Tuesday, May 1, 2012

How to Design a H-Bridge Circuit for Modified Sine Wave Inverters

In electronics H-bridge circuit refers to a configuration consisting of four individual switching devices like transistors or mosfets, such that these can be switched through external discrete signals from respective stages of the involved control circuit.

The above configuration is bridged or rigged in a form which resembles the letter "H" and hence the name H-bridge.
 The above special formation has a specific reason behind it.

Looking at the circuit diagram we see that the two arms of the bridge consists devices which are complementary to each other.

While switching, the complementary pairs from each of the arms switch together, meaning the NPN device from the left arm and the PNP device from the right arm of the bridge conduct together in response to an external applied signal.
Similarly the other complementary follow the switching pattern and the two pairs conduct in tandem at a given frequency.
The arrangement generates a push pull effect in the connected transformer winding of the inverter which in turn generates a required one full cycle of AC at the output of the transformer.

The H-bridge configuration has another great advantage with respect to the making of modified sine wave inverters.

The special arrangement of the devices make it possible for driving the devices through two individual signal sources, one which decides the switching rate of the output while the other decides the RMS value of the output from the transformer.

The given circuit idea shows how a H-bridge circuit may be designed for a particular modified sine wave inverter application.
 The two gates carry the 50 Hz signal from the source to T1 and T3 such that they conduct alternately.

Transistors T2 and T4 ae fed with the PWM pulses which switch ON together, however when T1 is conducting, T2 should remain shut OFF and similarly  when T3 conducts T4 should remain OFF, the two diodes from the respective NAND gates takes care of the issue and allow only the relevant transistors to carry out the PWM switching.

Thus the NPN transistors are responsible for producing the regular 50 or 60 Hz pulses while the PNP transistors are responsible for breaking the square waves as per the PWM sections.

The resultant pulses generate the exact intended, optimized modified sine wave waveform at the output of the transformer.



57 comments:

  1. Hi, congratulations on this configuration you can give the list of materials and see it says 12 + and 12 - this configuration is symmetrical with power? 12 + 12 - and gnd?
    thank you very much.

    nestor.

    ReplyDelete
    Replies
    1. Hi,

      Here I have just tried to explain the concept, I'll be publishing one complete example circuit based on the above concept very soon, +12 and -12 refers to the battery poles, so there's no GND involved here.

      Regards.

      Delete
    2. Thanks for congratulating me, appreciate that.

      Delete
  2. Hi Swagatam

    Thanks for all your generous efforts. I am waiting for an example circuit utilizing above. I am also waiting for load independent 220v support in your modified square wave inverters otherwise they are perfect to utilize and experimentation.

    Thanks again for each and every circuit diagram and explanation.

    Regards

    Sam

    ReplyDelete
    Replies
    1. Hi Sam,

      Great to see you back after some time! Yes I'm trying hard to design the circuits that you have suggested, once I make them I won't waste a second, and post them immediately for you....

      Thanks very much for keeping me motivated.

      Best Regards.

      Delete
  3. Hello again Mr Swagatam, I have to tell you that this is a nice article. I have previously explored h-bridge inverter designs but I left them because they mostly produce square waves and am glad you have invented a way to to produce modified sine waves using this configuration. a couple of questions though,

    1. At what frequency will the pwm be generated and what is the relationship between supply frequency and pwm frequency?
    I have q great deal of interest in designs implemented using ic 555 because it is easily and readily available here in Nigeria so am very much interested in this design. I went hunting for 4049 which was used in one of your inverter designs but I couldn't get it. couldnt find 4017 decade counter either.

    ReplyDelete
    Replies
    1. Thank You Kopje,

      The supply frequency is the one which is fixed as per the country specs, either 50 or 60Hz, PWM frequency may vary as per our selection.

      The PWM frequency should be generally 100 times higher than the supply frequency, however there's no standard rule because it's after all adjustable as per the needs.

      Regards.

      Delete
  4. Hello Mr Swagatam, thank you very much for your response. a few more quick questions.
    1. Won't it be necessary to add a low pass filter after the h-bridge to filter out the pwm frequency? Because from my understanding, you will have two frequencies (50hz + pwm frequency) at the primary of the transformer. Or is it unnecessary to add a low pass filter?
    2. Will it be possible to post an image of what the output waveform will be like?

    And thank you very much for your time and effort. You have been of immense help. Cheers.

    ReplyDelete
    Replies
    1. Hello Kopje,

      A low pass filter is not required here, because the question of filtering PWM does not arise, rather we want the PWM to remain in the waveform.

      I'll will try to update the waveform image soon.

      Regards.

