Shorting a Solar Panel to stop charging

If I charge a battery through a diode from a solar panel, I should be able to stop the charge by shorting the solar panel with an N-CH MOSFET (e.g. drain on PV+ and source on PV-). The diode should block the battery form the short. Does anyone have thoughts on doing that?

Crude way, but it should be possible.
You could PWM the mosfet, to have some sort of transistion phase.

What sort of battery are we talking about, and how big is the solar panel (voltage/current).
Leo…

Where is the power going to go? Which component will get heated up until it bursts?

I have an old 75W (SP75) panel that has a shorting current of about 4.8A, and an old lead-acid car battery which I think is about 50AHr. I will try to get it going on the bench first where I have a small constant current mode supply. I think the supply has capacitance on its output so I'm not sure shorting it is a good idea.

I want to use all six PWM's for LED drivers (e.g. dual PicoBuck), so perhaps I could bit-bang the solar to have some sort of transition (e.g. maybe some duty starting at about 14.3V and all the way off at 14.7V).

So to get far ahead of where this idea is at, I am sort of thinking two color changing lights (e.g. 350mA in each RGB LED) in an outdoor pool for the night, and perhaps the short could be swapped for a water pump that would run when the battery was charged.

@MorganS are you asking where the power goes when the solar panel is shorted? I have an idea that answering that is complicated beyond my abilities. I did have some solid state physics classes twenty plus years ago but I don't really know what happens to the power when carriers are generated in a junction with little voltage across it.

Maybe better to use an n-channel fet as a switch in the negative line of that panel.
Source to battery negative and Arduino ground, drain to solar negative, solar positive to battery positive.
Leo..

Thanks, @Wawa, I see what you are saying, let me think on this...

Wait there is a body diode to consider... I have been got by those a few times.

Why not just open the circuit to stop charging?

...R

ron_sutherland:
Wait there is a body diode to consider... I have been got by those a few times.

Tripped me up too.
Maybe a p-channel fet, high-side.
Source to solar+ and drain to +batt.
Resistor between source and gate, and gate pulled to ground with a transistor.
Leo..

Robin2:
Why not just open the circuit to stop charging?

…R

Yes, that is the elephant in the room.

ron_sutherland:
@MorganS are you asking where the power goes when the solar panel is shorted? I have an idea that answering that is complicated beyond my abilities. I did have some solid state physics classes twenty plus years ago but I don't really know what happens to the power when carriers are generated in a junction with little voltage across it.

The power gets wasted in parasitic resistances. The wires between the panel and the MOSFET have resistance. The panels themselves have resistance. The MOSFET has a resistance.

Now if it's the wires or the MOSFET heating up with (nominal) 75W then that's going to start a fire. If it's just the panels, then they have a lot of surface area to dissipate that heat and they're outside on fireproof mounts.

Open-circuit has its own dangers. The panels called "12V" will do 18V when there's no current flowing. That's pushing the limits of many MOSFETS, which will be damaged at 20V. If there's any inductance (even a long straight wire) then short-term spikes can easily exceed that voltage.

Why not just use an inexpensive charge controller, designed for exactly that purpose, like this one?

I've used the 15 amp version for several years, and it works very well.

This is the two options that I can make sense of at the moment.

My instinct is saying to go with the open circuit method. But I am not sure that is the better choice. If I PWM the open circuit it will stop and start the current flow, but if I PWM the short circuit the current flow between the battery and solar panel will not change much (e.g. bellow about 15V it is almost like a pure current source). The short circuit is simple and normally I like simple.

@jremington yep, I may do that in the end. I sort of think a low voltage disconnect is important so I am looking at something like a SunSaver-6L.

Both circuits will work.
Only difference is that the top one needs the Arduino to stop charging,
and the bottom one needs the Arduino to start charging.

I would also go for a dedicated charger.
They are much more than just voltage cutoff.
Your battery will last longer with a proper BMS.
Leo…

Not sure shorting a panel is good - unless it has enough bypass diodes the power will tend to dump
into a few of the cells unevenly and cause local heating. It just feels a wrong approach.

@MarkT I understand that it feels wrong, it does to me as well, but is it wrong.

Longer strings of panels can have shadow problems that need bypass diodes to work around. This will be a string of 36 cells each will have about 0.4V when the battery is charging at 14.3V. If I put a short on the panel then the voltage will drop but the current flow will stay nearly the same in each cell. If a bird parks on one of the cells then that cell will not generate current and cause it to act as a forward biased diode. That could be a problem if current flow crowded into one spot of the cell, but I don't think that is what happens.

The same problem happens if the bird parks while the battery is charging. Each cell needs to be able to dissipate the power from the panels short-circuit current at the cells forward drop. Are my wrong?

Good solar panels have AFAIK bypass diodes across groups of cells, to keep reverse voltage below the zener breakdown voltage of the cell.
http://www.pveducation.org/pvcdrom/7-modules-and-arrays/bypass-diodes
Leo..

Not sure about the zener issue, but that web site was good. I have been thinking in terms of using a single string of 36 cells (e.g. I have an old obsolete panel to play with). It looks to me as though the hot spots occur when there are strings in parallel and the combined current can add up to more power than a cell can dissipate. So with that in mind, I think the shorting idea is probably dangerous with strings in parallel.

ron_sutherland:
Not sure about the zener issue..

From what I've read/understood from other solar sites:
Without bypass diodes, the obscured cell gets full voltage from all the other cells in the panel, reversed.
Reverse breakdown voltage of a solar cell is about 6-7volt.
If there is no diode across a group, reverse voltage could reach that breakdown voltage,
and 6-7volt times panel current will be dissipated in that single cell.
That could be 7volt * 5Amp = 35watt in a single cell.
Bypass diodes across single cells are not needed. Just across a group.
So that if one cell is blocked, the other cells in that group can't produce >6volt.
Please correct me if I'm wrong.
Leo..

Hopefully, I understand the issue now...

had some help from SPICE.

SolarReverseBreakdownIssue.png

I doubt my model is right.

I need some time with this.

SolarReverseBreakdownIssue.png