Inverter design stupid

My background, I'm an EE who has worked on power supplies, but that was a very long time ago and I've spent the last decade in software system and control modeling....so I have a little idea of what is going on, but no significant experience.

My question, or idea, probably very stupid, is are there inverters that don't do voltage level conversions? Start off with a 340v battery bank and just do chopping on the voltage to produce a sine wave. I'm sure this is dangerous, but I think you could produce a very cheap (~50-100$) very high power inverter in exchange for having to have a large battery bank. Just a cheap atmel uP, a couple IGBTs and some caps. The industrial power supplies that I used to work on were a few kw, weighed a couple pounds and fit into a lunchbox. Without the magnetics to do the level shifting and subsequent filtering, why couldn't inverters be small, cheap and simple? I'd think it would also be maybe 5% more efficient.

Stupid idea? What am I missing? Anybody have any ideas or thoughts on the matter? I haven't found anything on google.

Comments

  • boB
    boB Solar Expert Posts: 1,030 ✭✭✭✭
    Re: Inverter design stupid

    One kind of inverter that is getting more popular are non-isolated grid tie inverters which are an effort to increase efficiency and reduce weight. They are sort of what you are suggesting, but do have output filters. Most of these use PV as their input source though.. There may actually be some industrial non-isolated inverters with batteries but I'm not hip to any names right off
    .
    So, yes, I think this is doable, but probably not without some magnetics, either for filtering or inductance for boosting and/or bucking.

    boB

    PS, At least with a simple PWM topology, that is....
  • azrc
    azrc Solar Expert Posts: 43
    Re: Inverter design stupid

    Thanks,

    I saw this yesterday and while not a lot of details, but this sounds a lot of what I'm working on.

    http://www.nrel.gov/features/20110729_powerblock.html
  • RCinFLA
    RCinFLA Solar Expert Posts: 1,484 ✭✭✭✭
    Re: Inverter design stupid
    azrc wrote: »
    My background, I'm an EE who has worked on power supplies, but that was a very long time ago and I've spent the last decade in software system and control modeling....so I have a little idea of what is going on, but no significant experience.

    My question, or idea, probably very stupid, is are there inverters that don't do voltage level conversions? Start off with a 340v battery bank and just do chopping on the voltage to produce a sine wave. I'm sure this is dangerous, but I think you could produce a very cheap (~50-100$) very high power inverter in exchange for having to have a large battery bank. Just a cheap atmel uP, a couple IGBTs and some caps. The industrial power supplies that I used to work on were a few kw, weighed a couple pounds and fit into a lunchbox. Without the magnetics to do the level shifting and subsequent filtering, why couldn't inverters be small, cheap and simple? I'd think it would also be maybe 5% more efficient.

    Stupid idea? What am I missing? Anybody have any ideas or thoughts on the matter? I haven't found anything on google.

    Large commerical UPS unit do this but they usually put an output isolation transformer. We had a one where I used to work that had 40 series connected 12v AGM batteries. In the ballpark of 500 vdc.
  • Solar Guppy
    Solar Guppy Solar Expert Posts: 1,989 ✭✭✭
    Re: Inverter design stupid

    It's obvious none of these folks have brought a product to market.
    NREL nerds wrote:
    "They just have to integrate the power block into their product. Until now, the inverter manufacturers have to build every inverter from scratch. They still have to add filters and protections and so forth, but we give them a large part of the inverter that is already integrated and tested."

    It is typical of academics ... if they had just once brought a product to market they would know the DSP based inverters aren't anything new, been around for over 15 years and all the work is in packaging, NEC/UL and FCC compliance.

    I know first hand you can have the converter running in 2 months from schematic to PCB but it takes 18 months to get it to pass UL , FCC Class B in packaging that can be produced afordably. Not a single one of these building block will ever be in a commercial product, its nothing more than a thesis for federal grant funding.
  • System2
    System2 Posts: 6,290 admin
    Re: Inverter design stupid

    What's inside most cheap high frequency modified sine inverters* is just this:

    DC buss => switching voltage converter => 170VDC buss => H-bridge => 120Vrms modified sine output

    Here's one I was messing with where the parts are quite visibly separated (unusual, to say the least, but kinda cute)
    AIMS Power 3000

    Sinewave inverters are pretty similar with the exception that PWM is used; an output lowpass filter may then be used to remove the high frequency components and leave you with clean 60hz sinewave. Ever opened up a Trace SW? There are two softball sized inductors inside as part of this LPF.

    Realistically, this is all you'd need:

    HVDC => IGBT bank being switched with PWM control to generate sinewave => Output LPF



    * available in bubble packs everywhere fine chinese rubbish is sold :p
  • RCinFLA
    RCinFLA Solar Expert Posts: 1,484 ✭✭✭✭
    Re: Inverter design stupid
    SolarGoth wrote: »
    What's inside most cheap high frequency modified sine inverters* is just this:

    DC buss => switching voltage converter => 170VDC buss => H-bridge => 120Vrms modified sine output

    Here's one I was messing with where the parts are quite visibly separated (unusual, to say the least, but kinda cute)
    AIMS Power 3000

    Sinewave inverters are pretty similar with the exception that PWM is used; an output lowpass filter may then be used to remove the high frequency components and leave you with clean 60hz sinewave. Ever opened up a Trace SW? There are two softball sized inductors inside as part of this LPF.

