The only electronic website that answers to every comment, and addresses all circuit related issues..

## H-Bridge Inverter Circuit Using 4 N-channel Mosfets - Modified Sine Wave

The following post describes an H-bridge modified sine wave inverter circuit using four n-channel mosfets. Let's learn more about the circuit functioning.

We all know that among the different inverter typologies, the H-bridge is the most efficient one, since it does not necessitate the use of center tap transformers, and allows the use of transformers with two wires. The results become even better when four N-channel mosfets are involved.

With a two wire transformer connected to an H-bridge means the associated winding is allowed to go through the push pull oscillations in a reverse forward  manner. This provides better efficiency as the attainable current gain here becomes higher than the ordinary center tap type topologies.

However better things are never easy to get or implement. When identical type mosfets are involved in an H-bridge network, driving them efficiently becomes a big problem. It is primarily due to the following facts:

As we know an H-bridge topology incorporates four mosfets for the specified operations. With all four of them being N-channel types, driving  the upper mosfets or the high side mosfets becomes an issue.

This is because during conduction the upper mosfets experience almost the same level of potential at their source terminal as the supply voltage, due to the presence of the load resistance at the source terminal.

That means the upper mosfets come across similar voltage levels at their gate and source while operating.

Since as per the specs, the source voltage must be close to the ground potential for efficient conduction,  the situation instantly inhibits the particular mosfet from conducting, and the entire circuit stalls.

In order to switch the upper mosfets efficiently  they must be applied with a gate voltage at least 6V higher than the available supply voltage.

Meaning if the supply voltage is 12V, we would require at least 18-20V at the gate of the high side mosfets.

The proposed H-bridge inverter circuit having 4 n channel mosfets tries to overcome this problem by introducing a higher voltage bootstrapping network for operating the high side mosfets.

N1, N2, N3, N4 NOT gates from the IC 4049 are arranged as a voltage doubler circuit, which generates about 20 volts from the available 12V supply.

This voltage is applied to the high side mosfets via a couple NPN transistors.

The low side mosfets receive the gate voltages directly from the respective sources.

The oscillating (totem pole) frequency is derived from a standard decade counter IC, the IC 4017.

We know that the IC 4017 generates sequencing high outputs across its specified 10 output pins. The sequencing logic shuts subsquently as it jumps from one pin to the other.

Here all the 10 outputs are used so that the IC never gets a chance of producing incorrect switching of its output pins.

The groups of three outputs fed to the mosfets keep the pulse width to reasonable dimensions. The feature also provides the user the facility of tweaking the pulse width that's being fed to the mosfets.

By reducing the number of outputs to the respective mosfets, the pulse width can be effectively reduced and vice versa.

This means the RMS is tweakable here to some extents, and renders the circuit a modified sine wave circuit ability.

The clocks to the IC 4017 is taken from the bootstrapping oscillator network itself.

The oscillating frequency of the bootstrapping circuit is intentionally fixed at 1kHz, so that it becomes applicable for driving the IC4017 also, which ultimately provides about 50 Hz output to the connected 4 N-channel H bridge inverter circuit.

The proposed design can be much simplified as given here:

More Circuits Recommended for You...

1. @swagatam, the values of the components is not shown. Pls what values are required?

1. click the diagram to enlarge, the values are given in the diagram itself.

2. Exactly sir, can the circuit be used for 220vdc inverting, like i can see some kind of arrangement in chopper transfomer inverters

3. yes it can be used.

1. I'll possibly do it soon...

3. But how could an input voltage of 220vdc be regulated to 12v in my quest to use your circuit to invert 220 to supply the circuit. I have a little diverted, pls bear with me.

1. I did not understand your question, please provide a clear description of your requirement, are you referring to compact ferrite-core inverter circuit??

4. Dear Swagatam . It is really an innovative idea.

5. Verily true quoating you, "I did not understand your question, please provide a clear description of your requirement, are you referring to compact ferrite-core inverter circuit??"
Yes because that is where they primarily use a centre tap ferrite-core to generate high frequency ac then rectify and smoothen it to give that kind of around 350vdc and then use a h-bridge connection to re invert to 220vac.
So how applicable is your circuit to drive the transistors assuming their voltage rating is upto 500v. That is why i want to assure myself b4 testing your idea. Thanks.

1. The mosfets can be connected to the high voltage DC in the above circuit for your application. the IC stages may be powered from the battery.

