solar power down? added a combiner box.

[richhagen]

In the building I live I have several panels connected to a 10 guage or so cable to my charge controller and battery box in the basement.  One night when I came home from work I was delighted to see that the UPS man had left me a small used panel that I had ordered.  Excited as usual about these types of things, I took it right away up on the roof and hooked it up.  It is a 12 volt system and the wires were connected in a small rainproof outlet box.  Thinking all was well I climbed down and went on to other tasks.  I noticed over the next few days that my power was down.  I got scared thinking that maybe somehow in the dark I had connected the panel backwards or something.  I didn't want to mess up the panels.  I built a box with diodes and fuses for each panel capable of connecting 10 panels.  





The Diodes mounted on the copper strip to the right are 12 amp 100 volt stud mount diodes.  The copper strips were made from 1 and a half inch copper tubing.  I drilled the holes with a number 21 bit and then tapped them with a 10-32 tap.  for the diodes I greased them and added a nut to the back side so they couldn't come loose, although being soldered to a 12ga. wire probably would have prevented that anyway.  The heatsink (copper strip) is not that large, but none of the diodes should see over 4 amps or so as the largest single panel is 65 watts.  In thinking about it if I were building it again I would have made that strip wider to allow for more heat transfer.  I put fuses for each panel, although they should never blow, it is an additional safety in case of something wierd or overlooked, and so that I can pull it an analyze each panel individually from the roof.  There is also a fuse on the transmission cable outside of the box.

I climbed up on the roof, took everything apart, and then tested each panel, all appeared at least open circuit to be ok. What a relief!  I'm guessing that the wire nut connection failed because there were too many wires and it was outside.  We'll see if the screw terminals hold up over time, at least there is only one wire per screw.  When I climbed back down and looked at the voltage on the charge controller, it was already up a 10th of a volt on my batteries.  So far, so good, we'll see how it holds up.  All good and fun, Rich Hagen

[Tom in NH]

Hi Rich,

I'm getting ready to bring six more hundred watt panels on line. Before reading your post I thought I would just parallel them all together and run one wire to my charge controller. I like your system. It's like the opposite of a distribution box, whatever that would be called. I think I should make something like it. Thanks for sharing. --Tom

[ghurd]

Looks good.

Whats the CU strip on the left for? Ground?

I do about the same thing. I solder crimp-ons to the wires.

The terminal strip screws tend to get loose over the first night. I tighten them again the next day. Some anti-corrosion grease (like Ox-Gaurd) on the fuses and screws helps if it's humid. I solder and heat shrink all the connections not on a terminal strip.

The stud diodes are overkill (I like that) and I bet you'll have no problems with them.

It would be interesting to test the forward drop at 4 amps. I'm guessing it's under 0.5v.

G-

[Snoprob]

Wondering what and where would best lightning protection be added to similar setup.

The semi-pigtail curls of wire off diodes to fuses got me thinking about it - old phone co trick to add a few coils in immediate proximety to ground where lines enter house circuts as cheap insurance againts lightning caused fires...

[richhagen]

Yep, the strip on the left is for the negative lines on the panels.  On my system, the negative terminals are ultimately tied to earth ground, although not in the box here.  I originally planned on the crimp on terminals for all the wires, trouble was I was getting very littl power and had to put it up and didn't know which building I had left my crimpers and such at.  I found I left them at the building with the 48 volt system on it.  I'll check the screws tomorrow evening/night, and then add the crimp on terminals to at least some of the wires.  Thanks for the tip on the screws loosening.  As for the voltage drop, I havn't measured it yet, I would bet its about a half a volt as is usually typical for these type diodes at lower currents.  I was trying to avoid the voltage drop, as I didn't need it with the controller to prevent discharge from the batteries, but came to the conclusion that it is better safe than sorry with a bunch of panels paralleled.  After I put the last one up and the output dropped, I got real worried that I had just messed up a bunch of expensive panels.  Rich Hagen

[richhagen]

I would think that passing the wire around a very well grounded pipe a few times and then adding a bank of varistors downstream from that and a reasonable sized capacitor still further downstream, all upstream from the charge controller would provide some measure of protection.  I havn't done that here yet, except for the turns around a water pipe.  The idea being that the inductance in the wire would slow the change in current through it and in the meantime, the huge voltage spike in the wire would find a place to go by arcing across the insulation on the transmission wire and into the ground.  Then the small percent of the energy that made it past the inductance would be absorbed by the varistors and then the capacitance downstream.  Don't hope to ever have a test if this would actually work.  I'll probably ultimately pick up a Delta Lightning Arrestor or something similar.  Even a nearby strike can introduce excessive voltages that can damage charge controllers and other electric equipment.  The panels themselves are P/N junction diodes and would probably be pretty easily damaged as well.  I am not sure how to protect them if they were hit, so I would guess the best strategy would be lightning rods, tall trees, or something taller far enough away that would make a much better path to ground.  

A lady I work with had a tree struck outside that did a few thousand dollars damage to her homes electric wiring and some appliances.  As near as she could tell it didn't strike the house directly, but a tree a hundred feet or so away.  Rich Hagen

[richhagen]

For a typical 100 watt panel in full sun you could see about 6 amps at rated current, maybe a little more when new under perfect conditions.  As I mentioned above, the only thing I would probably do different if I were you at this point at least is to make the strip the stud diodes are bolted to larger so that more heat could transfer on those sunny New Hampshire summer days.  Rich Hagen