What happens when a leaf falls on my solar panel?

[Justoine Knight]

There's a video over at Sunsil that demonstrates clearly what happens when a solar panel is shaded:

http://www.sunsil.dk/video/

More info here:

http://www.electronicsweekly.com/Articles/2010/07/12/49023/Solar-power-architecture-cuts-shading-losses.htm

This strikes me as a great idea!

[dnix71]

Sweeeeeet. The sunshine state (Florida) is anything but sunshine during the summer. All it takes is a little clouds to kill a panel output.

[WindriderNM]

The panels I have have a diode on the back that somehow keeps the power up if a lief covers one of the cells. When I asked the guy i got them from what the extra connection  tabs were for he told me about the diode. When i wired up the panel i used trial and error to place the diodes by covering a cell, when I got it right there wasn't much drop. I haven't taken the time to figger out how this works yet.

[mcline27b]

"I haven't taken the time to figger out how this works yet."
It works because the diode allows current to flow past the shaded cell.  Without the diode the impedance of the shaded cell impedes the flow of power from the working cells and drastically reduces the current of all the cells in series.

[ghurd]

"I haven't taken the time to figger out how this works yet."
It works because the diode allows current to flow past the shaded cell.  Without the diode the impedance of the shaded cell impedes the flow of power from the working cells and drastically reduces the current of all the cells in series.

Before anyone gets too excited:
Normal 12V panels have 2 bypass diodes.  If a couple cells are shaded, then it sounds like amps will flow around the shaded half and into the battery.  Not correct.
Half the panel makes about 10V open circuit, reduce that by 0.7V forward diode drop, and that only leaves 9.3V open circuit.
Pull a few percent of the rated current from the panel, and the voltage drops drastically.

Meaning if the bypass diode is "bypassing" then there is not enough voltage left to push a milliamp into a battery.

Bypass diodes in standard panels, in standard configurations, have no use in standard 12V or 24V systems,
and are best removed before they cause problems IMHO.
G-

[rossw]

Bypass diodes in standard panels, in standard configurations, have no use in standard 12V or 24V systems,
and are best removed before they cause problems IMHO.

I agree with the rest of your comments, Glen, but I'm a little puzzled by this bit?

If your module has two (or more, I've seen some with 3 or 4 - but they're rare) bypass diodes, what problems do they cause?

Before I'd put isolation (schottky diode in series with each array), when I connected the last 600W array to my existing 2.8KW, there was a sickening popping and tinkling sound... inside the junction boxes I found these:


And you can see where they came from here:


I still can't figure out what caused it - except perhaps avalanche from excessive dv/dt breaking the diodes down.
In any (normal) operation, I can't see how they'd have any (negative) effect?

[ghurd]

See what I mean?  If you would have snipped the little buggers out, they would not have popped! 

I never saw a 12 or 24V with other than 2.

Bypass diodes in standard panels, in standard configurations, have no use in standard 12V or 24V systems,
and are best removed before they cause problems IMHO.

I agree with the rest of your comments, Glen, but I'm a little puzzled by this bit?

If your module has two (or more, I've seen some with 3 or 4 - but they're rare) bypass diodes, what problems do they cause?

They like to fail into a short circuit condition.
The half with the shorted diode just runs in circles.  Nothing goes into the battery.
(possibly they very often fail to open, but that would have 0 effect, so nobody would notice?)

Many people would say this is a minor issue, until they are 60' up on a 75 degree roof, doing everything required to find the problem.
Not fun.

Well, it is almost fun to see a $3 HF meter slide down and hit a cement driveway... until you realize it could be you the next time.

In 12V systems, without high voltage input MPPT, I snip the trouble makes out before they cause me trouble.

I have no good idea why your diodes popped.
The only times I recall seeing them blow up like that was in a panel done by a "retailer" (instead of an 'understander').  He could get three 1A diodes cheaper than a 3A diode, so he paralleled 3x1A, and they blew after a few months or a year. 

<tangent> He sold a lot of panels, he could not fix them, so going cheap meant the customer had to pay me.
Whatever money they saved on discount panels from a discount retailer was used up on my fuel cost alone, plus they had crappy parts, crappy panels and a crappy controller.
A better system, with better parts, would have been MUCH less expensive if they bought it from me in the first place.

Back to your situation, is there a chance the diodes went over Vr with MPPT or something?
Most factory installed bypass diodes I see are 1KVr, some/many 600Vr, but never less than 600Vr.  Most are rated at >2~3x Isc.

At the moment I can't imagine how avalanche would have any effect... let alone for a duration effecting dt... but it is 3:00AM here... and I smell like pepperoni.

Is more than minute avalanche in bypass diodes even theoretically possible in standard systems?
I don't think it is, but...
G-

[DamonHD]

Ross: were those bypass diodes schottky?  If so, then their Vrmax is often only 30V--50V and if you were stringing panles in series without blocking diodes then that might have been the problem.  But as G says, I'd expect bypass diodes to have a high Vr especially if intended for grid-tie, and you know what you're doing, so ...

Rgds

Damon

[rossw]

Back to your situation, is there a chance the diodes went over Vr with MPPT or something?
Most factory installed bypass diodes I see are 1KVr, some/many 600Vr, but never less than 600Vr.  Most are rated at >2~3x Isc.

At the moment I can't imagine how avalanche would have any effect... let alone for a duration effecting dt... but it is 3:00AM here... and I smell like pepperoni.

Is more than minute avalanche in bypass diodes even theoretically possible in standard systems?
I don't think it is, but...

The diodes in question were Schottky, I forget their exact specs but I think they were 12A and 120V.
The rest of my arrays are indentical modules, and have the same diodes in them. (Obviously, using Schottky reduces the drop across the "dead" half)

My *supposition* is that in my haste to bring the new array online - it was around midday, the other arrays (which were all in parallel) were sticking out their 100V, and delivering around 2.8KW. I connected this array (by plugging in the final connector). At this point, they will have seen the full 100+ volts bus voltage, and almost instantly. The dv/dt gradient would have been very high. In all probability, beyond the avalanche zone. *IF* that happened, the diodes will have gone into a conduction mode. With 2.8KW behind the source, a very substantial current would have been available - more than enough to destroy the (lets say, 10A) diodes.

After this "event", I replaced the diodes (The supplier sent me 15 or 20 by DHL, at no charge) - and I then added a series diode to each array to prevent backfeed to one array - and I've never had any indications of any sort of problem since.

I take your point - but I think I'd rather have them there than snip them out "just in case"

[ghurd]

I take your point - but I think I'd rather have them there than snip them out "just in case"

Agreed, because it's 100V+.

[oldgit]

I wonder whatever happened to Sunsil. Sounded like a great idea but it never came to be.

[Mary B]

My guess would be high cost and cheap chinese panels undercut them so bad they went out of business...