It might work, but not sure.

6. Pls another observation.
How could constant voltage be maintained when more load is applied on the output.

7. This comment has been removed by the author.

1. for that we will have to add another controller stage, I'll possibly update it soon.

8. dear swagatam
everything looking very nice, i need your help with making pure Sine Wave inverter to build. i have mod-sine wave inverter 3500w and i was hoping to make it to pure sine wave out put. is this can be done?

regards
Mohamed

1. Dear Mohamed,

It may be possible, you may take the help of the following article

9. http://www.eeweb.com/electronics-forum/need-help-regarding-ups-as-a-final-year-project/
I have exported one of your ideas here. You may login to see.

10. sir swagatam,
what the value of transformer?
the out put is ac or dc?

1. shabbir, transformer rating will depend on your required specs, the output would be AC

11. Hi Kassim,

I know that you have done it with a good intention, but instead of exporting the diagrams it would be better if you invited your friends to my site by providing them the relevant links.

12. As you could see there on the circuit not a single value is given, the student needed to know that to get the values he needed to hit on the 'link' to extract the values to execute the project. That is the essence of the internet global village.

1. OK that's good, I didn't notice that.

13. Hello Swagatam
i did your circuit but it did not work and when i test the out AC Volt I find it only 4 Volt and i supply the circut by 12 DC Volt also i test the out of CD4049 it is only 8 Volt can u help me and l look for all lines and componenet and all is ok
thanks

1. It would be difficult to say without knowing about the entire set up that you have used....is this your first inverter project? are you well versed with the concepts? if its your first attempt then you would first have to know many things related to the subject for succeeding...

the output from the 4049 stage should be at least 18V, it's a tested design

14. congrats sir swagatam nw you can create this H-brigde circuits ,and what I can see no more preset this design that I realy find .thanks sir swagatam your great Engineer hoping you can devilope this circuit..

1. My pleasure Roderel.

15. Dear Mr Swagtam,
i need your help in his circuit i have 22 Volt on the Collector or the 2 BC547 but on the High Mosfets gate the volt is 4.0 Volt and i check all lines and connection and also i test the out of CD4017 by Led it is pulsed so i have only 4 Ac volt before the Transformer after it i have 80 volt only so i need your Suggest and help to solve this problem
thanks too much

1. Disconnect the mosfets completely from the transistors, and supply.

after disconnecting the gates from the emitter of the transistor, check the voltage at the emitter of the transistors.

tell me the results, then we can proceed.

2. Dear Swagatam

i Disconnect the gates of Mosfets from the Emitter of Transistors and i have the Collector Volt is 22 Volt and the Emitter Volt is 4 Volt
this problem make me Crazy so I change the Emitter 1 K Resistor by 10 K and the Emitter Volt is 8 Volt

:(

Plz Told Me where is the problem

3. Connect a BC557 transistor stage with the BC547 transistors in the following manner:

connect the emitter of the BC547 directly to ground.

connect the collector of the BC547 to the 22V source via a 1K resistor.

connect BC557 base with the collector of BC547 via 10K resistor.

connect collector of BC557 to ground via a 1K resistor.

Connect the mosfet gates directly to the collector of BC557

4. thanks Swagatam for ur help
But you forget to told me about the Emitter of BC557

5. emitter of 557 will go to 22V source.

16. hey i would like to build that circuit but where do i get
parts

17. Iam not sure ur configuration is ok, why not upload ur connection to identify the problem visually, anyway Swagatam is there.

18. my brother thanks for the nice idea,, but my question is that is this a sine wave inverter? is it suitable to carry inductive loads?

1. Hi Tessy,

By removing one or two diodes, the design can be made into a modified sine wave, but I am not sure about the exact results how it would be.

19. sir Swagatam, can I use this for ic sg3525 inverter circuit and will I need to use two of ic 4017 since sg3525 has two output, please help me on this, cos am having problem with ic ir2110 for full H bridge and I see yours is better. thanks.

1. Foma, if you want to include SG3525 then you won't require the IC 4017.

you may replace the 4017 IC with sg3535, the outputs of the SG3525 now directly gets connected with the mosfets and the transistors. The following diagram will provide a better picture.

The connection from the N1 and N2 gates can be ignored because the sg3525 will have its own separate oscillator.

20. hello sir. i have found in some website,the H-bridge was made of 2 n-channel and 2 p-channel MOSFET, but here H-bridge is made up of 4 n-channel MOSFET. is this given circuit is right?

