this kind of stabilizer eliminates the use of a bulky multi-tapped transformer.

First I will design 500W stabilizer and if everything works fine then I will add more parallel mosfets. There will be option provided on PCB for 5 parallel mosfets with proper track widths.

One design can be used for different wattage stabilizer.

My plan is to build a 10 KW stabilizer for a home.

 

[KlausST]

re: Has anybody done this ? (AC >DC >AC converter)

Hi,

Stabilizer range 180V to 270V AC

180 * 1.4142 - 2.0 = 252.556V

270 * 1.4142 - 2.0 = 379.834V

I have to get constant 230V for these VBUS voltage range.

How can you get stable output voltage of 230V AC with only 180V (edited) AC input?
The calculated voltages are for no load conditins only.

*****
820 uF capacitor for DC rail:
Did you calculate it´s initial switch ON current?
Did you calculate it´s full load ripple current?
What about PFC? (Your circuit is not allowed to operate in many countries. Is it allowed to operate in your country?

****
You didn´t take care about power factor of your load devices.
You didn´t care of load peak current during start up.
You dind´t take care about MOSFET peak current during switching.
What about output low pass filter?

Klaus

 

[Okada]

re: Has anybody done this ? (AC >DC >AC converter)

Yes, these have to be implemented but one by one. First I need to get constant voltage from 180V to 270V AC input.

How can you get stable output voltage of 230V AC with only 10V AC input?
The calculated voltages are for no load conditins only.

*****
820 uF capacitor for DC rail:
Did you calculate it´s initial switch ON current?
Did you calculate it´s full load ripple current?
What about PFC? (Your circuit is not allowed to operate in many countries. Is it allowed to operate in your country?

****
You didn´t take care about power factor of your load devices.
You didn´t care of load peak current during start up.
You dind´t take care about MOSFET peak current during switching.
What about output low pass filter?

Where is 10V ? Is it AC or DC ? Where did I mention it ?

First I will build the stabilizer 500W. If it works fine then I will add active PFC, filters etc...

Just for testing purpose I think they are not needed. I just have to get get constant 230V sine voltage.

I am referring this.

I am referring this.

https://tahmidmc.blogspot.in/2013/02/demystifying-use-of-table-pointer-in.html

https://tahmidmc.blogspot.in/2012/11/feedback-in-sine-wave-inverter-pic16f.html

I have successfully generated Full_bridge signals.

820 uF capacitor for DC rail:
Did you calculate it´s initial switch ON current?

I am new to practical power electronics.

Please tell me how to calculate the initial switch ON current. Should I use the basic transient analysis formulas for the capacitor circuit ?

I just need good filtering of the DC ripple.

What about PFC? (Your circuit is not allowed to operate in many countries. Is it allowed to operate in your country?

I am not sure about it but I do want to implement a Active PFC circuit using one Mosfet and inductor.

You didn´t take care about power factor of your load devices.

Yes, I didn't take the power factor into consideration.

What safe value of pf can I take for calculations ? 0.8 ?

Loads will be resistive or inductive.

You didn´t care of load peak current during start up.

I don't know the peak start up current of the refrigirator or AC compressor motor.

You dind´t take care about MOSFET peak current during switching.

How is Mosfet peak current different from the calculated peak current ?

What about output low pass filter?

Yes, I will add one later after stabilizer works fine on hardware.

 

[KlausST]

re: Has anybody done this ? (AC >DC >AC converter)

Hi,

sorry the "8" was missing. "10V" should be "180V".
I edited my post now.

Klaus

 

[Okada]

re: Has anybody done this ? (AC >DC >AC converter)

How can you get stable output voltage of 230V AC with only 180V (edited) AC input?

I have to use SPWM for getting it. I still have to test it practically. I am waiting to receive my Rigol Oscilloscope. I need to get the PCB layout designer, PCB manufactured and components assembled before doing any testing.

