Modified Wave Inverter - Page 2
 

Modified Wave Inverter

Started by Tikvah, November 30, 2013, 05:17:06 AM

Previous topic - Next topic

bevans6

If anyone (like me, for example) is curious as to how a pure sine wave inverter works, here is a good explanation at a medium technical level:  

Edit:  couldn't make the link work, sorry...


In even simpler terms, a pure sine wave inverter uses an oscillator to create an absolutely pure sine wave reference waveform at the frequency desired (50hz or 60hz), then the sine wave is sampled at around 50Khz (fifty thousand times per second), that sampled output drives a pulse width modulator that creates an on/off switched signal that mimics the sampled sinewave and controls the output MOSFET's (which are power switches).  The 12V or 24V DC battery input voltage is sent to an oscillator, transformed up to 180 volts (the peak voltage of a 120 volt RMS AC waveform) and turned back into DC to power the output stage.  The MOSFET's create a 360 volt peak to peak square-wave output at 50Khz that is sent to a filter (simple capacitor and the output transformer) and the output is smoothed to a pure sine wave again, but at 60hz and with whatever power capacity you designed the MOSFET stage to.  Output distortion is going to be present, but at around 1% to 3% total harmonic distortion compared to a theoretically pure sine wave, not an issue.  The waveform has smooth zero crossing points, which is what induction motors and microprocessors like to see, and the distortion is usually limited to a little clipping at the peaks.  If you use the output to power an induction motor, the CEMF of the motor will further perfect the waveform to a pure sine wave.

This little exercise is why I love these bus forums - I learn a lot, not only from things I read here but from the places my curiosity takes me from questions I find here.  This morning I had no idea how pure sine inverters actually worked, aside from a vague idea of switching power supples and MOSFET output stages, and now I do.   :o

Brian
1980 MCI MC-5C, 8V-71T from a M-110 self propelled howitzer
Allison MT-647
Tatamagouche, Nova Scotia

luvrbus

DonRowe.com answers a lot of questions about both type inverters on his site I have noticed that over the last few years the Energy Star rated appliances struggle with the modified sine

Keep your eyes open for a older SW Trace with load sharing it was the best inverter ever made
Life is short drink the good wine first

robertglines1

looks like there is no free energy ride here!  sine wave/ pure...   If I get it correct?  if it has  brushes or a toaster= it does not matter which?  if it has  electronic-computer controls=pure sine wave.  Some of the old  units were not as sensitive to sine wave as the new latest and ...   I'm  out of my comfort zone hear..And would like to understand.   Bob
Bob@Judy  98 XLE prevost with 3 slides --Home done---last one! SW INdiana

luvrbus

Now days in the new high end conversions you are seeing more and more of the Denmark and Dutch made inverters my doctor has the WTC or WTG load sharing inverters in his bus where they are made I have no idea

They are nice and look expensive lol but no matter the cost I probably paid for a set or 2 for him over the last couple of months
Life is short drink the good wine first

bevans6

Robert, the good news is that pure sine wave inverters are coming down in price quickly, in the smaller sizes and the consumer quality versions.  I bought a 1,000 watt model for $180 last year.  The high end commercial quality inverters are always going to cost more, simply because of quality and programmability.

It's hard to explain why some things work well on MSW inverters and other things don't work well, because it quickly gets into fairly sophisticated electronic and electrical theory.  Some things are easy - anything that is a plain resistive load, like an old-fashioned incandescent light bulb, a heater like a coffee maker, and so on, will work fine with MSW because the square wave output has the same basic amount of electrical energy as a pure sine wave would have.  Induction motors like on an air conditioner or a home style refrigerator work kinda OK, but have issues.  That type of motor is actually generating electricity as it turns, as well as using electricity - that is called the CEMF, or counter electro-motive force - and it pushes back against the input electricity as the motor turns.  The problem is that the incoming electricity is a square wave and the CEMF is a pure sine wave so they fight each other and the motor uses more power than normal.

Other kinds of devices change the electricity in order to use it, and somethings change the square wave easily while others can't change it well at all.  Something with a transformer on the input can take the square-wave and use it fairly well, but other things use filters, capacitors and other things, and they don't like the square wave at all - examples are small battery chargers and things that use some types of "wall lumps" for power. 

Another problem with the square wave is that it is actually made up of the sum of a sine wave plus all of it's odd harmonics, going up in frequency, so a 60 hz square wave has a big 180 hz component, a smaller 360 hz component, a smaller still 420 hz component and on up till they kind of fade away.  Many things can't deal well with the high frequency components of the signal, so even though it has the same amount of electrical power as the pure sine wave, it causes things (like induction motors and microwave magnetrons) to heat up or operate incorrectly.