      Delete
  5. Okay sir....thank you very much. I have decided to implement this design for a 5KV inverter using parallel mosfets to increase output power. I will update u the result of m work when am done. Thank you very much sir. Also waiting for the waveform image.

    Much regards sir.

    ReplyDelete
  6. Sir, I want to design a simplest FM transmitter circuit.. Will u help me out..

    ReplyDelete
    Replies
    1. Ho Pradeep,

      You can try the circuit shown in this link, the second diagram:

      http://homemadecircuitsandschematics.blogspot.in/2011/12/how-to-build-electronic-spy-bug-circuit.html

      Regards.

      Delete
  7. Mr Swagatam, good day to you sir. I am wondering if there is a way I can send you the schematic of my pwm inverter design which is based on your design.

    ReplyDelete
    Replies
    1. Hi Kopje,

      Good day!
      send it to hitman2008@live.in

      Delete
  8. Dear Sir Swagatam,

    thank you for your very nice circuit, and for sharing your knowledge with us, it's really great!

    I read in one of your comments that you were going to publish a complete version of this circuit with +/-12 V as supply. Did you already publish this? Where can I find it?

    I have to control the temperature of a little box using AC current, and I think that this solution can be attractive for me. How can I control the amplitude of the output sinusoidal wave though?

    Thank you again for your kindness.

    Fabrizio

    ReplyDelete
    Replies
    1. Dear Fabrizio,

      Thank you very much for appreciating mu work!

      If you could explain your need in details then probably I can come up with a much more suitable idea than the above, because I think there are much easier options than the above with similar implementations.

      Regards.

      Delete
    2. Hi Mr Swagatum
      I've been reading a lot of your posts on inverter circuits which control the centre tapped transformers (12v-0-12v to 220v) and came accross the H-bridge inverter circuit diagram for modified sine wave,which controls the normal 12v-0v to 220v transformer.I would like to know if you've published the complete example circuit based on the below concept

      I would like to build a 500w inverter with this concept,as the centre tapped transformers are very expensive and I have easy access to the normal transformers
      (http;//homemadecircuitsandschematics.blogspot.com/2012/04/how-to-design-h-bridge-circut-for.html
      Regards
      Robin

      Delete
    3. Hi Robin,
      The above concept is based only on my assumptions and has never been tested practically, therefore I am sorry I cannot provide you with the example.

      By the way you may refer to this circuit also, which is much simpler and looks more feasible.

      http://homemadecircuitsandschematics.blogspot.in/2012/10/simplest-full-bridge-h-bridge-modified.html

      Regards.

      Delete
    4. Hi Swagatam
      I'm still going to persevere with my original goal of including PWM in a H-bridge,the above circuit was maybe to complex for me to build at first,but now that we have built a simple H-bridge using P and N channel mosfet's and built a 400W mosfet PWM inverter (also been able to vary the output voltage as required)maybe now we can attempt the above circuit?
      The final circuit should have a good quality modified sine wave(maybe to run soffisticated electronic equipment) and have automatic voltage regulation for varying loads.
      If I can show my appreciation for what you have done for me in any way,let me know
      Regards Robin

      Delete
    5. Hi Robin,

      Your relentless hard work and valuable inputs are all commendable.

      I appreciate every bit that you have provided to this blog and the interested readers, and we all thank you for that.

      By the way if you would like to contribute your own written articles for this blog, you work will be highly appreciated.

      Best Regards.

      Delete
    6. Hi Swagatam
      With your guidance and expertise I've been able to build the above concept and it works.
      If you can foresee any problems please comment and fine tune it,it is by no means perfect.
      I will send you a circuit diagram to hitman2008@live.in,there is a 1k resistor missing just before the base of BC547.
      You can publish it if you want.
      Cheers Robin

      Delete
    7. Hi Robin, thanks a lot!

      The above circuit was posted a long time ago and since then until now my knowledge level has increased considerably, so surely the above circuit has room for improvements. I'll check your circuit and study it.

      By the did you see your waveform images, posted here:

      http://homemadecircuitsandschematics.blogspot.in/2013/10/modified-sine-wave-inverter-circuit.html

      Delete
  9. i post a message here dont know if u got it Mr Swagatam Alex

    ReplyDelete
    Replies
    1. Alex, all pages will respond correctly to the posted comments, so you can be sure it will reach me, but avoid pages that already have more than 200 comments in it.

