Ok so I select the day I want the logged data for. I select the values I want to chart and exported, volts, amps, watts and so on. I hit export and a file saves to my pc's desktop that I can open in excel.
Now when I open the file the detail (resolution) is every 2 second? scrolling down the page going dizzy for e v e r!! I :P
I have little spreadsheet experience so perhaps its a walk in the park for some to tone down the sample resolution to perhaps every 5 min or more .
Then there is the time stamps and some formula you can use to make it readable. I sent the file to a friend who deals with spreadsheets and data logging and he sent it back saying the following quoted below.
QuoteHi Kurt, I cant seem to get any sense from those csv files, which have a time stamp of milliseconds since 1/1/1970. I used the supplied formula and set the format to time, but it just goes:
1:58
3:58
5:58
.
.
.
23:58
then starts again at 1:58 after only 24 lines... which clearly is incorrect.
Any chance you can send it as an Excel or OpenOffice file?
I know less than he dose so I don't have the answer. Has anyone created a working template so the data is readable on excel?
I just highlight the columns I want to look at and make a graph of them in excel.
boB
Quote from: offgridQLD on February 06, 2013, 03:21:40 AM
Ok so I select the day I want the logged data for. I select the values I want to chart and exported, volts, amps, watts and so on. I hit export and a file saves to my pc's desktop that I can open in excel.
Now when I open the file the detail (resolution) is every 2 second? scrolling down the page going dizzy for e v e r!! I
I have little spreadsheet experience so perhaps its a walk in the park for some to tone down the sample resolution to perhaps every 5 min or more .
Then there is the time stamps and some formula you can use to make it readable. I sent the file to a friend who deals with spreadsheets and data logging and he sent it back saying the following quoted below.
QuoteHi Kurt, I cant seem to get any sense from those csv files, which have a time stamp of milliseconds since 1/1/1970. I used the supplied formula and set the format to time, but it just goes:
1:58
3:58
5:58
.
.
.
23:58
then starts again at 1:58 after only 24 lines... which clearly is incorrect.
Any chance you can send it as an Excel or OpenOffice file?
I know less than he dose so I don't have the answer. Has anyone created a working template so the data is readable on excel?
Insert a new column to the right of column A and insert this formula into the new cell B2. You will have to change the -8 to your time zone off-set. The -8 is for USA PST. Change the cell type to DATE and TIME and your B2 will now show the date and time for that row of data. B3 will of course have to use A3 in the formula and so on
=(A2/1000 + (-8)*60*60)/86400 + 25569
Thanks Resthome I will give that a go.
Below is a screen shot of some exported data from my classic to excel. My friend asked me what the (Input Current) column represents?
I was about to tell him its the current the PV is feeding into the controller but then when I looked at the (input voltage) at the time and multiply the input current first row 36.8A by the input volts in the first row 70v I only get 2576w yet its showing in the same row I'm putting 3334w into the battery. Can some one explain this to me?
If I take the battery voltage in the first row and multiply it my the output current in the first row 56.9v x 58.2A = 3311w close enough 3334w with a bit of rounding I'm happy with that. But input current and input voltage don't add up ???
(http://i1105.photobucket.com/albums/h352/kurtthrun/sampleone_zpsa93832d8.jpg?t=1360196423)
Kurt
Just giving this one a bump :-[
So can anyone tell me Looking at the above logged data.
Why input voltage 70v x input current 36.8A = 2576w yet at the same time its showing 3334w charge into the battery?
what dose input current and input volts represent? As its not adding up.
I understand the 3334w is coming from battery current x battery voltage, but if the input voltage x input current is only = 2576w thats 500w less than the charging watts number?
Thanks
Kurt
Quote from: offgridQLD on February 07, 2013, 05:12:31 PM
Just giving this one a bump :-[
So can anyone tell me Looking at the above logged data.
Why input voltage 70v x input current 36.8A = 2576w yet at the same time its showing 3334w charge into the battery?
what dose input current and input volts represent? As its not adding up
Thanks
Kurt
Sorry, cant offer a solution for this but I looked through my logged data (coming from serial port of the Classic to RaspPi) and its seems that the input wattage is not too different from the output wattage. I must graph it all out in case I have issed msomething obvious.
So I would suggest this apparent anomoly may be an issue with the data storage or processing in the local app
dgd
Thanks DgD,
its not just a once of thing I just exported this mornings data. I also looked over the past months data its showing the same error. Input wattage is always way off like 15 - 20% less than output wattage.
