I think everyone is kind of saying the same thing, but he can already over current his battery at 1.1 amps, but as Wayneh says he could use the extra power when the full load is on. The question is how much do you want to pay for another .1 amps, or so, when the load is on. The panel is already almost perfect.

 

Les,

If this cell's I/V behavior was like an ideal constant current source, then its curve would follow the blue line I added to your picture. I indicated the actual sweet spot where this particular cell should be operated.

Note that the MPPT voltage is about 100* 0.4/0.68 = 60% of its 0pen-circuit voltage. For charging a 12V battery (charging voltage 12-14.5V, the ideal open circuit voltage for the panel should be 14/.6 = 23V, ergo, no SMPS is needed. Just hook the panel to the battery and be done with it.

 

View attachment 73439

.......If this cell's I/V behavior was like an ideal constant current source, then its curve would follow the blue line I added to your picture. I indicated the actual sweet spot where this particular cell should be operated......

The blue lines you put on the graph are incorrect. Open circuit voltage (.68V) is only at full illumination and zero current, as in "open circuit." The maximum current is measured when the cells are in full illumination and shorted. The red arrow indicator is the maximum power point at full illumination. In order to efficiently utilize this V/I combination at a different voltage, a switch mode power conversion will have to be used.

Another word about "open circuit voltage". This the maximum voltage possible from a panel. It is a design guide for Max. Voltage, for any devices down stream. Draw any amount of current the voltage drops to the nominal .5 volts/cell.

 

Typical MPPT voltage is about 83% of open circuit voltage. But this is not very important. Nominal 12V panel (rated 12V because it is good for charging 12V batteries) has Vmp around 17.5V and Voc around 21V, but the battery is charged at most at 15V, so, from the battery viewpoint, the panel is a constant current source and produces the same current as if the panel was shorted.

The panel connected directly to the battery will always push in its Isc regardless of the battery voltage.

 

The blue lines you put on the graph are incorrect. .

You claimed that the cell for which you posted the I/V characteristic is like a "constant current" source. If it is a constant-current source as you claim, its I/V curve would fall along the blue line, which is a constant current source of 0.38A with a voltage limit of 0.68V. The full insolation line (red line) deviates from the blue line, especially for output voltages greater than 0.4V.

For charging batteries this particular cell should be operated at ~0.4V per cell, so it would take 14/0.4 = 35 cells to make the panel...

 

The panel connected directly to the battery will always push in its Isc regardless of the battery voltage.

Looking at Les's cell curve, the ideal operating point is at 0.35A while the short circuit current is ~0.38A

 

Looking at Les's cell curve, the ideal operating point is at 0.35A while the short circuit current is ~0.38A

Yes, but battery will never get to MPPT voltage - it's be always left of that point.

 

Yes, but battery will never get to MPPT voltage - it's be always left of that point.

Exactly. In order to get the most power out of the panel, the battery voltage has to go up to 18 volts, which the battery will not allow.

 

Yes, but battery will never get to MPPT voltage - it's be always left of that point.

No, that depends on how many cells are in the panel. If the panel is made of 35 of the cells for which Les posted the I/V characteristic, if used for charging a Lead-Acid 6cell battery, the panel will be operated at or very close to the MPPT voltage. You guys keep throwing out 18V, but where are you getting that value? It is not supported by the I/V curve that Les posted, nor my own experience...!

 

It is probably this one:
http://www.ebay.com/itm/20Watt-12V-...559?pt=LH_DefaultDomain_0&hash=item3f28efeb07
17.9v at mppt minus a diode drop say .4 volts = 17.5 volts at 1.11 amps 19.4 watts @90% for supply -17.46 watts. Without mmpt 14.5v@1.11A - 16.1 watts. About 9%. If the load isn't on it's just making bubbles.

 

Mike, you were close. Typical solar panels are made up of 36 cells. 36 X .5 volts = 18 volts max (full illumination, max power point). This voltage changes over temperature. Some people say 17 volts, some people say 18 volts. It's the same panel.

 

Mike, I can feel your frustration over this, so I am going to walk through some numbers.

Under full illumination: as stated the panel puts out 18 V at 1.1 amps.

