# DIY rotary inverter technical and economical feasibility

I’m setting up a solar panel system (200W) that’s used to charge batteries that will power small appliances in a converted van. Solid state inverters are expensive and difficult to fix and it seems something like a rotary inverter (generator) would be neither of those things. Perhaps I’m wrong, but that’s what I’m here to find out.

The requirements for the generator are that it can power sustained low power loads (<500W) and surges (900W) at ~120V 60Hz pure sine wave.

A soft requirement is the ability to change it’s output depending on the load. From what I’ve seen of solid-state inverters, they get very inefficient the lower you are from its rated load. The rotary inverter is a solution I’m exploring that solves that.

The reason I'm asking this question is because it seems BLDC motors can achieve high efficiencies (>90%), and if BLDC generators can do the same it seems like it may be worth it.

If they can, what implementation of this do you think is the best balance between cost and power efficiency assuming all new parts? Would permitting used parts change your answer?

• The reason solid-state inverters are used nowadays is because of low cost and size. You can get 200 watt inverter the size of a coffee cup for $25. You will not find anything like that from a generator, let alone the generator will be a hundred pounds. – MadHatter Jan 22 at 15:35 • Are you aware that a BLDC motor is just an AC motor with a solid-state inverter attached to it? – Hearth Jan 22 at 15:35 • Several companies I've worked at have used rotary converters where there was a need for pure sine power, usually for testing at frequencies other than the available line frequency, but do you really need pure sine? The pseudo sine wave you get from most cheap inverters is fine for the majority of loads. lifewire.com/modified-sine-wave-inverter-damage-question-534760 – Phil G Jan 22 at 15:44 • @PhilG you know I read that article, and for every article I read that says it’s not an issue there’s somewhere else that says it is. So confusing. At this point, I’m just trying to gauge the feasability of this option. If what I have to gain is the know-how and a more straight-forward system, I’m willing to put in the time. – shoe Jan 22 at 16:12 • The only advantage of a rotary inverter is the fixed frequency ratio but the inertia which depends on mass and energy to get to full RPM must be huge. It is identical to charging a supercap in kilo farads which would be >1000x bigger and more expensive than a good battery. e.g. one Li Ion 18650 cell is 10kF – Sunnyskyguy EE75 Jan 22 at 16:26 ## 1 Answer Chinese made grid tie inverters are currently being marketed at a price of roughly 1kW/100$USD It might be hard to find a only 1kW/1HP BLDC motor for this price, but if you can, then go for the motor generator converter. An AC generator would then need to acquired also. I don't know if the price get's better or worse with scaling (like a 5000kW rotary or a 1MW rotary) but I do know that at power generation facilities they typically use solid state conversion to convert from AC to DC (and the diodes are huge!)

You'll want to check out this paper Utilization of DC motor-AC generator system to convert the solar direct current into 220v alternating current, which describes an DC to AC rotary generation system and the efficiencies. They also show that it's cheaper and more efficient. However, the efficiency curve that they used might be outdated as I've found one for a solid state inverter that is much higher than they show:

I'd imagine that the efficiency for the plot above would scale to a 1kW system well. In the paper they describe losses from internal heating as being large.

A soft requirement is the ability to change it’s output depending on the load. From what I’ve seen of solid-state inverters, they get very inefficient the lower you are from its rated load. The rotary inverter is a solution I’m exploring that solves that.

One thing to note about efficiency is it doesn't matter what the system is, the efficiency will drop as the power goes down. Why? because any system has a controller or circuitry that uses quiescent current. Let's say the controller takes 2W to run (you'll have to have one for both systems) for a 1kW system. If the system is converting a full load or 1kW then 2W would be 0.2% of the power to run the system. If your converting 100W or 10% of the full load the system can handle then it's 1% and finally if your converting 10W it would be 10% so in each case the efficiency will go down as the rated load goes down (as you see in the plots above).

If they can, what implementation of this do you think is the best balance between cost and power efficiency assuming all new parts?

I think a solid state grid tie inverter would be much cheaper in all cases. The cost of two motors and a controller would be much more than a solid state grid tie inverter.

I don't think so but if you had some parts on hand you might be able to build one cheaper. Even searching around on ebay just one 1-1.5 HP AC motor cost about 100$to 150%, then you'd have to buy a controller and a DC motor. You can get a 1kW grid tie inverter for that much that is more efficient. Even at 500W a solid state grid tie inverter can be had for 60$ that would cover the cost of only a used 1/2 HP AC motor on ebay currently.

The other caveat is even with the motor generator converter with solar energy you would still need an MPPT tracker to get the best power efficiency of the system.

• Another thing to note about efficiency is that if you're willing to spend the , you can easily extend the efficiency range of any switching converters by making two or more that run in parallel -- one wimpy little one that's really efficient at low loads, and one honkin' big one that's really efficient at high loads. Then only bring the honkin' big one on line when it's called for. And, it won't add tremendously to the cost, because the wimpy one won't cost much or take up much room. – TimWescott Feb 5 at 22:26
• A rotary converter is not a motor tied to a generator as you seem to think. It combines both units within a single set of coils creating one that is much more efficient than the combination. It is probably better to think of it as a descendant of the induction regulator. – Edgar Brown Feb 8 at 23:29