    Realistically, this is all you'd need:

    HVDC => IGBT bank being switched with PWM control to generate sinewave => Output LPF



    * available in bubble packs everywhere fine chinese rubbish is sold :p

    Actually, a modified sinewave inverter's DC-DC converter puts out about 155 vdc. +155v for about 5 msec, 0v for about 3.25 msec, -155v for 5 msec, 0v for 3.25msec = 120 vrms @ 60 Hz.

    True sinewave high freq PWM inverters DC-DC inverter has to get a bit over 170 vdc to make sinewave peak allowing for some drop for H-bridge.
  • Vic
    Vic Solar Expert Posts: 3,208 ✭✭✭✭
    Re: Inverter design stupid

    SolarGoth,

    I've only been inside an X SW+ Invereter once.

    There are three output transformers, one LARGe, one Medium, and the last one small in size. The outputs are hooked in series (IIRC), and this is the AC output of the inverter.

    There must be an LPF function, but was not looking for it.

    The output of the SW+ inverter is fairly clean, but still has some HF switching artifacts associated with each step of the sine wave. Using an oscilliscope on the output shows these quite distinctly.

    RCinFla is quite knowledgeable on these inverters, and prob will correct my recollection of the inverter topology.

    Have Fun, Vic
    Off Grid - Two systems -- 4 SW+ 5548 Inverters, Surrette 4KS25 1280 AH X2@48V, 11.1 KW STC PV, 4X MidNite Classic 150 w/ WBjrs, Beta KID on S-530s, MX-60s, MN Bkrs/Boxes.  25 KVA Polyphase Kubota diesel,  Honda Eu6500isa,  Eu3000is-es, Eu2000,  Eu1000 gensets.  Thanks Wind-Sun for this great Forum.
  • GreenerPower
    GreenerPower Solar Expert Posts: 264 ✭✭✭✭✭✭
    Re: Inverter design stupid

    Vic,

    By control the direction of the current flow into the primaries of those 3 x-formers, the output voltages in series can add/subtract each other. This is an inexpensive way to approximate the sine wave with 27-level stair-step wave form, not quite as true sine wave with PWM but effective enough.

    GP
  • boB
    boB Solar Expert Posts: 1,030 ✭✭✭✭
    Re: Inverter design stupid

    The SW's three MSW inverters in series are added together to make up the stepped output sinewave approximation. Those two softball size toroids are common mode chokes but do offer some filtering when higher output currents are drawn from the inverter. You can see this if you monitor the secondary of each individual transformer with a scope like in the picture.

    Also, the amount of steps vary depending on the dc voltage and load so this picture is only representative of waveforms at one voltage and load.

    boB

    PS. Yes,


    image.php?u=47&dateline=1189381808 RCinFLA is right... It's the leakage inductance that gives some filtering of the sharp edges at higher AC load currents, not the CM chokes. The CM chokes are for conducted emissions reduction (RFI/EMI)...

    boB (again)
  • Vic
    Vic Solar Expert Posts: 3,208 ✭✭✭✭
    Re: Inverter design stupid

    OK,

    Well my recollection of the SW+ was that the CM Chokes were relatively small. About the volume of about 1.5 Golf Balls. There are a couple of Film Caps on the output IIRC, and they do trim the HF components of the outupt.

    Looking at the waveform of a lightly-loaded SW+, the voltage peak is absent, just a flat-top. A bit surprising to me, but guess that is within 5% Harmonic Distortion.

    I do not claim to know just how these inverter/chargers work, but have been happy with them here. Glad that the power room is 200+ feet away from the residence ... BBBUUUZZZZZZZZZzzzzzzzzzzzzzz ... the only complaint.
    Vic
    Off Grid - Two systems -- 4 SW+ 5548 Inverters, Surrette 4KS25 1280 AH X2@48V, 11.1 KW STC PV, 4X MidNite Classic 150 w/ WBjrs, Beta KID on S-530s, MX-60s, MN Bkrs/Boxes.  25 KVA Polyphase Kubota diesel,  Honda Eu6500isa,  Eu3000is-es, Eu2000,  Eu1000 gensets.  Thanks Wind-Sun for this great Forum.
  • RCinFLA
    RCinFLA Solar Expert Posts: 1,484 ✭✭✭✭
    Re: Inverter design stupid

    The SW series output filter is just for RF emission suppression. It does not filter the output stepped sinewave much. You can hear the steps as a low buzz in your ceiling fans.

    http://www.google.com/patents?id=2SgnAAAAEBAJ&printsec=frontcover&dq=5373433&hl=en&ei=7mZNTrDgK6Tq0gGaueCUBw&sa=X&oi=book_result&ct=result&resnum=1&ved=0CCwQ6AEwAA

    Uses principle of low leakage inductance transformer. In simple terms, if primary is short circuited then secondary is short circuited. Actually there is a little inductance and resistance, couple of mH and tenths of ohm. All the up's and down's must be equal in time over the cycle on each transformer or the core will become magnetically biased with a DC field offset which is a bad thing.

    In contrast, the single low frequency transformer in XW series and Outback inverters that is high frequency PWM'd on the primary side is designed with some amount of leakage inductance (higher then a low frequency transformer would normally have). This leakage inductance of low freq transformer becomes the series inductor, in effect, for the output filter to remove the high frequency PWM chopping. With the leakage inductance, only a capacitor needs to be placed across output winding to accomplish the L-C low pass filter.