A contrary view: There is no wasted power when charging a lead acid battery with a solar panel. While the battery is accepting charge, the battery terminal voltage (6 cell) is between 12.5V and 14.5V. That happens to be near the MPPT voltage for an 18 to 22V panel. So, no "regulator" is needed, effectively the panel should be connected directly to the battery.

It is only after the battery reaches full charge (indicated by the battery terminal voltage climbing north of 14.6V) that a regulator is needed. Since the battery is already charged, who cares if now the regulator is producing heat instead of modifying the electro-chemistry of the battery.

The regulator should be the kind that uses a PFET in the path between the solar panel + lead and the + pole of the battery. During the initial charge phase, the PFET is turned on hard, meaning that the voltage drop across it is only a few mV, and the wasted power is nearly zero. After the battery reaches full charge, the PFET progressively is turned off, so now the PFET begins dissipating heat. Same would be true if the using a shunt regulator; no power is wasted initially, charging the battery in the shortest time. Who cares if you make heat after the battery is charged...

Forget the LM317. It is not suitable for this task.

 

It is only after the battery reaches full charge (indicated by the battery terminal voltage climbing north of 14.6V) that a regulator is needed. Since the battery is already charged, who cares if now the regulator is producing heat instead of modifying the electro-chemistry of the battery.

I generally agree with your analysis, especially for small systems. But what if there is a concurrent load, so that you want all the power you can get going forward? You wouldn't want to be throttling the output just because the battery is charged. requires

I said it in another thread - optimizing alternative energy systems requires knowing the end-to-end parameters. Optimum for one is not for another.

 

... But what if there is a concurrent load, so that you want all the power you can get going forward? You wouldn't want to be throttling the output just because the battery is charged...

Over a Billion cars cant be wrong...

Cars use flooded-cell lead-acid batteries with a simplistic voltage regulator that uses just the battery terminal voltage for its sensing input. There are many loads (engine, lighting, air conditioning, electronics) connected in parallel with the battery while it is being charged by the alternator (a current source, not unlike a solar panel).

The voltage regulator throttles the alternator output current so that when the battery voltage is 14.5V (set-point +_delta), the alternator current perfectly matches the total load current, and zero current flows in/out of the battery (unless stopped at a traffic light). Up to the point where the battery voltage is less than the regulator set-point, the difference between the max. alternator output (typically 40 to 60A) and the total load current is available to charge the battery.

 

But reducing the power draw at the source (by adjusting current in the field windings) is different than burning off excess power in a transistor. In a car, unused alternator power means un-burnt gasoline. With a solar system, you generally want to extract as much of the available power as possible, and for this OP that would be max current at 17V or so. Throttling at 14.5 V would reduce peak power output.

Of course, the OP may not care one wit about that and just wants a charged battery.

 

Over a Billion cars cant be wrong...

Cars use flooded-cell lead-acid batteries with a simplistic voltage regulator ...

But car batteries hardly ever get discharged.

 

But reducing the power draw at the source (by adjusting current in the field windings) is different than burning off excess power in a transistor. In a car, unused alternator power means un-burnt gasoline. With a solar system, you generally want to extract as much of the available power as possible, and for this OP that would be max current at 17V or so. Throttling at 14.5 V would reduce peak power output.

That exactly the main idea for this topic.

I like read your discusses

My summary, I must have switch regulator with good capacitor at input, or find MPPT controller.

 

...

My summary, I must have switch regulator with good capacitor at input, or find MPPT controller.

I dispute that. If you choose the right panel, while the LA battery terminal voltage is between 12.0V and 14.5V (as the battery charges), the panel can be operated at so close to its natural MPPT operating point that adding a MPPT switch-mode converter adds almost nothing to the overall efficiency. The most you might gain is ~5%. The losses in the SMPS will be >5%.

 

But car batteries hardly ever get discharged.

Right, they are always operated between about 85% of full charge to 99% of full charge (unless you leave the headlights on overnight). That doesn't effect how the voltage regulator does its job, however...

 

I dispute that.

Me too. The capacitor is irrelevant and only adds cost. And the benefit of the MPPT controller depends on the load profile. If you're only charging a battery, the benefit is small at best or non-existent as Mike has noted. If the battery is just a small part of the overall load, the benefit of MPPT might be greater.

But aren't we talking about a 20W panel? No way you can justify the complexity of a fancy system just to capture a small percentage of those 20W. It would be cheaper, I think, to buy another panel than to incrementally boost the one you have.