Kurt
Kurt,
I suppose the good news is that the Classic is ok, its the reporting in loacal app that is weird. I have spent too long staring at the MNGP and often go to the status screen that shows input voltage and amps along with battery volts and amps. The wattage figures seem to be the battery VA and input wattage is always somewhat higher by 3 to 5%.
dgd
A little higher is fine as I would expect some eficancy losses from input to output. But in the exported data input watts is 20% lower than output watts its imposable as you cant make something from nothing.
I will have to check at the actual classics interface this afternoon when I am at the house. Its just a little annoying when you have a application that's sole purpose is to give you data so you can monitor your system. Yet you cant trust what its telling you as a lot of it is gibberish!
I would appreciate it if some one could check there exported data and see if this (input watts vs output watts) is a common APP bug or just mine?
Kurt
The battery power (VA) is quite a bit more accurate than the input current and voltage accuracy
so your input power calculation will be off.
I think you will find the input current calibration is more of an offset inaccuracy.
BTW, don't try to figure out unit efficiency by looking at meters. It will always give you erroneous
results no matter whos charge controller you're using.
The battery current is actually pretty good.
I suppose that if you needed to, you could take the exported data and use a correction table in
excel to make it dead nuts right on, or much closer. The input current and voltage should be very stable
with temperature. The only thing that might change, and I need to fix eventually, is the
input current zero offset. The battery current is zero corrected at wake up.
boB
Thanks for that answer.
I don't need to use input V & A data its not of much use other than working out the regulators efficacy I already trust its efficient.
I am just trying to make a list of what data from the local app is reasonably acculturate and what should be ignored or used as a vague guide.
Like I said before I am happy with the output volts and amps data to the battery it matches what my external shunts are telling me.
From now on I just wont export input data into the excel files.
Kurt
I usually export the input voltage because I find it useful to get an idea of there is partial shading
or to see of the tracking is basically working correctly. Input current is not very useful I don't think.
I find that I can tell quite a bit about a system just by the voltages.
For instance, if the system is in current limit, absorb or float voltage regulation, the input voltage
will go up. If there is no partial shading, the input voltage will be roughly 75 percent of the Voc
but you can't always tell what the Voc is.
boB
Kurt,
The local app does seem to get the input and output amps resonably accurate. At least in the live display part they are
quite close, in the pic below showing amps only the input from PV is 32v and battery is in ABSORB at 29.4v. The in and out wattages are quite close.
So it would seem that its the exporting of data process that has a problem...
dgd
DGD,
I guess it depends on what you call close I see a good 16% difference in your example .
This is a sample from my system yesterday you can see I have a good 30% difference at times.
(http://i1105.photobucket.com/albums/h352/kurtthrun/Inout_zps07afe920.jpg?t=1360359564)
Yet every now and then like in this sample below at 10:58am you can see input and output Amps are identical.
(http://i1105.photobucket.com/albums/h352/kurtthrun/inout2_zpse4da2333.jpg?t=1360359563)
Honestly though as long as classic can read Amps being fed into the battery accurately (within reason) and detect the battery voltage reasonably accurately then I am happy (and I know it dose this as I have measured it myself) . They are the two basic readings that matter when the controller is making decisions on what to do or not to do to your 15K worth of battery's ;)
Kurt
Ahhhh... I see now !
The input current will ALWAYS be somewhat less than battery current.
How much lower depends on the ratio of input voltage to output voltage
plus some losses in the controller.
When you see I/O current very close together means that the input voltage and battery output
voltage are also very close together.
The higher the input voltage goes, the lower the input current goes.
boB
We may not be on the same page but...
I understand that the input amp number show will most of the time be lower than the output amp number due to the fact the input voltage is usualy higher (amps x volts = watts)
So I can see for a given wattage if the volts are higher the amps will be naturally lower.
The issue is when you do the math (amps x volts) for the input voltage and input amps shown you obtain a input wattage number. At the same time you take the output amps & output volt numbers and multiply them to obtain a wattage number. You end up with a big difference between input and output watts for the same sample time I see up to 30% difference
Kurt
OK, I see. Yes, they will be kind of hard to get to agree closely. It's mainly the input measurements
which are not nearly as important to be as accurate as the output measurements.
I would still like the input current/voltage to be much closer though and that will
eventually happen, I hope.
boB
Sounds good to me :)
Kurt