1. using 4 n-channel mosfets is more efficient, therefore 4 n-channel bridge config is always better.

21. can you please tell me how 4 mosfet n-channel h bridge works in few sentences.i am in a great confusion.

1. with every alternate pulse from the IC, the diagonally situated mosfets switch ON together alternately.

22. hello sir swagatam
i have a question about the circuit that i already build. when i test my circuit, one irf 540 from pin #6 get blowed. would you like to tell me why it was happened.
(sorry for my bad english) thx before

arif from indonesia

1. Hello Arif,

23. Swagatam
Thanks for the information,
I will try it and give you feed back

24. Thanks
I am Tunde,
I have been trying to build an inverter with this H-bridge concept and you have finally come to my rescue
I am very great full Swagatam.

1. you are welcome Tunde

25. Hi Mr. Swagatam,

I've recently taken interest in your blog as I'm working on a similiar project like this. What I'm trying to do is feed in 12VDC into a Flyback converter which will step up to 240VDC. From here, I wish to invert this supply to 240VAC using a H-Bridge inverter.

Can I email you my schematic and perhaps teach me a few things? If yes, where can I drop my email to?

Apologies for commenting as Anonymous but this is written in genuine intention to learn from you, Sir.

Thanks

1. Hi, thanks for reading my blog!

yes, you can send it to homemadecircuits@gmail.com

26. using solar panel as input can i get 600W at output side in this inverter circuit...??
and what should be the circuit and specification of transformer and solar panel....???

1. for 600 watts, the solar panel must be at least 200 watts rated and the transformer at 800 watts.

The battery should have minimum current capacity of 200ah.

27. Hi, Swag,
I am have built an inverter but the problem lies in the charging area in that when there is power from the mains and the inverter is charging, do we have to build a different charger for the inverter or we should allow the power mosfet to charge the battery.
Also what are the side effect of charging with the mosfet. And is it advisabe to build a different charger for the inverter.

Regards

1. Hi Itoro,

Which circuit are you referring too?
All inverters mostly use a separate transformer or winding for charging.
Without actually inspecting the circuit it will be difficult for me to comment precisely.

28. hi swagatam
this is srinivas cani use this design for 3kva transformer (0-36 volts)220volt out put,please suggest me

regards

1. Hi Srinivas,

yes you can use it.

29. hi sir in this circuit i want to ask you that the IC4049 please sir is N1 start from (2 to 3) and (N2 4 to 5) up to 15? i guess you are understand what i mean cos im confused about it i build it but it don't work that is why im asking maybe i didn't connected the pins correctly thank you sir im looking forward to hear from you son.

1. Hi bianzz, all gates inside the 4049 are identical so you can organize them as per your choice, only the input and outut pins of the selected gates needs to be connected correctly as per the diagram.
check the datasheet of this IC you will get the idea regarding the involved gate pinouts.

30. ok sir thank you but sir please im still in the 4049 some of the pins is not shown in the circuit such as pin8 and pin16 i can see the pin8 is gnd or ground please sir where is the pin16 connect to? that is all my problem im looking forward to hear from you.

1. bianzz, what does the datasheet say about pin16? Please follow exactly as mentioned in the datasheet

31. hi swagatam,
thank you for sharing your knowledge
I have implemeted your circuit but it has some problems.
the first is oscillating frequency. How 1 khz on the 4017 ic will be 50 hz on the h bridge?

The second is that the output is 220 volt ac without load but with load the volt decreases suddenly to 40 volt ac.
How could I preserve the volt with load?

1. Hi Ahmed,

I have divided the 1khzwith the total number of outputs used from the IC 4017, so 1000/10 = 100Hz, and this fvrequency when divided across two mosfet arms gives 50Hz.

You can adjust it as per your own calculations if the above results look incorrect.

The output voltage will depend on trafo wattage and battery AH capacity, they must be sufficiently rated well above the load wattage, otherwise it will drop proportionately.

32. Hi Swagatam
I got the fets switching properly now,I removed the 1k resistor from the voltage doubling circuit to the bc557 emitter and put a wire there.I didnt get any voltage there but got 21.2v at the collector.
In the other H-bridge circuit there is a resistor there,maybe there should'nt be?
There is an extra resistor at the top right high-side fet which i left out.
I'll do some further testing tomorrow

1. Hi Robin, that's great keep it up...

appreciate your outstanding efforts in verifying my circuits.