The well known problem with clocks is down to the wave form that the square-wave inverter puts out.  Basically, as it goes through it's 60 hz pattern it sits at zero for a while, then it jumps up to around 180 volts, then it jumps down to zero again, then it jumps down to negative 180 volts, then it jumps back to zero and sits there till it starts over.  So there is a fair amount of time spent at plus or minus 180 volts, and a fair amount of time spent at zero volts.  Clocks, and timers of all sorts use the zero crossing point of a sine wave to trigger their clocks - it's easy to detect and very accurate.  The square wave doesn't have a zero crossing point to speak of - it just sits around at zero then jumps up or down - so the clock gets very confused.

There are all sorts of other things that work well, kinda OK, or very poorly with MSW inverters.  For example, some kinds of fluorescent lights have ballasts that can actually cause the inverter to overload, they act like a short circuit in a way.

I hope this helps, it's not particularly technically perfect as an explanation but I tried to make it relevant to a lay-person who probably wouldn't have a degree in electronics...

Brian
1980 MCI MC-5C, 8V-71T from a M-110 self propelled howitzer
Allison MT-647
Tatamagouche, Nova Scotia

edroelle

Brian,


To reinforce your comment about capacitors - Earlier, I had said I burned-up 2 electronic defrost timers in my Whirlpool refrigerator.    The component was a capacitor that burned-up each time, as you commented in your last post.    I considered putting a larger capacitor in, but decided to eliminate the whole electronic board.

Ed Roelle
Flint, MI

Oonrahnjay

Quote from: bevans6 on December 03, 2013, 05:39:52 AM... There are all sorts of other things that work well, kinda OK, or very poorly with MSW inverters. ...

    I don't know why but I bought a (fairly nice and large but not expensive) Panasonic "inverter" microwave.  It works fine on the shore power and on the generator but through the Outback inverter, it acts like it's not happy* and seems to pull a lot of power but doesn't seem to get much of anywhere.  Have other people had similar problems?


(* That's the technical term that those of us who knows lots about them electrictrons use.)

BH  NC   USA
Bruce H; Wallace (near Wilmington) NC
1976 Daimler (British) Double-Decker Bus; 34' long

(New Email -- brucebearnc@ (theGoogle gmail place) .com)

bevans6

Which model Outback do you have?

Brian
1980 MCI MC-5C, 8V-71T from a M-110 self propelled howitzer
Allison MT-647
Tatamagouche, Nova Scotia

Oonrahnjay

Quote from: bevans6 on December 03, 2013, 12:09:23 PMWhich model Outback do you have?  Brian 

    I have an Outback FX 2012 MT. 
Bruce H; Wallace (near Wilmington) NC
1976 Daimler (British) Double-Decker Bus; 34' long

(New Email -- brucebearnc@ (theGoogle gmail place) .com)

bevans6

That inverter should have no issues with any microwave oven, so all I can say is call the manufacturers help lines.  Sorry...

Brian
1980 MCI MC-5C, 8V-71T from a M-110 self propelled howitzer
Allison MT-647
Tatamagouche, Nova Scotia

Oonrahnjay

Quote from: bevans6 on December 03, 2013, 02:23:12 PMThat inverter should have no issues with any microwave oven, so all I can say is call the manufacturers help lines.  Sorry...
Brian 

     Thanks.  I know I have a battery issue -- somebody who should have known better came very close to killing my house batteries (wet cell) so I'll get "real batteries" and try again.   BH
Bruce H; Wallace (near Wilmington) NC
1976 Daimler (British) Double-Decker Bus; 34' long

(New Email -- brucebearnc@ (theGoogle gmail place) .com)

technomadia

FYI - the only current manufacturers of load sharing inverters that I know of are Victron, MasterVolt, and now at last Magnum.

We have a Victron and mostly love it, and I am about to start beta-testing the new Color Control GX control panel for it.

  - Chris
Cherie and Chris / Bus tour: www.technomadia.com/zephyr
Full-time 'Technomads' since 2006 (technology enabled nomads)

Oonrahnjay

Quote from: technomadia on December 03, 2013, 07:31:13 PM
FYI - the only current manufacturers of load sharing inverters that I know of are Victron, MasterVolt, and now at last Magnum.

We have a Victron and mostly love it, and I am about to start beta-testing the new Color Control GX control panel for it.

  - Chris 

      Thanks, Chris.  I am thinking of going over to the Victron.  They have some nice high-capacity inverters.  Is the quality good?
Bruce H; Wallace (near Wilmington) NC
1976 Daimler (British) Double-Decker Bus; 34' long

(New Email -- brucebearnc@ (theGoogle gmail place) .com)