- put a 16.3 ohm resistor on the output, the current will be 1.1 amps and it will dissipate 19.8 watts. The resistor will see 18 volts.

- if you put a 12.3 ohm resistor on the output, the voltage will drop to 13.6 volts and the current will be 1.1 amps and it will dissipate 15 watts.

- you could put a 5.7 ohm resistor on the output, the voltage will drop to 6.3 volts and the current will be 1.1 amps and it will dissipate 6.9 watts.

Nothing changed except the load resistance.

In the above descriptions the currents, 1.1 amps, are all the same. This is because the illumination is the same. So, the illumination determines the current output.

The load resistance is different in each case, which yields a different voltage. So, the load resistance, based on the current output, determines the voltage output.

There are small variations due to voltage, but the above is pretty much right on.

 

You guys keep throwing out 18V, but where are you getting that value?

This is the value which comes handy for charging 12V batteries. You may need up to 15V to charge a battery. Plus FET/diode voltage drop. Plus allowances for Vmp being lower in less than full light and/or hot weather. So, smaller panels (designed to be coupled with 12V batteries) are done this way.

Bigger panels don't follow this pattern because they're designed to be used with MPPT controllers or with grid-tie inverters and voltage doesn't matter any more. In Europe, they allow solar arrays up to 1000V. To get this, you need lots of panels in series, and you can get your desired voltage +- 30V with any kind of individual panels.

 

The specs of the EBay panel that Ron posted are:
Voc = 22.41V
Vmppt=17.9V
Imppt=1.11A
Output@mppt=20W
Isc = 1.22A

I made a piecewise linear LTSpice model of this panel to fit the data above. To verify the model, I plot panel current vs panel voltage (Green Trace) and panel power output vs panel voltage (Red Trace).

Check I(V1) when V1=0 (short circuit), V(out) when I=0 (open circuit), and that peak power (20W) occurs at 17.9V and 1.1A. Does anybody dispute that this is a good model?

 

Looks good to me Mike.

 

I agree, it is right on.

 

Someone explain it to me, the point (green) when (20v) I=0.8A?
thanks

 

Someone explain it to me, the point (green) when (20v) I=0.8A?
thanks

The Green trace shows how much current (Left Y-axis, units of A) the panel puts out as a function of the voltage across the panel (X-axis, units of V). If asked to deliver 20V into some load, the panel would put out ~750mA. If charging a battery whose voltage is 12.0V, the panel delivers 1.2A.

The right Y-axis is for Panel Output Power delivered vs voltage, units of Watts.

 

If charging a battery whose voltage is 12.0V, the panel delivers 1.2A.

What if i put simple lamp? or Led with resistor?

 

Using the panel model shown above, here is a comparison of how long it takes to charge a huge capacitor from 12V to 14.5V , first using a simulated MPPT controller and second, just connecting the panel to the battery. Using a 25,000Farad capacitor as a proxy for the battery means that at a charging current of ~1A, it will take about 60Ksec (17h) to raise the voltage from 12V to 14.5V. (i.e like charging a 17A-hr battery).

I am using a behavioral current source to simulate a MPPT charge controller. The expression that controls the current source is I=0.9*20/V(batm). The 0.9 factor comes from the MPPT being at best 90% efficient. The 20 is from the maximum of 20Watts the panel can put out if loaded at the MPPT point. The divisor is the battery voltage, so I=P/E, which means that the current into the battery automatically decreases as the voltage increases.

Look at the Red and Green traces, which are the voltage vs time across the two simulated batteries. Note that it takes the MPPT controller 46ks to charge the simulated battery from 12 to 14.5V, while the panel connected directly to the battery takes 52ks, which saves 6ks or the MPPT is about 10% faster. The Light Blue trace is the panel power going from 14.5W to 17.5W as the battery accumulates charge for the direct connection, while the Dark Blue is a constant 18W (90% of the 20W max for the panel).

This shows that for this example, the complexity and cost of using an MPPT controller to get a 10% reduction in charging time is pretty marginal.

btw- something else occurred to me. My earlier statement about not using the LM317 as a charge regulator is bogus. Placing the dropout voltage (~3V) of the LM317 between the panel and the battery will not effect the charge time because the available charging current doesn't change much with or without the added voltage drop of the regulator...