Power Meter ADC problems

[Mike Walne]

Three things:-

a) At this stage, are you using the turbine as your power source or mains?
b) Your formula is the way to go to calculate power.
c) Zero drift from you current sensor is not an issue.

I'll come back to you with a more complete analysis when I have a little time to spare.

Mike

[Can]

[Ttelmah]

Only as an example to others....

Standard 1/4 carbon film resistors, normally have a maximum voltage rating of something like 250v. Now remember what the peak voltage is on 220V ac, and cringe....

Isolate, before (if you are _lucky_), you 'live to regret it'.

Best Wishes

[Mike Walne]

OK. So we understand each other.

When we meet someone new on the forum we don't know who the chancers are.
I don't like answering questions directly.
I'd rather answer your question with a different one.
That way you may be persuaded to work out the answer to the original for yourself.
This is a public forum, your tutor can monitor our conversation as well as the next person.

As I see it you still have a problem with interaction between ADC channels.
In which case you can tackle the problem by disconnecting from the dangerous sources.
I.e. use low voltage benign ones instead.
That was my thinking behind the suggestion to apply DC to one channel and AC to the other.
For this scenario you can use battries, DC power supplies and signal generators.
That way you (and your nearest & dearest) won't have to worry about your imminent demise.
You can monitor what's going on with a 'scope, and stream loads of data to a PC.
Without good raw data, you can't progress.

I was not being sarcastic about Vrms*Irms.
The hint was that you needed to do some research.
I intended to be helpful.

You go to class and they tell you Power = Irms * Vrms
A few weeks later, another guy says:-
We should have said Power = Irms * Vrms * cos(theta)
(Or something similar)
Then they let you out into the real world.......
The voltage isn't sinusoidal and the current isn't even the same shape as the volts.

What you've forgotten (or didn't notice) was, at the beginning of the class, they gave you the integral formula for calculating power.
The emphasis was on getting the nice neat equaton you recorded in your notes, and expected to duplicate in exams.

Sorry but the neat equations are special cases only.
Welcome to the real world.

Mike

[Ttelmah]

Read the data sheet.
Look at the section entitled 'A/D Acquisition Requirements'.
The second paragraph of this explains what is wrong with your code.
Study the examples of the forum here and from CCS. Notice what they all do when changing channels.

Best Wishes

[Can]

Ok, 80 us cost me a really long time. So, now I can read the voltage and the current independently. Now I will be trying to isolate the source from the rest of the circuit and then I will try to arrange the power calculation parts.

When I finalize the code with the power measurement and everything, I will put it here.

[Mike Walne]

Assuming that you've got your ADCs biassed somewhere near centre.

To calculate one ADC RMS you:-

1) Create variables sum_of_samples, sum_of_squares, mean, mean_square, rms, n.
2) Reset n, sum_of_samples and sum_of_squares to zero

3) For as near as possible an integer No of complete cycles, i.e. n samples:-

4) Add each new sample to sum_of_samples.
5) Square each sample and add to sum_of_squares.
6) Repeat for n samples.

Then calculate :-

7) mean = sum_of_samples / n.
8) mean_square = (sum_of_squares - n * mean ^ 2) / n
9) rms = square_root of mean_square.
10) Convert to engineering unit.

You can do the above for Irms, Vrms as shown, then a similar process for power.
Do all the maths as integer maths except for the final ones and conversion.

I can see the wide frequency range from the turbine will create some problems.
But that shouldn't stop you getting it to all work at mains frequency first.

Best of luck.


Mike

[asmallri]