Understanding & Using Solar DC-AC Inverters – Part 2

Output regulation

We take for granted the fact that our mains power is very well regulated . so

you can plug almost any appliance into a standard point outlet, and it will

operate correctly. That.s because the electricity supplier has enormous generating

plants, with automatic regulation systems to keep the mains voltage and

frequency very close to constant, despite load variations of many megawatts.

Inevitably you can.t get this kind of performance from a much smaller electronic

inverter, connected to a modest battery or solar panel as the energy source.

However most modern inverters can provide reasonably good regulation for loads

of up to their rated capacity (given in watts) . assuming of course that they.re

running from a well-charged battery.

In this type of inverter it isn.t feasible to control the peak-to-peak output,

because this is largely fixed by the battery voltage and the transformer.s step-up

ratio. So in most cases the regulation is achieved in a different way: by varying

the width of the rectangular pulses, to control the .form factor. and hence the

RMS value of the output voltage.

This is called pulse width modulation (PWM), and is usually done by having a

feedback system which senses the inverter.s output voltage (or load current).

When this feedback senses that the load on the inverter.s output has increased,

the inverter.s control circuitry acts to increase the width of the pulses which turn

on the MOSFETs. So the MOSFETs turn on for longer each half-cycle,

automatically correcting the RMS value of the output to compensate for any

droop in peak-to-peak output.

The resulting regulation is usually capable of keeping the RMS value close to

constant, for loads up to the inverter.s full rated output power. However this

approach does have limitations, mainly because it can generally only increase the pulse width to a certain point. (In the extreme, the output becomes a square wave.) This may not be sufficient to allow the inverter to deliver enough RMS output voltage

in short-term overload or .surge. conditions. When many types of appliance are first turned on, for example, they draw a .startup. current which is many times greater than the current drawn when they.re running. This type of surge can overload the inverter, and its protection circuitry may .shut it down. To prevent damage to the transformer and MOSFETs.

Some types of inverter incorporate special .soft start. circuitry, to allow the inverter to cope with this type of short load current surge. The output voltage and power may drop, but at least the inverter keeps operating and allows the appliance to start up.

Even so, there are some appliances and tools that are simply not compatible with inverters, because of their tendency to draw an extremely high startup current.

Examples are refrigerators, freezers, air conditioners or any other appliance where a motor is driving a compressor or pump. As the motor in these appliances often has a very heavy load right at switch-on (with the compressor near .top dead centre.), it can need to draw a huge current simply in order to start rotating.

This type of appliance and tool should really be powered using a suitably rated engine-driven alternator, not a DC-AC inverter.

Voltage spikes

Another complication of the fairly high harmonic content in the output of

.modified sinewave. inverters is that appliances and tools with a fairly inductive

load impedance can develop fairly high voltage spikes due to inductive .back

EMF.. These spikes can be transformed back into the primary of the inverter.s

transformer, where they have the potential to damage the MOSFETs and their

driving circuitry.

The risk of damage is fairly small during the actual power pulses of each cycle,

because at these times one end of the primary is effectively earthed. Transformer

action thus prevents the .other. end from rising higher than about twice the

battery voltage.

However as you can see from Fig.2, there are times during every

cycle of operation when neither of the switching MOSFETs is conducting: the

.flats. between the rectangular pulses. It.s at these times that the spikes can

produce excessive voltage across the MOSFETs, and potentially cause damage.

It.s for this reason that many inverters have a pair of high-power zener diodes

connected across the MOSFETs, as shown in Fig.1.

The zeners conduct heavily as soon as the voltage rises excessively, protecting the MOSFETs from damage.

Another approach is to have high-power standard diodes connected from each

end of the primary to a large electrolytic capacitor, which becomes charged up to

twice the battery voltage. When the ends of the primary attempt to rise higher

than this voltage, the diodes conduct and allow the capacitor to absorb the spike

energy.

Thanks to this type of protection, most inverters are fairly tolerant of moderately

inductive loads. However they may not be able to cope with heavy loads that are

also strongly inductive . like heavy duty tools and machinery, or more than

one or two fluorescent lights.

Quite apart from the generation of voltage spikes, heavily inductive loads tend

to demand current which is strongly shifted in phase relative to the inverter.s

output voltage pulses. This makes it hard for the inverter to cope, because the

only energy available to the load between the pulses is that stored in the

transformer.