Inverter Efficiency Errors

And?

A DMM won't give correct AC voltage reading on a square wave, but I expect you know that....
Not sure if you are asking or telling...
Still it's more interesting than some posts
Del

We are telling, just sharing some information on the subject should anyone else get puzzling results.

The test was done on various inverters, including pure sine wave inverters. Getting one such set of results seemed too good to be true.

There were a lot of things we didn't know about these inverters though. The most important thing we discovered is that when they go into alarm mode (beeping sound), they draw maximum current from the batteries, in this case 2.6 Amps. A particular brand would also draw over 2 amps in when first turned on. But instead of dropping back down to meet the load it was carrying, a 1.4 Amp load, it remained at 2.2 amps with no alarm indication. It had to be cycled on and off a few times before it went down to 1.4 Amps.

So for smaller systems where you are not getting anywhere near the performance from your batteries, the inverter could be the culprit. For systems that don't have a low voltage cutoff, some of these inverters will rapidly drain your batteries beyond 80% DOD.

Also when more than 1 inverter is used, the current draw from each inverter when carrying no load has to be factored in. These small unit draw 0.110 A while larger ones draw up to 5 amps when no load attached. This information was stated anywhere by the manufacturers.

We will be looking at dedicated 'solar' inverters next month to see how they operate.

Greenja

The waveform shown is from a simple "modified" sine wave inverter. NO ONE should ever trust power and efficiency calculations based solely on DMM measurements. Precise voltage and current waveform measurement is the only reliable method. The 85% efficiency calculation seems realistic.

For reference, I regularly see 350W inverters for $25. Just last week I passed up a 2000W inverter that was on sale for $120 (I just don't need one right now).

Regarding your proposed 2000 Watt Sine wave (true Sine?) inverter design:

All electronic parts (including power and protection components), fans, PCB's, chassis, terminals, wiring, and magnetics for $30 ?

Having worked in the electronics design and manufacturing industry for ~20 years, I believe your $30 parts cost estimate for a 2kW inverter would require large volume parts pricing. Fine for a company, but extremely difficult for the average electronics hobbyist.

If you can design, manufacture, and sell your 2kW inverter at a profit, you should do it!

I got a real eye opener while in the Caribbean. The fans, terminals, plugs are so inexpensive I had to check to make sure they were actually made out of plastic and not sugar!

Apparently there is so much over capacity right now and the fact that the middleman is pretty much eliminated, you get the parts for the manufacturers asking price. As far as approvals UL, CSA they all have both. Component cost and IC are very cheap. I'm assuming it has more to do with shipping cost than anything else, the shipping cost has to be passed on to the consumer. Also we found out that these manufacturers produce equipment for several major names, all on the same assembly line so the components are coming from the same place.

We are cautious and do a lot testing before we recommend or install the equipment at sites. Our primary concern is how well the IC do during electrical storms, lots of zapped equipment out here which is not entirely the fault of the equipment, but the lightning protection of the buildings.

Greenja

Hi Greenja,

the inverter efficiencies are well known, as are the mis-reading of "modified sine" waves with non TRMS meters. I get a couple of calls a year from new yacht crew who read the meters on the panel and see typically 185V in place of the expected 220V. I, in my honesty and lack of business sense, take along a TRMS meter, demonstrate the real RMS value and then try to sell them a sine wave unit!

These modified square (I prefer that to modified sine) waves normally have a peak of about 1.414 x the rms therefore I would have expected about 160V for a 115V inverter. However as the battery volts drop with discharge the peak volts drop and the pulse widens to compensate to maintain the rms 115V. They normally cut off when they can no longer maintain the 115 or 230V which is when the waveform has deteriorated to a square wave.

These inverters will run almost anything however when the peak voltage drops some appliances will not run (microwaves being a typical example). This computer and its peripherals, the sat TV, our low energy compact fluorescents all run flawlessly with a mod'd square. The water pump runs a little noisily and normal inductive ballast fluorescent lights usually will not run or stop when the peak starts to get low.

Regards

Chas

Hello Chas,

There were no problems with the mod square wave inverters in terms what they could power. We were a little disappointed when we discovered their shortcomings though, that is excessive current drain when no alarm was present. The only way to know if there was a problem was to meter it. I suppose an analog amp meter in the setup would be a good idea.

We did notice the extra noise coming from the fans. It appears that the fans are running faster than when connected to the AC line. Not sure if the extra noise is from increase fan speed or due to the mod square wave input. We will have to do an RPM measurement on them to find out, then run them at the equivalent AC voltage that produces the same RPM as the inverter.

