Non sine wave inverter failure

Discussion in 'General Electronics Chat' started by profbuxton, Apr 12, 2014.

  1. profbuxton

    Thread Starter Member

    Feb 21, 2014
    233
    68
    Greetings to you knowledgeable folk,
    I have been given two failed inverters, 24v Dc powered, which have failed in service.
    The load was apparently a refrigerator. One inverter had blown fuses on the dc side and four MOSFETS? shorted and a circuit track to the driver transistors burnt off board.
    I have not yet opened up the other unit but it was used on the same load.
    My query relates to the suitability of non sine wave inverters driving loads such as fridge motors or other such inductive loads.
    Can you enlighten me as to any restrictions on the loads attached to these inverters before the owner buys another one and it fails also.
    I doubt if they are worthwhile repairing but I am interested to find out the failure mode.
     
  2. wayneh

    Expert

    Sep 9, 2010
    12,086
    3,024
    How certain are you that the inverters were not simply too small for that load? The peak load when a fridge compressor fires up, while the door is open with the light on, is quite large. At first blush I'd want an inverter rated at ~2X the continuous load of the fridge.
     
  3. #12

    Expert

    Nov 30, 2010
    16,247
    6,744
    Sorry to disagree, but...The condenser for my air conditioner hums along at 6 amps while the start surge is rated at 76 amps. It only gets 73 amps because of resistance in the power wires, but 2X is a fantasy for a refrigeration compressor.
     
  4. wayneh

    Expert

    Sep 9, 2010
    12,086
    3,024
    Haha, well I meant "at least" 2X but I see that was grossly inadequate. So how does one size an inverter to run a fridge?
     
  5. #12

    Expert

    Nov 30, 2010
    16,247
    6,744
    Personally, I'd read the compressor label for "LRA". That's Locked Rotor Amps, and that's the start surge. You can probably light it up for 90% of its rating if the inverter will self-protect, but most of them won't. They'll just blow their guts out trying.

    Bad label? Try an Amp Clamp meter and add 20% to compensate for the slowness of the needle movement if you don't have a peak hold function.

    Oh yeah, you can intentionally let the compressor get the pressure up, then unplug it, plug it back in, and measure while the rotor is really locked. It won't hurt to do a 2 second measurement with a locked rotor.
     
  6. wayneh

    Expert

    Sep 9, 2010
    12,086
    3,024
    I believe the result of this will be the discovery that you cannot afford the proper inverter. :(
     
    #12 likes this.
  7. #12

    Expert

    Nov 30, 2010
    16,247
    6,744
    Right. It's absurd how over-powered you have to go to get a compressor to start without taxing the limits of the inverter.

    Once upon a time, I did the math for a gasoline powered generator with a 10KW surge rating and a 6000 B.T.U window air conditioner. The windings were good enough, but it would need me to add about 20 pounds of flywheel to deliver enough power for a couple of seconds. Makes me wonder...how can you rate a generator for 10KW surge if the motor stalls?
     
  8. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,515
    1,246
    It would help to know how long they were in service before failing.

    Also, consider that an AC motor of any kind has a power factor way less than unity. Forgetting the huge startup current rush for the moment, this means that even if the motor said 3 amps and the inverter said 4 amps, there is an excellent chance that the inverter will see a load of 5 amps.

    ak
     
  9. #12

    Expert

    Nov 30, 2010
    16,247
    6,744
    During locked rotor, the phase angle approaches zero. There's nothing in there except the resistance of the copper.

    Well...almost nothing.:rolleyes:

    Anyway, the amp clamp will show the actual current, both parts, and the labels show the actual current. No need to run the math on it to get the worst case measurement.
     
    Last edited: Apr 12, 2014
  10. profbuxton

    Thread Starter Member

    Feb 21, 2014
    233
    68
    Thank for the informative replies.
    I suspect you may be correct in the overloading. I currently have opened up the first one to fail and am having a little "probe" (like the aliens do) around. Its a 2300 watt rated unit with a 6000 watt surge capability. The first one I found damaged was only a 1500 watt rating unit.I am always a bit leery of these ratings, too much like audio amps with their 10,000 watt "music power" ratings from the makers.
    I believe this unit worked for some time before failing. I will "interrogate" the owner to find out the fridge rating.
    I guess the questions remains about the advisability of using non sine wave inverters on this sort of load. Does anyone have any experience with these inverters on inductive loads?
    Thanks for your interest.
     
  11. #12

    Expert

    Nov 30, 2010
    16,247
    6,744
    The owner won't know. You're going to have to crawl behind the refrigerator and read the label.

    I don't know why anybody builds or buys those lousy 4 step fake sine wave inverters. They don't have the peak voltage to run a rectifier the way they are designed to work and the same defect means they can not deliver the peak power required to drive an inductive load the way it should work if it had the alleged sine wave voltage. I'm amazed that they will drive anything more complicated than a light bulb.
     
  12. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,515
    1,246
    The watt ratings are accurate as long as you remember that they mean real (as in non-imaginary power) watts. Pure resistive load, zero phase angle between voltage and current. I think the inductive nature of your load is a clue to the failures.

    ak
     
Loading...