Hello Everyone,
I have some doubts and need some help with the below?

As per CEA(Central Electricy Authority) guidlines ,they state:
"The generating station shall not inject DC current more than 0.5% of full rated output of interconnection point"

We need to make sure that our on-grid inverter is not injecting any DC into the main grid.

Now for a design with a isolation transformer, there would be no DC injection.
For a transformer-less design with a series decoupling capacitor, there would be no DC injection.

For a transformer-less design without a series decoupling capacitor, there would be some DC injection.
My first thought was to put a clamp-on meter and check the DC current or maybe put a DC current probe and check it on an oscilloscope?
Is this the right way?
Are there specialized instruments which measure the DC current injected into the main grid?

Thanks and regards,
AK

 

Edit:
If i take the average of one AC current cycle, Will that give me the DC component?
Ideally it should be zero,but if there is some difference between the positive and negative halfs,the should give me the DC component?

 

Some older meters have the ability to do TRMS AC as AC and AC+DC. Keysight's does both AC and DC measurements. https://literature.cdn.keysight.com/litweb/pdf/34460-90901.pdf?id=2345839@page=119

The offset of your current transducer will have a large impact.

I have no idea of what the max current of your generator is or what 0.5% is. A little digging came up with these: https://www.panucatt.com/Current_Sensor_for_Arduino_p/cs-200a.htm

 

If i take the average of one AC current cycle, Will that give me the DC component?

1. You need to measure average of RMS current of one AC cycle.
2. You should be sure that measured DC component belongs to inverter, but not injected from main grid to inverter.

 

1. You need to measure average of RMS current of one AC cycle.
2. You should be sure that measured DC component belongs to inverter, but not injected from main grid to inverter.

I don´t think so, since RMS of a pure sine wave is not zero, therefore any averaging of that rms will never give you 0Vdc, even though sine wave has no DC component and the result of your operation should yield zero.

 

RMS has to be aware of zero crossings. Line operated meters will usually measure in terms of n*PLC or n*(Power Line Cycles).

Non-RMS meters precision rectify and then take the long term average and multiply by a fudge factor to read RMS of a sine wave.

You can;t really measure the Average of the RMS because they are the same number. Now you can use the Average function of scope with MATH functions. The average will be the DC current.

 

As a first step just to see if there is something to worry about, try this.

Float the scope with an isolation transformer.
Connect the probe directly across the AC line.
Carefully adjust the GND or zero location.
Measure the positive and negative peaks.

They should be the same. 0.5% may be difficult to resolve on a scope screen, but a more serious problem like 5% should be visible. If it is a digital scope and it has peak and peak-to-peak measurement built in, that probably is more accurate than eyeballing it.

ak

 

@ak52 is your question aimed on how to measure the DC current in order to verify your device, or how to measure it in order to correct for it in the feedback loop of your drive circuitry?

 

I don´t think so, since RMS of a pure sine wave is not zero, therefore any averaging of that rms will never give you 0Vdc, even though sine wave has no DC component and the result of your operation should yield zero.

Edit: picture replaced. Thanks to kubeek, post #11

 

That picture is not very correct, RMS of a negative current is still positive, since it is the square root of mean of square of the current, and any scalar squared is positive.
You said RMS of one AC cycle, which is RMS of the whole wave form, therefore 6.73A. If you said average of (RMS of half cycle times the polarity) then you would be correct, but that is a non-trivial task to pull off in a circuit, since where exactly the edge of the half cycle lies will have great impact on the result.

 

That picture is not very correct, RMS of a negative current is still positive, since it is the square root of mean of square of the current, and any scalar squared is positive.
You said RMS of one AC cycle, which is RMS of the whole wave form, therefore 6.73A. If you said average of (RMS of half cycle times the polarity) then you would be correct, but that is a non-trivial task to pull off in a circuit, since where exactly the edge of the half cycle lies will have great impact on the result.

