Modified Sine wave inverter issues
System Posts: 2,511 admin
I have never installed a modified sine wave inverter for any of my systems, but this time I went with the Trace DR series inverter due to money issues. The owners are running a small window AC unit draws about 5 amps, and a small freezer draws about 4 amps, that all for the load side. I know about the issues many have had running things using Modified Wave inverters. Has anyone ever installed an Isolation Transformer say 5 kW unit, between the inverter AC output and the appliances. I might be spinning my wheels but I was thinking about a capcitor inductor network that might help to smooth out the sharp edges on the modified sine wave. Sounds like Arc and Spark engineering but that is one thing I am pretty good at. Just thought I would ask for advice before I let all the smoke out of the appliances. arcandspark
0 · Share on Twitter
It would probably clean it up until you turned on that heavy load, then you would let the smoke out of the cap more than likely
the introduction of the transformer alone would reduce some of the harmonic content, but the capacitor added for resonance would be very good. i would be concerned with the high value needed to make the circuit resonant as it will entail allot of caps for it as you probably already know. also the voltage rating is of concern as well. if you can't do a resonant circuit at least a low pass filter passing some of the higher frequenies to ground would help too. it would be interesting for you to scope this if you have one.
Niel, I have the isolation transformer, 5kW should work just fine. As for the cap yes I can only get to about 2000 UF and 400 vac. Probably wont do much to help. Interesting though the part about the higher harmonics. I have loads of .1 through .001 UF 600 volt caps and I have a couple of scopes so I think I will try it and see using the scopes if it helps. I fugure it won't because if it did work then Trace would be recommending these things to users of Modified Sinewave units to help smooth things out, but I will try it and let you know the outcome. My concern about using the modified wave inverter is causing damage to the AC unit and the freezer compressor. Picked up the Trace DR 2400 watt inverter new for $770.00, not sure if I got a good deal, but I am hoping I did. arcandspark
Hey, I'm NOT suggesting anyone try what I've been up to, nor am I pushing any theory, - - just the facts -
For over a year now, I've been running a 7 cu ft., Woods freezer, converted to a chest fridge, on modified square wave, also, for over two years, my computer, washing machine, a bunch of CF lights, stereo's and sometimes a 1/2 HP, 220 volt submersible pump through an auto transformer. To date, have had no problems. The Chest Fridge only runs about 5 minutes out of every 40 minutes, so the compressor only just starts to get the chill off it before it stops - no chance to overheat. I have however, cooked the CPU power supply in the breadmaker, twice! That happened within 2 minutes. It has no transformer, just a capacitor, which loves to pass those sharp edges, a diode, a fuse-resistor and to limit the voltage, a zeener diode. The capacitor "passes" too much current, the zeener diode overheats badly and finally, the fuse resistor opens, preventing a fire.
When I have time, I plan to build a transformer based CPU power supply with the several different voltages required. In the meantime, the bread is only made using grid power, as I'm running out of fuse resistors LOL.
Oh, and the Chest Fridge? I'm VERY happy with it.
DR series has the isolation transformer in it already. That is what makes it so heavy compared to portable units that do not have a groundable neutral. The transformer helps a little.
As to putting a capacitor in parallel, I would definately not recommend doing that. Beside the resonance issues there will be high surge current at the modified sinewave switching points. If you want to play you can put some small series inductance before the capacitor similar to what is done on high frequency PWM inverters like ProSine. The value and size depends on load and you have to make sure the series coils don't saturate during startup surge current demains.
You did not describe the actual problem but I would be very careful of the starting surge from the window A/C unit. I would not be surprised if it has a 35-40 amp startup surge current that last for half a second or so. A 2KVA DR unit is just barely enough to starup a common refrigerator compressor. Beyond that, most A/C compressor use a continuous split phase capacitor motor design. It is somewhat incorrectly referred to as starting cap and starter winding but it is on all the time. It provide a phase lead for the motor, sort of like a carrot on a fishing pole held in front of a horse. The modified sinewave will cause higher peak currents in this phase lead winding. These winding do not have the wire size of the main run winding so you might run risk of burning it out over time. A slight reduction in this capacitor will help reducing this winding stress but it will lengthen up the actual startup surge time before the compressor spins up to run speed causing the inverter to overload trip before compress drops to it run value current
Check if the compressor has a lock rotor ampereage rating. You will see this value, to slight higher for initial turn on. (LRA typically spec'd at lowest ACvoltage operation spec, therefore it is higher for nominal to high line voltage). My 42K BTU central unit compressor has 104 A LRA on name plate with 205 vac minimum operational voltage. At 240 vac I have measured a wopping 120 A startup surge that last for 300 msec using a high frequency response closed loop hall effect pickup sensor. Run current is 15 amps (at 240vac) My two series stacked SW5548plus's barely gets it started.
yes if those caps are in parallel to the voltage it will pass current to ground so to make the circuit work it must be resonant and would have to be series not parallel. i believe the X sub C(ac resistance) is linear with frequency if i remember right so it won't pass too many harmonics to ground in parallel without the fundamental passing a good deal to ground as well. forget the caps then unless you can come up with enough to resonate it in series.
another possibility would be to have the content of the 2nd and 3rd harmonics passed to ground with series reasonant lc circuits at those frequencies(notch) and those 2 would account for the majority of the harmonic energies. that would be a series resonant lc circuit at 120hz to ground and another one of the same, but at 180hz to ground. personnally with that much inductance and capacitance needed i wouldn't do it.
I played around with that, experimenting with a 1000 Watt modified square (sine) and even with a cap as small as 0.8 mfd, the inverter instantly shut down on overload - every time.
Never did get to play with inductors and capacitors together, although, I intended to play with that too. Not enough hours in the day.
Thats why I enjoy this web site. You can get good feedback from great people. You have all saved me lots of wasted time effort. Sounds like nothing can really help to improve the modified sinewave inverter so I will just have to use it as is. Worse case I can use the inverter just for lights and ceiling fans in their home. Thanks again, arcandspark