Question on Micro-Inverters

[sdold]

I could see how a differential amplifier could do this if the other input was derived from the inverter output current. The job of the differential amp would be to source current in proportion to grid voltage. Not saying it works that way, but I see how it could.

[solarix]

The control loop (diff amp circuit) in the inverter is what makes the inverter output match the grid voltage. Actually, there is an infinitesimally small difference between the inverter's output and the grid's voltage and this difference is sensed to control the circuitry. Maybe a better way to think about it is from an Amps perspective. The inverter is constantly having to figure out what current (amps) to push out onto the grid in order to output the momentary power level at whatever voltage the grid happens to be at that instant.
The whole subject can be a career for the engineers involved and is indeed sophisticated so am sorry if I'm not boiling it down very well.

[jflorey2]

A basic circuit in electronics is the difference amplifier (or diff-amp). It is the basis of all control theory. One input is connected to your reference signal (grid AC) and the other input is connected to your inverter output. The output from the diff amp is then proportional to any difference between the two and is used to drive the inverter's power circuitry to minimize this difference. This "negative feedback loop" thus works to make the inverter's output just follow the grid's AC wave.Now, inverters may well have some digital sinewave approximation involved in the process to give it some help or smooth the signal etc., but the diff amp is the key.

That doesn't make any sense. The AC input voltage is one voltage reference you could use. What's the other? (And you can't say "the inverter output" because the inverter output is the same as the AC input voltage. There's no difference; they are the same potential.)

There are indeed a lot of control loops within the inverter. What you have described is not one of them.

[SunEagle]

....

Grid-tie inverters can't operate without a grid to follow and when the grid goes down, the inverter goes down too. POCO's that require "safety" AC disconnects to isolate the (evil) solar inverter from the grid so their lineman can work on powerlines during outages are just implementing stupid OSHA rules in an effort to make their competitor's installation costs as high as possible.

Sounds like you have no idea why solar inverters have to disconnect from the grid for safety issues.

It does not have anything to do with competitor's or costs. It is purely a SAFETY issue for anyone that may be working on the power lines.

Saying anything else is just plain stupid.

[solarix]

OMG. First - realize that micro-inverters can't make AC on their own. They have to have a reference to synchronize to.
and second - they have to have a low impedance (dynamic resistance) load that can absorb however much power the microinverter is generating.
The grid is both of these. All grid-tied inverters are like this - whether micro or string.
A basic circuit in electronics is the difference amplifier (or diff-amp). It is the basis of all control theory. One input is connected to your reference signal (grid AC) and the other input is connected to your inverter output. The output from the diff amp is then proportional to any difference between the two and is used to drive the inverter's power circuitry to minimize this difference. This "negative feedback loop" thus works to make the inverter's output just follow the grid's AC wave.
Now, inverters may well have some digital sinewave approximation involved in the process to give it some help or smooth the signal etc., but the diff amp is the key.

Grid-tie inverters can't operate without a grid to follow and when the grid goes down, the inverter goes down too. POCO's that require "safety" AC disconnects to isolate the (evil) solar inverter from the grid so their lineman can work on powerlines during outages are just implementing stupid OSHA rules in an effort to make their competitor's installation costs as high as possible.

[jflorey2]

Thanks for the explanation, Do most grid-tie inverters do this? I didn't realize so much was going on.

There is indeed a lot going on. Grid tie inverters have to convert DC to AC and send power back to a power grid; that is about the simplest thing they do. They also have to run an MPPT algorithm to maximize power extraction. They must comply to the standards of IEEE 1547 and meet the requirements of the UL1741 test suite; this means they must meet requirements for startup and shutdown timing, rapid disconnect from the grid, power quality etc. Some must also comply with Rule 21, a set of new rules from the Smart Inverter Working Group; these are newer requirements that help with grid support on utility grids with large percentages of solar penetration.

And of course they must do all this even on power systems with poor power quality, higher than expected impedance and high levels of electrical noise without creating much noise themselves.

[sdold]

Thanks for the explanation, Do most grid-tie inverters do this? I didn't realize so much was going on. Edit: So if I'm understanding this correctly, it sounds like it does act like a sine-wave current source that really does try to be "sinusoidal", even if the grid waveform is distorted.

[jflorey2]

What's a Sine Wave Estimator? I googled it and didn't find anything useful.

It's just something that can calculate a sine wave inside an inverter. In its simplest form it is a table that contains the values from a quarter of a sine wave, populated to whatever resolution is needed.

There's a reason for having this sort of functionality. An "ideal" resistive load will take whatever the voltage is at that instant and draw a specific current. One could therefore think that an "ideal" DER inverter would send back current based on voltage; this would be the inverse of the resistive load, and resistive loads are "good" loads for the grid as opposed to peak-rectification or inductive loads.

However, that is both harder for the inverter to do and worse for the grid overall. In cases of peak rectification, for example, the top of the AC waveform gets "flattened out" by poorly designed AC to DC converters that do not have power factor correction. This leads to higher than expected neutral currents in three phase systems, more harmonic energy and higher peak currents.

An inverter that implements the inverse of a resistive load doesn't make this any worse but it also doesn't make it any better. During the flattened peak the inverter simply outputs the same current. An inverter that outputs current in a sine wave helps overcome the problem by supplying more current during the peak, which helps restore the sinusoidal waveform. It's also easier; rather than a feedback loop that keeps current proportional to (often odd) voltage waveforms, the inverter need only determine what the sine wave's phase is (usually a relatively easy task) and then sync the sine current to that phase.

[sdold]

There is actually a sine wave estimator in the inverter.

What's a Sine Wave Estimator? I googled it and didn't find anything useful.

[jflorey2]

I've always wondered this too. Does anyone here really know? I have always suspected that there is nothing in the inverter that generates a sine wave, but that it instead just sources current to the grid based on whatever part of the cycle the grid is on.

There is actually a sine wave estimator in the inverter. That's needed due to requirements for power quality and power factor; the current is required to be fairly sinusoidal (and neither lagging nor leading in most cases) even if the voltage waveform is not.

[Mike90250]

The inverter has a sensor in it that senses the phase of the Grid AC waveform, and synchronizes the inverter to the Grid. A very simple circuit function to slave sync one source to another.

[sdold]

I've always wondered this too. Does anyone here really know? I have always suspected that there is nothing in the inverter that generates a sine wave, but that it instead just sources current to the grid based on whatever part of the cycle the grid is on.

[npaizis]

Got it. That answers my question.

Thanks again.

[jflorey2]

So you're saying the AC line has a signal on it that the micro-inverter can sync up with. Is that correct?

Yes. However, the "signal" is the voltage waveform; there is no separate signal sent to make that happen, just the same AC voltage waveform that any other AC device uses.

[npaizis]

jflorey

Thanks for replying. So you're saying the AC line has a signal on it that the micro-inverter can sync up with. Is that correct?