Does reverse current degrade a solar cell?

[GreenHorn Greg]

Hi All,

This question comes about by seeing a set of panels wired to a motor in such a way that the panel generating the most power determines which way a small motor turns.

So, small DC motor has two leads: A & B. A is connected to the Positive of Panel X, Ground of Panel X is connected to Ground of Panel Y, Positive of Panel Y is connected to B.

When Panel X produces more power, current will flow in one direction (thus reverse through Panel B) & Vice-Versa.

Here is a link where something is wired up that way: http://www.youtube.com/watch?v=HkhVomoD47g

The concern is damage to the panels but I've not found anything that talks about that directly so here I am askin!

The flip-side of the question might be: Is there a circuit that can reverse polarity to the motor without reverse flow through the "weaker" panel?

Thanks,
-g

[inetdog]

Hi All,

This question comes about by seeing a set of panels wired to a motor in such a way that the panel generating the most power determines which way a small motor turns.

So, small DC motor has two leads: A & B. A is connected to the Positive of Panel X, Ground of Panel X is connected to Ground of Panel Y, Positive of Panel Y is connected to B.

When Panel X produces more power, current will flow in one direction (thus reverse through Panel B) & Vice-Versa.

Here is a link where something is wired up that way: http://www.youtube.com/watch?v=HkhVomoD47g

The concern is damage to the panels but I've not found anything that talks about that directly so here I am askin!

The flip-side of the question might be: Is there a circuit that can reverse polarity to the motor without reverse flow through the "weaker" panel?

Thanks,
-g

You have one serious problem with your design: The voltage output of a solar panel is almost independent of the incident light level. It is the current which changes. So your motor will not turn unless one of the panels it less than moonlight dark.
Also, the panel with the lower voltage will conduct in the reverse direction, but with a resistance so high that you would probably not have enough energy to turn the motor even if the voltage behavior were not an issue.
The normal solution is to use one panel or set of panels to provide power to the motor and two smaller sensor panels or cells whose outputs are connected to resistors to provide a signal proportional to the current. That signal goes through IC or relay logic to switch the motor direction.

In the YouTube video, if the two sets of panels are 12V each, then the voltage driving the motor might be as much as 1VDC unless the panels used have a very high internal leakage resistance. If they have a high leakage resistance, or have a load resistor connected, then that will carry the motor current and it will not actually drive current through the panels in reverse.
Note also so that the "perfect" alignment is not really perfect as seen by the shadows on the flat area of the mount.


And welcome to Solar Panel Talk!

[PNjunction]

And yes, reverse bias on a panel or cell(s) damages it over time.

Some solar panels are actually two or more panels wired in parallel inside a common frame. In the junction box are bypass-diodes so that if one or more of the cells on the panel become shaded, the full current of the rest of the enlightened panel will not be applied to the reverse-biased / shaded cells. We are talking about mono / polycrystalline cells.

In small thin-film panels, there are usually no internal bypass diodes, and reverse bias is not as much of an issue, but it does not go away completely. Shading part of a thin film panel will not *immediately* damage that section that is shaded. Long term shading however, will eventually degrade that part of the thin film panel.

[GreenHorn Greg]

Thank you to both of you.

I wasn't sure if it would degrade or just be inefficient. I was pretty sure the diodes that I've seen in most solar cell schematics wouldn't like it but not sure if it would actually hurt them. I was actually surprised that the setup even worked.

Anyway, thanks again. I'll ask another question about recommended circuits for a tracker shortly.

[inetdog]

PN: There seems to be some confusion here between the effects of reverse bias (applying a + voltage higher than the panel Voc to the + terminal of a panel or string) and forced forward current (applying a + voltage to the - terminal of the panel to force more current through it than the current light level will allow.)

Forced forward current will indeed damage most cells, although mainly by localized heating, I believe.
Reverse bias can damage cells or not damage them depending on the voltages and currents involved and whether the voltage exceeds the breakdown voltage of the cells involved.

[bcroe]

Forced forward current will indeed damage most cells, although mainly
by localized heating, I believe. Reverse bias can damage cells or not damage them depending
on the voltages and currents involved and whether the voltage exceeds the breakdown voltage
of the cells involved.

OK how is forcing forward current much different than running them shorted in sun? Different
mode for the cell I suppose.

Number 83 on things to do, try forcing forward current as a possible means of heating and
causing snow to slide off the (hopefully vertical) panels. I'm expecting them to act as
simple diodes (once the bypass diodes are disconnected). That would mean a diode drop
at each cell, multiplied by the current to get heating power. I don't see any part of the
panel having uneven voltage, is there a problem with current distribution over a cell?

Bruce Roe

[inetdog]

OK how is forcing forward current much different than running them shorted in sun? Different
mode for the cell I suppose.

Running them shorted in the sun, the charge carriers pumped up into the conduction band and across the band gap just flow through the circuit under the resulting electric field.
Without sun striking the panel there are no loose charge carriers and so the voltage across the cell has to rise to the "breakdown" voltage, which may just be the forward voltage drop, but I don't think so, since that would not damage the cells.
The whole issue of what is the forward and what the reverse direction of the diode junction gets tricky when looking at a PV junction.

