Local app - the mystery deepens

[vtmaps]

When you double the voltage you quarter the power loss.

Yes.  But... does it matter?  The cable loss goes as the square of the current.  During bulk charge is when the current is highest and the loss hurts the most.   As current tapers off during absorb, the cable loss becomes less and less important.  In many systems, charging current starts out low in the morning, and by the time mid-day full power is possible, the batteries are absorbing.

When designing a system we usually use Imp (which is seldom achieved) to calculate cable loss.  The real loss is usually much less.

Regarding equalization, currents should be low (relative to bulk current) and cable loss is often blamed for the voltage drop of the hot panels.

Please understand that I am not arguing for thin cables... but just trying to put them in perspective.  Raising string voltage very often just shifts heat production from the cable to the controller.

The Vmpp is 35.5v. They might struggle in summer when it's hot but the days are so long and sunny it doesn't matter.

And that is the bottom line 

Things could change if jimbo wishes to harvest full power all day long with opportunity loads... thicker cables and/or higher string voltage may needed.

--vtMaps

[jimbo]

Great explanation 

[zoneblue]

Vt, youre spot on in principle, no argument. However the perspective was 30m at 30V? If i did that here on my 6awg, id lose 16% of my  average peak power.
http://www.calculator.net/voltage-drop-calculator.html?material=copper&wiresize=1.296&voltage=30&phase=dc&noofconductor=1&distance=30&distanceunit=meters&amperes=60&x=48&y=15

Using your model, if we use instead the average (sun hours) throughput ( which for say yesterday was 2566Wh over 6 hours,) is 430W, then id still get 4% loss.
http://www.calculator.net/voltage-drop-calculator.html?material=copper&wiresize=1.296&voltage=30&phase=dc&noofconductor=1&distance=30&distanceunit=meters&amperes=60&x=80&y=16

But that will be low because the losses will bite harder in the top half of the 'average'. Youd need to do a proper integration to get a true answer.

Becasue of the above, the risk is that the bank never reachs absorb at all.

I actually think Jimbos modules are 72 cell modules, so the voltage is there in principle, but id be watching the losses in the period just prior to reaching the absorb setpoint. Why not do an experiment: measure the voltage at the array, and at the input and output of the controller just as absorb is reached. Its easy to do.

Edit: distance fix

[jimbo]

Yes i would like to test it one day but the wiring up on the roof is very messy and hard to get to.

Off the top of my head there are 2 runs of 5 AWG for 100 feet (you had 300 feet) which gives a drop of around 1.88v. If it is 6 AWG then the drop is 2.37v. Not much but does affect the system when equalizing with high loads on.

As stated this is not ideal and will be changing the wiring when I get around to it.

So how much efficiency is lost through the midnite while dropping the voltage?