      Delete
  10. ok Sir i am glad we are still in touch. As i said i did not get to fully complete ur circuit yet because of some problems I am having. I m using a h bridge driven by a 60hz oscilator and then pwm is induced as u stated with the two dioes. This works fine so far on 12v supplying the h bridge but as soon as I raise the voltage to even 18v it blows the fets in the h bridge. I am using irfp260n which are 200v rating Sir. Are there any other changes that we are to do when raising the h bridge voltage I am using 12v regulator to supply the rest of the circuit. Please help us here Sir just dont know but it works ok on 12v. I am awaiting ur reply Sir. Thank u Alex

    ReplyDelete
    Replies
    1. Alex, the topology is suggested by the IC manufacturer so it's beyond my control to troubleshoot the problems, moreover mosfets could be prone to many factors which will need to be sorted out and rectified step wise, I have discussed a few of them in the below article, you may have a look at them

      http://homemadecircuitsandschematics.blogspot.in/2013/09/mosfet-protection-basics-explained-is.html

      Delete
  11. hello Mr Swagatam I had given the inverter a break because I was having some problems. I am not sure u understand my post. What I was saying is I am still supplying the 12 section of the circuit with 12v the only part i increase is the supply to the drains of the fets in the h bridge to 18v and I start to blow fets. If I use the entire ircuit on 12 v it is fine but when i increase the fet v I have problems so I was asking if I am going to raise the fet supply v if there is any other change I need to make in the circuit also to compensate for the increase hope u understand me better now Sir alex Thank u for ur support

    ReplyDelete
    Replies
    1. Hello Alex, I hope you have gone through the design shown in the following article.

      http://homemadecircuitsandschematics.blogspot.in/2014/01/simplest-full-bridge-inverter-circuit.html

      It clearly shows that the fet supply input could be as high as 600V, so 18V is nothing compared to that.

      That's why I suggested that your problem could be specific and cannot be troubleshooted without practical examination

      Delete
  12. ok Mr Swagatam i will recheck again, I did not get the ic u recomended but I was using an npn transistor to drive each high side with boostcap config and then take the signal straight from the output of a sg3524 to drive the low side so please instruct me id the design is wrong Sir. thank u

    ReplyDelete
    Replies
    1. If you are not using a specialized driver IC then I am afraid troubleshooting could become even more difficult. your configuration may be correct but it could have many hidden issues. Try using BJTs instead of mosfets, if these also become hot would confirm a faulty design.

      Delete
  13. thank u sir could u recomend a number bjts thank u

    ReplyDelete
    Replies
    1. you can try TIP122 initially and a smaller 3 amp transformer for experimenting

      Delete
  14. thank u Mr swagatam for ur support. I want u to help me unerstan something fully . I read in one of ur post that h brige is the most efficient topology ok . You see I am having problems with building the h bridge and getting it to work with higher voltages as u see so I want to know if I use the pushpull that u designed and I built using the PWM concept with the diode principle and still get a efficient inverter. Sir I am at a cross road now and I want u to help me decide if the pushpull can still be efficient for my house and on battery consumsion. Thank u for ur support .

    ReplyDelete
    Replies
    1. Mr. Alex, if you are not able to make it work, it could be your personal problem, the concept cannot be blamed,

      center tap push pull topologies are also as efficient but it will make things bulkier that's the only inefficiency with it, battery consumption efficiency is more or less equal for both the concepts.

      Delete
  15. ok Mr swagatam now I see things much clearer now as u explain. i was not blaming the concepts because I know they work but that I was having some problem with the h bridge but I was getting better results with ur puspull design working as a sine wave . I went back to ur design this morning and I am getting less heating now I am using a totem to drive the fets and I am doing some more tests on bigger heat sinks so as soon as everything is good I will reply to u and others who are on this project but for sure ur pwm works.Before as I read I thought the pushpull design was so poor on conversion from Dc to Ac but as u say bulhier but if its bigger transformer no problem it is for my house purpose Sir. I will let u know as the progress goes on. Thank u for ur support

    ReplyDelete
    Replies
    1. Sure Mr.Alex, you can continue with the center tap topology, and make it as efficient as an hbridge if a bigger transformer is not an issue

      heating up of the mosfets could be reduced by using additional BJT stage at the 4017 outputs as given here:

      http://1.bp.blogspot.com/-NgqlieM8TX8/U7zL0tGH6qI/AAAAAAAAHgE/WCKLVG9pVVs/s1600/mosfet+transistor+driver+circuit.png

      Delete
  16. thank u Mr Swagatam really appreciat ur time for helping us we really appreciate ur support. Earlier u sent me a transistor number as bjt but I think this is not correct Sir could u check and repost the number. U sent me tip122 but this looks like a darlington insted. Please help me here because I am getting better results now . At rest I am having normal heat. I only start to have heat now when I plug a compressor in drawing 12 amps on 48v system sine wave so I want to try the bjt to seee if it runs better Sir but Its coming to come ur project is working. i await the number thank u

    ReplyDelete
    Replies
    1. Mr Alex, the number is TIP142, a Darlington is specifically required otherwise the transistors will not trigger from an ICs low current outputs

      Delete
    2. thank u Mr Sagatam I will try these. So u say these should give me less heat than the fets Sir. one other question do u think the frequency i am using could be a problem I am using 50KHZ so the fets are switching at 25KHZ and they are irfp260n. the same inverter being used at square wave runs perfect with no heating. please help us to complete ur project Sir. thank u

      Delete
    3. Yes, the high frequency could be the problem, you can try using very low resistance at the gates (10 ohms) and reverse diodes parallel to the gate resistors, with a BJT you will get no such issues anyway.