I am referring these.

https://tahmidmc.blogspot.in/2013/02/demystifying-use-of-table-pointer-in.html

https://tahmidmc.blogspot.in/2012/11/feedback-in-sine-wave-inverter-pic16f.html

 

[KlausST]

re: Has anybody done this ? (AC >DC >AC converter)

Hi,

SPWM can not increase voltage...

Klaus

 

[Okada]

re: Has anybody done this ? (AC >DC >AC converter)

Here is doing it here in an inverter.

He takes battery voltage as 10V to 12V and based on that VBUS voltage varies and he uses SPWM and varies its duty to get constant voltage.

https://tahmidmc.blogspot.in/2012/11/feedback-in-sine-wave-inverter-pic16f.html

 

[KlausST]

re: Has anybody done this ? (AC >DC >AC converter)

Hi,

please read the complete article.

The SPWM can´t step up. In the article he used a transformer to step up.

But in your circuit there is no transformer.

Klaus

 

[Okada]

re: Has anybody done this ? (AC >DC >AC converter)

See in his circuit a battery is used to get VBUS voltage after pulse transformer. So, VBUS voltage varies as battery voltage varies. The output voltage of the Inverter varies based on two factors that is

1. load and
2. VBUS voltage.

So, to correct this (whether he uses output transformer or not)

he varies the SPWM duty to get constant voltage at the inverter output. He is feeding back part of the output voltage rectified to the adc.

I want to do the same but only difference is instead of using battery I am using mains to get VBUS voltage.

Or should I use a buck-boost converter to get constant 325V DC from 180 to 300V AC ?


I have different threads for Pure Sine Wave Inverter, SMPS Battery Charger, Buck Converter, etc... but actually these all belong to the same project that is
to design a 10 KW Constant Output Voltage Stabilizer Cum Inverter which uses the same H-Bridge circuit for the output.

Now, I think that I have to design 180V-300V AC input to constant 325V DC VBUS Voltage Buck-Boost Converter.
A 180V-300V AC input to 12V (this may vary depending upon the voltage required for multiple batteries) Buck Converter for Battery Charging
One common h-Bridge circuit for Stabilizer and Invereter.
No bulky stabilizer ot Inverter Output transformer.
I also need Active PFC.

- - - Updated - - -

One circuit to get VBUS voltage from mains and another to get from battery or battery bank.


I just have some doubts regarding precautions to be taken like

What happens if terminals of a 12V 200Ah battery are shorted ? Will the battery explode ?

If I use a 230V AC to 14.4V DC buck converter for battery charging then what happens if PWM duty is 100% or if mosfet shorts. Will the 325V DC appear across the battery terminal and explode the battery ?

 

[KlausST]

re: Has anybody done this ? (AC >DC >AC converter)

Hi,

Please read what I have written:

SPWM can not increase voltage...

For sure you can lower it, you can regulate it, but not increase.

And there is no part in your circuit that is able to increase voltage.
If you want to get an output of 230V with an input ov 180V then you simply need any part that is able to increase voltage.

It can be a transformer, it can be a (something..) boost converter.. many things are possible, but I don´t see it in your design.

Klaus

 

[Easyrider83]

re: Has anybody done this ? (AC >DC >AC converter)

Polarity of capacitors mixed

 

[Okada]

re: Has anybody done this ? (AC >DC >AC converter)

I have decided about it. I can't and will not use bulky transformers. If it can't be done with just rectifying the mains AC then I need a buck-boost converter to get constant 325V DC VBUS voltage from 180-300V AC.

In my design as 4 PWMs of different frequencies are needed I will be using 4 different micrcontrollers of different pin count.

One for the H-Bridge
One for the Buck-Boost Converter to get VBUS Volatge
One for the Buck Converter used for battery charging
One for the DC-DC Converter to get the 325V DC VBUS voltage from 12V battery

I have another thread for Buck Converter but nobody is answering in that thread.

@Easyrider

Which Capacitor ? C17 and C20 ?



Another question I have. Should I use the mosfet mentioned in the circuit or should I use STE40NC60 ?

This mosfet has a little high Rds(on).

For 7A the mosfet in schematic dissipated 2W and this mosfet dissipates 6W but this mosfet can be easily mounted on heatsink with ring terminal wires and no worries about PCB track width.