Greenja.

Hi Greenja,

the noise you are hearing is probably the effect of the high frequencies present in the waveform. Effectively you are trying to jerk the rotor around instead of pulling it smoothly.

The high power draw you sometimes see is most likely the result of a faulty unit. I have never encountered this in twenty years working with inverters, however I am normally only working with the established big name brands.

One problem with inverters is they are a little like a jet engine. If you don't get it right first time they have a high predisposition to blow up! One problem with most brands is repair-ability. Most manufacturer's won't give out schematics, for the reasons above, and as surface mount and programmed chips are used it makes them almost unservicable except at the factory. That is maybe the only reason for building your own.

Patagoniakid's remarks are no longer true. Older sine wave units made square waves and then filtered them to sines with big inductors. This made them very inefficient and noisy. Modern sine techniques have a slightly lower efficiency than the mod'd square but the motors work more efficiently with it.

kind regards

Chas

You are right about not being able to repair the units easily. Some of them sand down the part numbers on the IC, MosFets and Diodes.

Measure the current draw when your inverter goes into alarm mode, if it does not have a low voltage cutoff. Our results came when the batteries were just at the threshold, 10.5V. Then the inverter would go into alarm, draw high current and then just completely drain the batteries as it pulsed in and out of alarm mode.

The part about the old sine wave units making square waves still applies to the modern sine wave inverters. The older ones made a large square wave(mod square wave) and filtered it to a sine wave. The modern inverters make a lot very small square waves (PWM high frequency square waves) and filter them into a sine wave (low pass filter).

We are testing a few designs of single phase inverters using Maxim Mosfet drivers, MC908 micro with Maxim battery charging ICs. So far most of the problems encountered by the smaller units can be eliminated through code.

Greenja.

I think your testing has identified the problem. Some "cheaper" modified sine inverters either completely lack or have poorly adjusted overload and low battery alarm/cutoff features. These safety/protection features NEED to be present and working correctly for all applications.

I would expect/want a low battery alarm warning to start sounding at about 11.5 V and a complete shut-down to occur when the battery voltage dropped below10.8V. IMHO, a well designed inverter would allow the user to adjust these thresholds to slightly higher levels to prolong battery life.

Hi Greenja,

I think the unit you have is either faulty or at the bottom end of the market because all the units i deal with would normally cut out at that voltage, specifically to stop the fatal drain on the battery. They should also latch off with a hysteresis reset of >13V so that it will not switch on until the battery has received some charge. The symptoms you describe indicate a very small hysteresis, as it seems it is dropping in and out of operation as the battery volts recover.

Modern sine wave units do work on a pwm principle but it is not like the old low frequency units. They don't filter away the HF components, they integrate the pulses which means the energy is not being wasted.

regards

Chas

Unless something changed in the last day or so you will find modified sine wave or square wave inverters are not the total solution. There are problems with them that can be offset by price depending on the application. They are cheaper but that is where the comparison ends. I will still go on the record not recommending such inverters unless it is for a specific use that is known to tolerate such a wave form and you are willing to pay the price in inefficiency.

Current information (non technical) can be found here. I rest my case. They work fine but not as efficiently for more than 90% of the cases but...

You will find what doesn't work well with modified sine inverters listed here:

http://www.donrowe.com/inverters/inverter_fag.html#modified

http://www.eirbyte.com/inverters.html

http://www.energymatters.com.au/renewable-energy/inverters/

http://www.windsun.com/inverters/inverter_selection.htm

It really comes down to price and what you are willing to accept from the inverter. Recently I have seen pure sine wave inverters really cheap. Check e-Bay and avoid the modified sine wave problems....I rest my case.

Patagoniakid N6EXQ

Hi Patagoniakid,

Firstly, my apologies, I mis-read your first comment. I thought you had written that the sine units were inefficient a compromise, and so my remark was unfounded.

The modern PWM sine wave inverters are a little less efficient than the modified square wave in terms of energy conversion from battery to 230V or 115V so to use for a resistive or rectifier load (smps power supplies) they are actually still better.

Apart from the items mentioned previously the other main failing of these units is that on some domestic equipment (eg washing machines) with program functions the power supplies for the electronics are derived from a simple capacitive dropper /resistor/ zener circuit which is designed to work with a sine wave of a fixed frequency. The HF present in these waveforms passes too much through the cap and it burns out the resistor or zener after a short time.

I run my house from a modified square unit but I only sell sine units.

Entonces tengo la firma al fondo de los comentarios míos. ¿No sé si les usan lo mismo frase en Patagonia como en España?

Saludos

Chas