I absolutely agree with your comments about math terminology. Thank you.
We need zero crossing circuit, connected to out of current sensor and counter of
current cycles. It will start ADC for reading series of instant values of current and sending them to memory, then, after counting presetted number N of current cycles, stop ADC.
Then we calculate separately two values of RMS current: I_pos for all positive readings and I_neg for all negative readings.
Possible error in zero crossing detection will decreased N times, where N is number of current cycles in time of measurement .
Use Closed loop Hall Effect Sensor for AC, DC, & Complex Currents, allow us to have
information about AC+DC current in voltage mode.
http://www.tamuracorp.com/uploads/currentsensor/ClosedlloopPDF.pdf
LA 55-P $24.86 SENSOR CURRENT HALL 50A AC/DC
https://www.digikey.com.mx/product-detail/en/lem-usa-inc/LA-55-P/398-1010-ND/409823.

 

Thanks guys for all your replies.

Some older meters have the ability to do TRMS AC as AC and AC+DC. Keysight's does both AC and DC measurements. https://literature.cdn.keysight.com/litweb/pdf/34460-90901.pdf?id=2345839@page=119

The offset of your current transducer will have a large impact.

I have no idea of what the max current of your generator is or what 0.5% is. A little digging came up with these: https://www.panucatt.com/Current_Sensor_for_Arduino_p/cs-200a.htm

We are already having all the sensors for measuring the Current.We are using a 0.1 class current trasnsformer,so accuracy is not compramised.

@ak52 is your question aimed on how to measure the DC current in order to verify your device, or how to measure it in order to correct for it in the feedback loop of your drive circuitry?

Possibly both,First i need to measure the DC current.If it is beyond the stated limit,i will have to adjust/tune my PI control accordingly.

They should be the same. 0.5% may be difficult to resolve on a scope screen, but a more serious problem like 5% should be visible. If it is a digital scope and it has peak and peak-to-peak measurement built in, that probably is more accurate than eyeballing it.

ak

I did look at the pk to pk ,the positive peak and negitive peak are matching corretly,there is a difference of around 50 to 150 mv,but i think that can be neglected?

I did take an average of the instantanious signals.Took about 10 cycles at different load levels i am getting different results.
If the load is less that 15% ,i am getting values between -0.32v to 1.2v.
If loads are above 15%,i am getting values between 0v to 50mv.

 

Edit:
One more point i am simultaniously doing some harmonic analysis here is an attached screen shot of the THD and indivisual harmonics at 50 % of the load.


I understand 1 is the fundamental and 2,4,6 and 3,5,7 etc are even and odd harmonics respectively.
This question may be dumb,what is the zeroth order? Is that the DC is the waveform?

 

Possibly both,First i need to measure the DC current.If it is beyond the stated limit,i will have to adjust/tune my PI control accordingly.

What I meant was measurement on the fly and immediate correction, sort of how a DC servo in audio amplifiers is used.

 

We are already having all the sensors for measuring the Current.We are using a 0.1 class current trasnsformer,so accuracy is not compramised.

But, by definition, a transformer removes DC and isn't that what your trying to measure? n.nn * 0 is always 0.

 

But, by definition, a transformer removes DC and isn't that what your trying to measure? n.nn * 0 is always 0.

Maybe TS need to use DC, AC current sensor, something like this:

 

The accuracy of that sensor is very close to what OP is trying to measure, and as I recall these cost quite a lot of money. I think that a nice and stable current shunt, instrumentation amp and an averager will produce better results.

 

United States Patent 5055773
DC offset is measured by separating the positive and negative half cycles of the AC waveform, obtaining a measure of the energy content of each type of half cycle and subtracting one from the other and accumulating the result. The accumulated result is measured against a threshold which is selected to have a known relationship to the units used to measure the DC offset. Each time the threshold is exceeded, this increments a counter and at the same time resets the accumulator. The contents of the counter can then be multiplied by an appropriate scaling factor. The square root of this result gives a measure of DC offset.
Added:
Robust Completely Isolated Current Sense Circuit with Isolated Power Supply for Solar Photovoltaic Converters.
http://www.analog.com/en/design-cen...circuits-from-the-lab/CN0280.html#rd-overview

 

Hello all,
We are facing the same problem in our three phase grid tie inverter , the average of our ac current waveform is not zero which is around 300mA .Can anyone please suggest any alogrithm or block diagram to eliminate dc current injection into grid.
thank you.