Connect with us

Homemade DC to AC inverter questions

Discussion in 'Electronic Basics' started by HC, Feb 29, 2008.

Scroll to continue with content
  1. HC

    HC Guest

    Hey, all, I want to build a homemade inverter to convert a DC source
    into AC. Ultimately I would like to try to build a grid-tie inverter
    (240 VAC in the US), but in the meantime I hope to get a better
    understanding of AC and just simply build a small unit (where I won't
    care what the input voltage is or the output voltage is; probably 6-12
    VDC input, and similar output but as AC) with that purpose in mind.

    I'm sorry this is so long; it covers an area I do not yet understand
    well and I want to explain what I don't know (as well as I can) and
    what I've tried.

    I'm frustrated at my lack of understanding of the basics of AC and I
    am trying to gain that understanding through hands-on work (reading
    books on electronics (Understanding Basic Electronics is finished and
    I'm about 1/4 through Basic Electronics Theory) hasn't answered my
    questions).. I get that AC switches directions, but I built an H-
    bridge to switch a 6 VDC source and it has no negative signal portion
    (according to my oscilloscope) but when I hook up the 'scope to a 12
    VAC wall-wart it DOES have negative signal portions; both have two
    wires, both are supposed to be alternating but one is zero-line and
    above, one has negative and positive signal portions. There's gotta
    be something basic I'm not grasping.

    So, in an attempt to learn this and to build the inverter I ultimately
    want I have built this (Thanks to Tony van Roon for his site): (without the
    transformer; I just wanted to see the signal output) and it gives me
    AC with negative signal portions. Since it runs the output through a
    capacitor, would it be correct to think that ALL the current that
    flows out of this device will have to pass "through" the capacitor (I
    mean, it wouldn't really go through it but it would run into it then
    out of it cyclically, and the current output would seem to be limited
    by it)? If that is correct, would that mean that the maximum current
    output of this type of inverter would be limited by three items: VDC
    supply, TIP41 (or whichever power transistor is used) AND the 2700 uF
    capacitor? The part there that is a big question for me is the
    possible current limiting of the capacitor (I can figure out what the
    current capacity of my supply and the transistor are, that's easy;
    I've never seen a capacitor marked for it's current carrying
    capability). I mean, ultimately, I will need to know how much current
    this setup can produce.

    I tried the H-bridge, but while I get a pulsed output, it's all zero-
    line and above, never a negative output. I thought about using an H-
    bridge with 4 power transistors and feeding one side a sine-wave maybe
    from a function generator, and the other side a similar signal but 180
    degrees out of phase. I would think that would swing the voltage
    across the bridge in a wave similar to the wave of the function
    generator. But, since a square-wave H-bridge didn't yield a negative
    portion of the signal I'm pretty sure that doing something like making
    it swing slowly from one polarity to another won't do it, either.

    I found a schematic for a dual-voltage supply on that site: which might be useful
    for my purposes; to maybe use the zero volt output as the reference,
    and then try to swing an output from +12VDC to -12VDC. I could feed
    the input transformer the square-wave "pulsed" DC I can generate with
    an H-bridge. I would still need a way to gently switch the output
    from + to - but maybe I could put a totem pole of complementary power
    transistors on it and feed it a sine wave (the "unified" base). That
    would give me the gentle swing from from + to - (maybe), and the 0
    volt line would be the "neutral". Ultimately I could feed that output
    into a transformer to step it up to the line voltage I'll finally go
    for. In theory, I could parallel several voltage regulators to
    increase current capacity, and several power transistors to get the
    power output I want (5-10 amps). I don't have a center-tapped
    transformer (yet) to play with so I can't try this (yet).

    Thanks for your help.

  2. Bob

    Bob Guest

    Voltage is relative.
    Voltage is the electric field between two points.

    An H bridge has two output connections. Whent he H bridge switch the
    voltage between the two output connections changes polarity.
    If you consider one direction to be positive then the other direction
    is negative.

    If you measure the voltage of one of the H-bridge output relative to
    the zero volts connection of the h-brige then the result is zero volts
    or six volts.
    If you measure the output relative to the h-bridge positive supply
    the h-bridge outputs are either zero volts or minus six volts.

    You may be getting confused because the zero volts connection of your
    scope is connected to mains earth and the zero volt output of your
    power supply
    is also connect to earth. If this is the case when you connect the
    of your scope lead to an h-bridge output you are shorting it out
    through the earth connection.
    Measure with a multimeter across your h-bridge output.
    You will see the voltage reverse when the h-bridge switch s(assuming
    you havnt blown the transistors by shorting the output with your
    scope ground).

