Battery voltage sensor placement TrisStar 45 MPPT

[ChrisOlson]

Don't think there is an easy answer here. In order to come up with a conclusion you will need to data log your sources over a large period of time (large enough that you can average out and hope you have a decent best day / worst day comparison)

I got that for the last five years.  But stuff changed in the meantime.  We put in a whole new system last March and it's 24 volt, where our old system was 12.

The Schott panels were just added in September in the hopes they would help out more in the winter.  The Sharp panels are 5 years old.  All the solar now runs thru the same shunt that measures and logs the data in my APRS logger.  So I would have to get another shunt and separate the two arrays out for logging purposes.

Looking back thru my logs from the last five years I see there was many times in January I got more power from the Sharp panels on 12 volt than I'm getting from all of them this January.  But that doesn't mean the weather wasn't different back then.  I remember in 2009 the lake (Superior) was froze over and we had more clear days then.

The only way I will really know is to put a different controller on and see what it does.  I can easily wire the two strings of Sharp panels from parallel to series to get the voltage up to 218 just by changing a couple wires around in the combiner box during the night when they're dead.  But re-wiring them in pairs for series parallel would be a chore to try them direct-hooked on 24 volt.  But I think on 24 volt they would be not as good as they are on 12 because Vmp is like 17 point something.

So I don't know for sure.  I was curious what you are seeing.  I'm going to try a different controller.
--
Chris

[rossw]

Don't think there is an easy answer here. In order to come up with a conclusion you will need to data log your sources over a large period of time (large enough that you can average out and hope you have a decent best day / worst day comparison) .

Add a pyranometer to your solar logging and you start to get some reasonable data fairly quickly.

If the pyro tells you there are (x) watt-seconds per square metre of available power during the day, it's a fairly simple calculation of the area of your panels * their efficiency compared to the pyro's output, to see how you're going.

This can be done over fairly short periods of time (minutes even), to give an indication of how your arrays are working under different lighting, temperature, load and controller conditions.

[ChrisOlson]

If the pyro tells you there are (x) watt-seconds per square metre of available power during the day, it's a fairly simple calculation of the area of your panels * their efficiency compared to the pyro's output, to see how you're going.

Whoaa here, Bessy!

This is the kind of info I'm looking for.  <light bulb comes on over noggin> My weather station has one of those but I don't have it hooked up and it's still in the box.  Mainly because I don't know what the data means even if I had it.  It measures UV and a bunch of other stuff too.

Thanks!  That's something I could do that would probably be a lot better than making stabs in the dark trying to figure it out.
--
Chris

[Rover]

I agree with Ross .. to a degree.. thing is you have other sources boosting battery bank voltage, all of the solar controllers (mppt) will plot against some form of algorithm based on the voltage.

You have other things conflicting.... yes you can put up a solar sensor.... but you are trying to evaluate a complex equation, it is not longer  how much from solar .. because you can't isolate. Ross suggestion gives you potential.. but cannot take into account the SOC of the bank etc  for what you  are looking for.

Rover

[rossw]

On re-reading that, I've missed a couple of words. They make a change mostly to the accuracy of my terms.

The pyro output is usually watts per square metre.
Your panels also produce power, in watts.
However since you will be recording over time, not instantaneous output (I infer that from the use of a logger), then using instantaneous values over a period of time is largely meaningless.

If you monitor every second and accumulate, you get instantaneous-watts, every second, which is close enough for our purposes to being "watt seconds". Add them up and store the values every minute or 5 minutes or whatever, and you get a reasonably accurate "watt seconds" per unit time.

Here's a graph from my own from yesterday and today:


The blue area is the number of watt-seconds in each 5 minute period.
The red line is the PEAK watts during each 5 minute period.
The black line is a calculated estimate of the total incident radiation that would be available to a perfectly aligned array.

To explain some things that are evident.
1. The dips in the blue, where the red appears more-or-less unchanged is from something casting a shaddow. Possibly a bird sitting on the sensor, or a small cloud.