 

I think Mike is right on this one. For example in the circuit I showed I will charge with the full panel current (which is actually a little high for his battery). If there is a load it will be supplied first with the remainder going to the battery. My high voltage set point may be to high, especially if it is a sealed lead acid battery, so we might want to lower it. With the battery at rest it is fully charged at 12.72 volts, so anytime it goes below fully charged it will switch on again.
With this method there is almost no loss. I'm not sure what you might do with a switching regulator that would be better.
I don't think you want to set the terminal voltage as high as 17 volts due to gassing.

 

I did some crude calculations: Switching regulator verses direct connection

Assume
- Standard insolation at the equator, 100% for 8 hours, 50% for 2 hours, 0% for 2 hours(voltage too low to charge battery).
- Typical solar panel I-V characteristics from TI MPPT app note.
- 90% efficiency switching regulator.
- No MPPT correction applied

Switching regulator:
8hr X 18watts =144 WH
2hr X .45A X 14.4V = 13WH
total = 153 WH

Direct connection
8hr X 1.1A X 14.4V = 126.7WH
2hr X .5A X 14.4V = 14.4WH
Total = 141 WH

The switching regulator wins by 12 watt hours.

 

Les, I am confused. If no MPPT correction is used, what is the SMPS doing? How can you gain any efficiency if the losses in the SMPS are 10%?

 

If there is a load it will be supplied first with the remainder going to the battery.

And I keep niggling over what happens then, after the battery is charged. Imagine the battery is small or non-existent and you want the biggest load you can supply with your panel. In that case, MPPT could be worth it. In a larger system. If the OP even cares.

 

Les, I am confused. If no MPPT correction is used, what is the SMPS doing? How can you gain any efficiency if the losses in the SMPS are 10%?

Because when you connect the solar panel directly to the battery you loose power. At max sun light the panel produces 18V at 1.1 amps (20watts). You connect a battery to it and the voltage drops to a max of 14.4Volts (15.8 watts). At 50% sunlight the direct connection wins.

Remember, solar cells are constant current devices.

 

I dispute that. If you choose the right panel, while the LA battery terminal voltage is between 12.0V and 14.5V (as the battery charges), the panel can be operated at so close to its natural MPPT operating point that adding a MPPT switch-mode converter adds almost nothing to the overall efficiency. The most you might gain is ~5%. The losses in the SMPS will be >5%.

Which assumes you can select a panel with a small known range of terminal voltages.

The problem is you can't find that panel. The voltage varies with many factors, temperature being among them.

 

Most 12vdc rated panels Sharp-NE-80EJEA-80W-12V-Solar-Panel below a few hundred watts will operate fairly close (~10%) to the MPPT during bright sunlight direct connected (PWM full on) while charging in BULK mode in a practical system with voltage losses in wiring/electronics unless the battery is completely discharged. Once you leave bulk mode charging MPPT gains almost nothing if the panel is matched to the needed absorption charging current as MPPT will be disabled when the charger starts limiting voltage during the the absorption/float stage.

If light is marginal MPPT can help but today it's cheaper to just buy another panel if you have the space until you get above about 500W of system power at 12vdc.

 

The subtleties of an MPPT are generally misunderstood. It is not a simple SMPS yet it requires switch mode electronics. MPPT is where the battery voltage is monitored and the load on the solar panel is adjusted to yield the highest charging current at the battery. It is a constant ballet between the state of charge of the battery and the available power from the panels.
The MPPT function mainly comes into play when the sunlight is waning. When light is waning the MPPT will adjust (lower) the battery charge current so that solar panel will put out its max power thus charging the battery at the highest rate possible for the conditions. A good MPPT will only increase your power yield by about 20%.

 

...

Remember, solar cells are constant current devices.

No they are not. If they were, then running them at the highest possible output voltage would produce the maximum power out of the panel. In fact, the MPPT operating point (where it produces the maximum output power) for a 20V open-circuit Panel is about 16V.

 

No they are not. If they were, then running them at the highest possible output voltage would produce the maximum power out of the panel. In fact, the MPPT operating point (where it produces the maximum output power) for a 20V open-circuit Panel is about 16V.

It looks like a constant current function to me.

 

No they are not. If they were, then running them at the highest possible output voltage would produce the maximum power out of the panel. In fact, the MPPT operating point (where it produces the maximum output power) for a 20V open-circuit Panel is about 16V.

It's not a perfect current source but almost every electrical model has a current source in it because it's easy to simulate and characterize cells using that equivalent circuit with a few components.
Solar Cell