[bcroe]

Test

Running them shorted in the sun, the charge carriers pumped up into the conduction band and across the band gap just flow through the circuit under the resulting electric field.
Without sun striking the panel there are no loose charge carriers and so the voltage across the cell has to rise to the "breakdown" voltage, which may just be the forward voltage drop, but I don't think so, since that would not damage the cells.
The whole issue of what is the forward and what the reverse direction of the diode junction gets tricky when looking at a PV junction.

Come to think of it, I have a panel down the basement (no sun) and with no bypass
diodes. And a pretty healthy DC power supply. I think the test priority to determine
the forward conduction voltage just got a lot higher. Bruce Roe

[Mike90250]

I think Midnight Solar is exploring a mode for their charge controllers called "Snow Melt" to be used in conjunction with a generator, to clear the ice off arrays.

[mapmaker]

I think Midnight Solar is exploring a mode for their charge controllers called "Snow Melt" to be used in conjunction with a generator, to clear the ice off arrays.

Here's a discussion of the snow-melt feature:

http://midniteforum.com/index.php?topic=5

--mapmaker

[inetdog]

Here's a discussion of the snow-melt feature:

http://midniteforum.com/index.php?topic=5

--mapmaker

And what is being discussed there is really applying a high enough voltage (of the same polarity that the panel would normally produce) to cause current to flow in the reverse direction through the panel leakage resistance and/or breakdown resistance
This is the same direction that the tracker would be forcing current, and would not be affected by the presence or absence of bypass diodes.

[bcroe]

Melting snow electrically

Today we had good sun, and I hauled out my latest DIY solar panel. It incorporates
every trick I know, and tests indicate it is fault free. its a 7 X 9 array of 3" by 6"
cells, and it managed to deliver proper voltage and current. It put out 3.55A, about
all I could expect.

But today I mainly wanted it, because it has no bypass diodes.

I took the panel in a fairly dark basement, and tried applying DC to it.

With the power supply + connected to the panel -, I got no appreciable current, like
a reverse biased diode. I raised the voltage to several times operating voltage, no
change.

With the power supply + connected to the panel +, the panel had normal operating
voltage. But raising the voltage just above normal (1/2 volt per cell), it started to
conduct. It conducted large currents, so I ran the attached curve. It looks to me
just like raising the voltage across a forward biased diode. No problem breaking
anything down, it just needed a bit more than normal voltage.

See figure.
My conclusion, the cell works like a forward biased diode, and the sun supplies an
output current. If the external load absorbs enough current, the voltage across the
diode won't be high enough for much conduction. If the cell is left open circuit, all
the generated current will flow and be absorbed by the cell itself. We like to
operate at the bottom of the corner, to get the most power.

It appears possible, that externally feeding in full current is no more damaging than
setting it in the sun disconnected. Since the cell efficiency is only near 20%, the
heat will only be 20% of that produced by full sun. I'm expecting it to be
spread evenly over the cell, and the panel.

One point here, is this effect won't be bothered by bypass diodes.

The first thoughts are, get the first string cleaned by any means. Now (via the
combiner) let it connect to the second string. The first string will generate the
power to warm the second string, and if it is vertical, the snow will slide off. If
it is really cold, use 2 cleaned strings to warm a 3rd. Once this starts working,
it can expand to eventually clean all strings.

I suppose I'll be out at the combiner this winter trying this out. If it works, some
more automated system will be needed. Maybe the first string will be cleaned
by external power, then mostly use sun power. Maybe the sun produced voltage
won't be high enough, so there will need to be some kind of boost. Could add a
bit of series DC from the line, or could short some panels in the string to be melted.

With 12 panels in a string, I could short 2 panels with a relay briefly, then short
the next 2, then the next 2, so the voltage is low enough to suck all the current
over, but on average to every panel in the string.

Maybe someone has a working system, but the info hasn't made it to this list. The
Midnight stuff says someone thought about it and ran an experiment. But that was
years ago, and I see no follow up.
Bruce Roe

[mapmaker]

The first string will generate the
power to warm the second string, and if it is vertical, the snow will slide off. If
it is really cold, use 2 cleaned strings to warm a 3rd.

Vertical is key. If you can get your panels vertical, you will rarely have much snow on them, and you won't need to melt any.

As far as using two strings to clean a third string... you may trip the string breaker on the third string (or damage the third string) if the current exceeds its series fuse rating.

--mapmaker

[bcroe]

Vertical is key. If you can get your panels vertical, you will rarely have much snow on them, and you won't need to melt any.
As far as using two strings to clean a third string... you may trip the string breaker on the third string (or damage the third string) if the current exceeds its series fuse rating. --mapmaker

Agreed; it depends on the sun & elevation angle. The panels are rated 8A, but were fused 15A;
I considered going to 12A or 10A. There are no breakers, running at 420 VDC.

Maybe power to melt the first string could come from the inverter building 300' away.
Like, 420VDC at 8A.

I'll see by end of winter if I have damaged any strings. It would be cool if the
entire operation could be made automatic. Or, if an infrequent event (wind blowing
against the vertical panels), at least manual controls are set up, with auto timers to
complete each stage. Bruce Roe

[PNjunction]

.. I took the panel in a fairly dark basement, and tried applying DC to it.

That's not really a real-world situation.

What we're mainly concerned with here is shading on one or more cells creating localized hot spots, while the rest of the cells in the panel are fully illuminated.

pveducation.org has some good material on bypass diodes used to prevent this problem. Section 7.2 goes into detail and has a far better explanation than I do!