The mosfet in schematic is not good for 10KW stabilizer cum inverter I think because it is PCB mount type and if I mount them on PCB then I have to worry about PCB track width and if I mount it on heatsink and not PCB then I have to solder wires but I am not sure if the solder handles 20A or more current.

Also I want to use one mosfet for both H-Bridge and also for the buck and buck-boost converters. In both buck converter used for battery charging and buck-boost converter used for VBUS voltage 20A or more current flows and I am thinking about using parallel mosfets in these converters also.

It is a 14 USD mosfet.

I have to buy them say 2 + 2 for the buck and buck-boost converters and 12 for the H-bridge. I need to buy all components in one go from mouser because mouser charges 40 USD shipping charges.

So, help me in deciding the right mosfet.

- - - Updated - - -

@Easyrider.

Thank you for pointing out the reversed capcitor connections.

I changed the connections of C17, C19, C20 and C22. Now they are correct.

 

[FvM]

re: Has anybody done this ? (AC >DC >AC converter)

Before ordering components you should calculate some parameters of the AC/DC part.

- Input RMS current
- VDC voltage ripple
- Capacitor ripple current (check against datasheet rating)

The input and capacitor current will depend on power grid impedance but is usually quite high in relation to output current. The present rectifier circuit isn't suited for output power above 1 to 1.5 kW.

 

[Okada]

re: Has anybody done this ? (AC >DC >AC converter)

Ok FvM. I will do.

Why nobody has answered what happens if

Buck converter's (used for battery charging, 230V AC to 14.4V DC) mosfets shorts ? Does the 325V DC appear across 12V battery ? Will the battery explode ?

I have only exploded small capacitors and coin cells.

I need to know these to take precautions.

 

[FvM]

re: Has anybody done this ? (AC >DC >AC converter)

Why nobody has answered what happens if Buck converter's (used for battery charging, 230V AC to 14.4V DC) mosfets shorts ? Does the 325V DC appear across 12V battery ?

Everything else (cables, transistors, PCB traces, bus capacitor, fuse) will explode before. Nobody will make a 325 to 12 V buck converter without transformer due to the dreadful efficiency, at least not above a few W output power.

 

[Okada]

re: Has anybody done this ? (AC >DC >AC converter)

I am new to Buck, Buck-Boost, Cuk Converters and thats why I asked the question. I need to know these correctly to build the 10KW Stabilizer cum Inverter.

Saftey is first.

Are there buck converters with transformers ? If yes, are they ordinary step down transformers are switching transformers ?

Can I use ordinary 230V to 15V step down transformer for my battery charger buck converter ? If yes, the transformer will be bulky because I need 20A charging current for 200Ah battery. Then should I go with SMPS battery charger ?

The present rectifier circuit isn't suited for output power above 1 to 1.5 kW.

Then should I use high voltage high current rated SCRs for the rectifier ? For 10 KW Stabilizer/Inverter ?

rectifier ok for 500W stabilizer/Inverter ?

 

[FvM]

re: Has anybody done this ? (AC >DC >AC converter)

Problem of the rectifier circuit is the large "crest factor" (peak to average respectively RMS current). It's won't be better anyhow with a SCR rectifier. 500 W should work well. Still need some kind of inrush current limiter, e.g. NTC or series resistor with bypass relays.

10 kW inverters with single phase power input are very unusual, we see up to 2.2 kW for industry standard VFD inverters with single phase 230 V supply.

 

[Okada]

Actually I was thinking about buying 5 KVA Inverter for home but price is very high. I was thinking to buy this APC Inverter.

**broken link removed**

The price is too high without batteries.

APC makes Inverters upto 20 KVA.

https://www.apcindiastore.com/sep17...3YgBty5_P3i6hjEXBhNI5uqULO59CNDBxgaAs5L8P8HAQ

Price are very high without batteries.


Still need some kind of inrush current limiter, e.g. NTC

I am planning to use Littlefuse PPTC Resettable fuses.