  3. Bob Monsen

    Bob Monsen Guest

    AC is simply a current that occasionally changes directions. So, imagine
    your "H-Bridge":

    V1 ---- RRRRR ------ V2

    When V1 is 0, V2 is 6V. So, the current flows right to left. However, when
    V1 is 6V, and V2 is 0V, the current flows left to right.

    As a consequence, the load RRRRR is seeing AC.

    Your oscilloscope is measuring voltage, relative to some ground. However,
    since I = V/R, then the current is proportional to the V across the

    So, if your oscilloscope has a 'subtract channel' feature, put one probe at
    V1, and one at V2, then subtract the channels. You'll then the voltage that
    the resistor sees across it. That voltage will first be positive, then
    negative, then positive, etc...
  4. JeffM

    JeffM Guest

    If you think the utility is going to allow you to connect to their
    using uncertified gear, you are deluded.
  5. HC

    HC Guest

    Hey, Bob, thank you for your reply. What I had was a control circuit
    (555 timer and 4013) driving the gates on the MOSFET's, running them
    on 5 VDC. The "load" across the H-bridge was driven by a wall-wart
    that put out 6 VDC. I interconnected the ground from the proto-board
    and the wall-wart, so there was a common ground. I connected the
    probe of my oscilloscope to one totem-pole pair output and the ground
    lead to the other totem-pole pair output. What I got was a square-
    wave signal that was always zero or above and flickered back and forth
    (time, left to right) on the display. If I'm understanding what
    you're saying about the voltage being relative, what I needed was to
    have a voltage point that was halfway between the +6 and ground to
    attach the oscilloscope ground lead to. That would be a reference
    point that was 3 volts less than the +6 and 3 volts above the ground.
    Further, if I'm getting any of this right, then it's always only a
    matter of reference whether a signal has a positive and negative
    component or is always in the 0 to positive or 0 to negative range.

    If that's right, how could the oscilloscope show me a sine wave with
    both positive and negative components off the 12 VAC wall-wart when
    there are only two wires, which, if I understand AC transformers,
    should only be one piece of wire wrapped many times around a common
    iron-core (common, in this case, to the winding from the supply
    voltage of 120 VAC)? There would not seem to be a "reference" line,
    yet the oscilloscope did show me positive and negative portions of the

    I feel like I have a brick for a brain. Thanks for your help.

  6. HC

    HC Guest

    Thank you for the reply, Bob. I will check it to see if it does (I've
    only been using the most basic of its features, so I'm not sure; I'll
    check the docs). For clarity, and maybe this is a problem, I have
    been checking the voltage without a load in place, just wanting to the
    wave form.

    What you've said helps, thanks again.

  7. HC

    HC Guest

    First, we're in a rural area and it's a community of "good ol' boys";
    I graduated our local HS with some of the people who now work for my
    local utility; rules get very gray when you know, and have history
    with, the people who enforce them. Second, I care to try to do things
    right and if I ever did actually hook something up that I built I
    would take great pains to ensure it was right first and had a boat-
    load of safety features and auto-disconnects to prevent islanding (big
    worry) or noise and interference. And finally, I'm going to do it
    regardless of what they allow, if I think I've got a functioning
    device that I believe is safe, just say I did it, just to learn how to
    do it, and then maybe I'll go from there to see about getting the
    thing (or things) certified.

    I have to start some place; today I have an idea and I'm exploring
    it. I'll worry about which GPS nav system I'll need to sail the
    Titanic AFTER I raise it, ya'mean?


  8. The biggest safety issue for the utility company is that your device will
    be a dynamic load that can and will generate voltage and reverse current
    flow, so they need to be aware of that if they need to service the lines.
    Also, if there is a fault in the distribution system, and a protective
    device opens, your system will try to power the entire load, so you will
    need a way to sense overcurrents and disconnect quickly. There are also
    lots of transients and momentary brownouts that could wreak havoc on a
    solid state inverter.

    A safer way to do this might be to connect a suitably sized motor to the
    incoming service, and connect its shaft to another motor which you can
    control with your system. If you advance the speed of your motor just over
    its normal speed, it will make the motor on the utility side into a
    generator. This gives you safe isolation from the utility power, and if
    their power is disconnected, your system will no longer generate power into
    the grid (if you are using an ordinary induction motor).

    This should work fine if you have 3 phase service. For single phase, a PSC
    motor might be OK, but you might need to remove the capacitor, at least
    when you start generating. But maybe not.

  9. HC

    HC Guest

    Hey, Paul. I really do care to get it right and to not hurt anybody.
    I will better understand the precautions that will be needed when I
    understand better what the device does and how it does it (i.e., when
    I've built it). Even then I would only hook the thing up for a few
    minutes to test it and then I'll work it through the right channels to
    do it by the book. And before I'd test it for "a few minutes" I would
    compare its output to the line voltage to verify phase and voltage and
    waveform and so forth (running it connected to my power supply, but
    not to the utility).