      Delete
  17. Thank u Mr Swagatam. I specifically told u I was using IRF260 because i wanted u to tell me if the rds on this fet was ok to be used in this sine wave inverter being designed to handle 1500 watts. Could u explain to me which fet has a lower rds one which says .04ohm or one which says 8 megohms

    ReplyDelete
    Replies
    1. Mr.Alex, it's not "megohms", it's "milliohms" (1/1000 ohms) always, so probably now you can research and identify yourself which has the lowest milliohm rating.

      Delete
  18. oh thank u Sir I had it wrong all along. I have a few questions for u Sir . At rest about how much volts should I be getting at the gates of the fets that are driving the transformer I am refering to the pushpull I am building. Next question I always hear that it needs about 10 v to turn on a fet fully ButI am wondering If I am using say a Sg3524 to drive a fet its output is normally about 5 or 6v how would I achieve any volt over 6v from this Sg3524 ic Sir to drive the fets to fully turn on . Please explain this to us . Thank u Sir

    ReplyDelete
    Replies
    1. Who said the output of SG3524 is 6V, all ICs will produce peak DC at their outputs that may be exactly equal to the supply voltage across their supply pins.

      6V may be the average value which is normally 50% of the peak value because the meter would always show the average value when the IC outputs are oscillating.

      The same is true for all such topologies including IC 4017, IC 4047, TL 494 etc.

      Delete
  19. Thank u Mr Swagatam I have learnt so much from ur post SIr now I know how to select the fets for my inverter I sent a post before this about gate voltage if I dont see it i will resend sir thank u

    ReplyDelete
  20. Thank u for answering the question on the sg voltage sir. So as I was askking at rest with no load on the inverter about how much volts should we try to get at the gates of the fets before load is applied and I am refering to the design we are working with using the diodes at the base of the fets I await ur reply Thank u

    ReplyDelete
    Replies
    1. Without the diodes connected, the DC voltage at the gates would be 1/2 of the supply voltage.

      with the PWM diodes connected it could be anywhere from 10% to 40% of the supply voltage

      Delete
  21. Thank u Sir I am trying to work out this problem in stages to see where the problem could be I am still figuering out to myself that it seems sine wave inverters run with more heat than square wave inverters. The example is if I use this same inverter as square wave and power a solder iron there is no heat from the fets but as soon as I induce pwm and use the same solder iron as load it starts to run the fets hot the only difference is inducing pwm into the gates Sir am I missing something here Sir

    ReplyDelete
    Replies
    1. Alex, with PWMs there should not be much heating, the heating will come into effect when a sinusoidal input is employed and the fets forced to respond to sine inputs..
      Please read the suggestion provided at the end of the following article page which gives the possible remedies for PWM topologies:

      http://homemadecircuitsandschematics.blogspot.in/2013/08/homemade-100va-to-1000va-grid-tie.html

      Delete
  22. Thank u Sir. I did some test today and what I found out is if I power the load without the filter I have minimal heating. I have two ferite coils on each leg of the Ac out from the transformer and A 2uf cap accross the output toward the load side.I am having real trouble and cant get . Could u send me ur email so I can send u the circuit and u can simulate it and see if there is a problem I would really appreciate ur help Sir. Thank u

    ReplyDelete
    Replies
    1. Alex, I don't use simulators, moreover checking practically is always recommended, my my simulation won't give you exact solutions, so troubleshooting practically would be much better.

      Delete
  23. I read the post Mr Swagatam but I am still not sure how to deal with mine because they are not using a filter in their design as I am using Sir this is so discouriging Sir

    ReplyDelete
    Replies
    1. Not sure how to deal with your problem, going by the rules is only option we must follow

      Delete
  24. could u give me an email so i can send u the circuit Sir

    ReplyDelete

Readers are requested not to include external links while commenting. For consulting a diagram, upload it on Google Drive and provide the link here.


Readers are advised to proceed with the construction of the presented circuits only after understanding the concepts from the core. Not adhering to this can lead to failures and frustrations.