2. Dips in the blue, with an accompanying spike in the red line shows the cloud-edge effect, magnifying the sun causing a higher "peak" power, but then the shaddow causing a reduced total watt-seconds in the interval.

3. The black line is complex. My pyranometer is a cosine-corrected instrument. It takes into account the angle of the sun.
Some of my arrays track the sun, so I want to know how much "should" be available to them. This black line is calculated backwards from the power measured and corrected, and the calculated position of the sun.

Tip for you, chris - when you install your weather station sensor, install it close to your solar panels, and most critically - install it facing exactly the same way. That will NOT be the way they tell you to install it, which will be perfectly level. If you want to use it for this purpose, you want to know the sun relative to the panels, not relative to the ground.

[ChrisOlson]

Tip for you, chris - when you install your weather station sensor, install it close to your solar panels, and most critically - install it facing exactly the same way. That will NOT be the way they tell you to install it, which will be perfectly level. If you want to use it for this purpose, you want to know the sun relative to the panels, not relative to the ground.

LOL!  I was reading the instructions for it as we speak.  It says to mount it on a pole in an unshaded area.  It shows a picture of it and it's supposed to be at the top of the pole so it doesn't get any shading from the pole during the day.

I'm going to make a bracket and bolt it to one of my solar panel frames at the same angle as the panels.

Yes Rover, I have other sources of power that may affect what the MPPT controllers do.  But to me this makes sense to measure the solar power available vs what I'm actually getting.  My APRS logger logs to the SD card every 2 seconds.  And I have to dump the data into a spreadsheet or MySQL database to graph it or analyze it.

My weather station does the same thing - it logs the raw data to csv, except every second instead of two seconds.

I'm not seeing why I shouldn't be able to compare the two, irregardless of what the bank voltage is or what the rest of the system is doing at the time.  If I'm only getting half (or whatever) of the raw kWh over time of what the available solar power and efficiency of the panels says I should get, I got a problem someplace.  And that's all I want to determine, really.  Is the solar power actually there and I should be getting more power from those panels?  Or is the real problem just that the sunlight isn't getting thru because of conditions?
--
Chris

[rossw]

LOL!  I was reading the instructions for it as we speak.

Shhh! Don't admit to that, or everyone will think you're a wierdo!

I'm not seeing why I shouldn't be able to compare the two, irregardless of what the bank voltage is or what the rest of the system is doing at the time.  If I'm only getting half (or whatever) of the raw kWh over time of what the available solar power and efficiency of the panels says I should get, I got a problem someplace.

Rover has a valid point that you're perhaps missing.

Unlike a wind turbine, were you have to keep it tamed - if you're not *USING* the power, you have to *WASTE* it, solar panels are different. If you can't *USE* the power, you simply don't *TAKE* it.

So if your batteries are at a point that the charge controller has moved into "absorb" or "float", it won't be taking all the power your panels *COULD* make, so whatever you're measuring won't be indicative of what you could achieve. It's a fair point, but if you can compare "potential" to "realised" outputs, you will at least get a decent idea of what's what when the panels ARE able to deliver as much as they can to the batteries.

[ChrisOlson]

So if your batteries are at a point that the charge controller has moved into "absorb" or "float", it won't be taking all the power your panels *COULD* make, so whatever you're measuring won't be indicative of what you could achieve

Yes, correct.  I know the wind systems can walk all over the solar panels on a good day.  If the turbines are putting out 200-250 amps the controllers will have the panels "throttled" back before they even get a chance to do anything.  I plan on looking at this data on a day when the turbines can't keep up so the controllers are running the panels at "full throttle".  Those are the days when I'm disappointed in their performance.

When the wind is blowing I could care less if I even got those solar panels.   But when the wind doesn't blow hard enough to get the bank all the way up, that's why I got them.  Several guys have told me that I should be seeing 25-30% from those panels on cloudy days when the bank is trying to get to absorb, and I don't even get close to that.

The only way I'll really know is if I put that sensor in for my weather station so I can measure it.  I've pissed and moaned about it since last fall.  It's time to do something to see if my pissing and moaning is justified, or if my solar setup has a problem   
--
Chris