    But before I can do any of that, I need to understand AC a *%#$# lot
    better than I do now. :-( My lack of understanding is quite
    irritating. And I still come back to my posted example of the H-
    bridge, the AC wall-wart, and the readings I took from them both. I
    get AC with positive and negative swings from the AC wall-wart; there
    is no "reference" line, just two lines from the wall-wart. However,
    when I place the oscilloscope across the leads from the H-bridge IN
    PLACE OF the load (there is no load in place) I get a waveform with NO
    negative swing, just a left-right (time) jitter squarewave.

    If, instead of measuring the wall-wart, I had been measuring the line
    voltage I could maybe understand those results; one line would be
    neutral, the other would be AC (one line would be connected to the
    center tap on my transformer from the utility, fed by a single phase
    of power from them (I have verified by talking to the utility company,
    and posts to the great folks in sci.physics, how this works in the
    US), the other to the "phase" which would alternate from 120 V to 120
    V below (+120, -120) from the center tap. But this is 120 VAC that
    has been run through a transformer so I would think of it as a big U
    for a visual; as the line current moves one direction the transformed
    side moves one direction, and then they both switch. But since the
    transformed side has no center tap, how can it have one line be
    "reference" and one be phase?

    Years ago, when I was a kid and absorbing whatever I could get my
    hands on, I learned a little about electricity. Someone told me that
    current flows from positive to negative. Okay, gotcha. They told me
    that electrons move. Okay, gotcha. And I knew electrons were
    negative. So, I always wondered, since current flows from positive to
    negative, and since electrons move, not protons; did that mean that
    the positive terminal of the battery is actually where the electrons
    come from? Much frustration later, and too many animated discussions
    with people to mention, I finally found something in a book somewhere
    that made it all come into focus. A ridiculously simple distinction
    that nobody had thought significant enough to mention to me; there is
    conventional current flow and there is electron flow and they are
    opposite. Pow. Yes, the electrons move, but before anybody knew what
    an electron was they thought current flowed from the positively
    charged areas to the negatively charged areas, so they considered
    current to come from the positive. Only later, when science advanced,
    did we learn that, in fact, the electrons move, so the real current
    comes from the negatively charged area. So, we have two ways of
    viewing what happens; conventional current flow and electron flow. I
    think we all take that as self-evident now, but when I started to
    learn this stuff that jammed me up; nobody told me that so I was left
    postulating about silly stuff like maybe batteries were labeled + when
    that terminal really gave off the - electrons.

    My point is that I think with the AC stuff there is something really
    basic that I do not know yet. I expect I'll "get" this sometime and
    look back on this and think, "boy, what a doofus I was". But right
    now I don't know what the "Pow" is. I take a reading from a device
    that is supposed to reverse the polarity of two wires, from positive
    to ground, and get a 0-6V squarewave. I take a reading from a wall-
    wart that puts AC out on only two wires (so I would think it just
    reverses the polarity there, too, since there is really only one wire
    that runs to the transformer, around the core, then back) and I get a
    nice sinewave.

    Tomorrow I will set my H-bridge back up and do some more testing.
    Tonight I'm going to keep reading online to see what I can find.

    Thanks for everyone's time.

  10. Hello there .
    Last year i been building an inverter 50 Hz 230 volt .
    Although i didn't use the 555 it was a so called modified sinus
    inverter .
    As squarewave generator i used the sg3525 ,cheap 50n06 mosfets &
    normal ringcore 11 to 230 volt transformer (center tapped)
    The schematic looks like this .
    (its not my schema its something i quickly looked up on the web)
    This is the standard push pull stage .
    By switching one mosfet you make the current flow up after a short
    break the next mosfet makes it flow down .
    The end result is that you have an ac squarewave come out of the
    secondairy side .
    Now about your schema .
    Its not a halfbridge ,Halfbridge configuaration has capacitors to
    devide the powersupply in half .
    (what it is i cant tell you exactly) the totempole output they made on
    the 555 connects it to the plus on high output to minus with low
    (minus Vce ofcourse).
    About the capacitor ,yes it does limit your maximum power.
    In halfbridges you use it to prevent staircase saturation by a dc
    offset ,in that case a small capacitor is there to block the dc .
    In this case the capacitor is way to big for that purpose but the way
    its connected makes me think of the quasi resonant powersupply's .
    Sorry that i can't shed more light on the schema with the 555 .
    If you want a homebrew inverter i would like to advise the sg3525 .
    Its cheap easy and easy adjustable to high powers .
    Last year i builded it for 500 watt's but i know its build up to
    several KW's .

    Be carefull and good luck :)
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day