33. Hi Swagatam
Just one other thing,the Zener diode part seems to bring the voltage down so i left that out and connected directly to the battery
cheers

1. Thanks a lot Robin,

if possible please do send the images of the working prototype!!

2. Hi Swagatam
I had my oscilloscope on ac instead of dc for channel 1 therefore could'nt pick up the voltage doubler reading(big mistake lots of time wasted)

Ok i got the fet's switching 100%,Trafo used=12W {220v in and 12vac 1000ma out}.
output voltage=180v if i used trafo=220 in and 9v out i would've got higher output voltage
I will e-mail you the images of 1) low/side fet gate voltages
2)high/side fet gate voltages
3)trafo output waveform(there are 6*10k resistors across the output,so i put the scope over one 10k)
I will e-mail you the images immediately
cheers

3. OK no issues Robin, you are doing a great job

so far it's been very interesting, let's see what finally transpires from the efforts.

yes a 0-9V trafo would surely yield the required 230V

34. Hi Swagatam
The moment i connected a trafo i relized the fets were still not switching

I replaced the 1k resistor that i removed from the voltage doubler to the 557 emitter with a 100k resistor and got 23.5v at the emitter(which i think is correct?)But now it does'nt switch through to the collector.Remember previously i just had a wire.
I need that voltage to switch through to the collector and thus the High/side fet gates
Shoo,but i think we're getting there
cheers

1. Hi Robin,

you may have to check the stages separately.

first disconnect the bases of the mosfets from the transistors, next disconnect pin14 from the 4049 clock and feed the clocks from an external oscillator at a very slow rate, may be at around 0.5 Hz.

Next, you can confirm whether the transistors are providing the required voltages at the intended levels over the indicated points or not.

Once this is confirmed then you can integrate the mosfets and continue checking the results without altering the 0.5 Hz sequence rate.

If there's any fault somewhere the above method will be able to trace it out.

2. Oh,I just want to mention that i am working with a 100AH battery,its big enough and fully charged

35. Hi swagatam
I put another diode on each leg of the 4017 outputs ie:2,4,7,10 and 5,6,9,11 and the output voltage rose to 220v.
When i put a load(4w) on the voltage drops considerably even after increasing the trafo wattage from 12w to 120w i still get quiet a big voltage drop, now to prevent this voltage drop would puting fet's in parrallel(ie doubling up on the mosfets) so as to entice more current to flow through the trafo solve the problem?
And would the 4017 ic be able to switch for instance 8 mosfets comfortably
Regards

1. Hi Robin,

If you are using a 120 watt transformer with a 100Ah battery and fets specified to carry currents above 20 amps then the voltage should not drop unless the load exceeds 120 watts.

so i don't think paralleling the mosfets would improve the results because the problem could be somewhere else.

yes the 4017 outputs and the BJTs are all capable enough to make any msfet trigger fully since the gate impedance of an FET is very high,

May be you could switch to a professional full bridge design using a single IC fully bridge driver as explained here:

36. Dear Swagatam Sir,

I tried to use 2 N MOS and 2 PMOS for inverter circuit.I used IRF540 and IRF 9540.But when I increase the supply voltage beyond 8 V,the current increases abruptly and my DC voltage source get shorted.Can you please figure out the problem?

1. Dear Faris, it will be difficult to troubleshoot because an H-bridge using N and P mosfets can be extremely unpredictable, that's why the use of specialized H bridge driver IC is recommended for implementing such topologies safely...

37. hii Mr. Swagatam... I have checked my HCF4049UBE on the google and found of its vdd is 10V, so .. this device still can be used for 12V vdd or needed to regulate it to 9V or should i go for 4049 with the specification more high (15V) VDD ??

1. Hi Mr.Hui, the IC HCF4049UBE is specified to work with supplies from 3 to 20V, so it can be used safely with a 12V supply, just as any other IC 4049 variant

2. ow.. im sorry.. may be i little careless to see datasheet.. after i saw next page .. yess.. ur right sir the voltage supply is so wide .. .. i have been tricked with vdd 10v has pF ..on the first page.. i will careful next time.thx , sir..

3. no problem Hui, that sometimes happens:)

38. dear Sir, how i control my inverter output voltage level. my inverter output voltage is 330 volt but i want 240 volt. please tell me about that. thanks

1. Dear Pizush, you can incorporate the following circuit with your inverter for correcting the issue: