MPPT

[PG1995]

Hi

Could you please help me with ? Thank you.

Regards
PG

Attachment, "mppt1", deleted by PG1995 temporarily for some reason. Will upload it later.

Note: This thread also contains relevant material from post #13 to post #18.

 

[KeepItSimpleStupid]

The sentence beginning with "Batteries like to be charged..." is utter GARBAGE. It depends on the chemistry, Have at it: www.batteryuniversity.com

OK, I'll tame this statement a bit. The batteries used in a solar system are lead acid. Your article, I think, missed that point all together.

 

[ronsimpson]

A 12 V battery will need 50A to reach 20V.
1) The panel can not make that much power.
2) This will result in the battery boiling over and exploding.

Adding a resistor wast power.

In a low cost version, connect the panel and battery and get what you get.

In a good version, insert a switching power supply and micro that runs the panel at the (voltage & current) where you can get the most power out AND charge the battery where it works best.

 

[KeepItSimpleStupid]

MPPT:

Browse some specs. You first will find that they have to use as it's input current at the short circuit current of the panel and be able to handle the open circuit voltage of the panel. So that's a given part of the design and selection process, so in essence it's told.

I was involved in instrumenting a "demonstration system" for energy demand side management research that "never worked". Since it was a solar demo system, it just had to look pretty. Roof presence of panels was nearly enough except for the tours. The "house" was initially "built" to do basic research for thermal storage etc.

So, the inverter would never start because the open circuit voltage was too high for the inverter. In any event, they had to add and reconfigure the panels.

The MMPT generally has to "disturb" the voltage operating point to figure out which way it needs to go.

I think some of the feedback paths are missing in your pic, but in essence, if you read some of the battery university lead acid charging algorithms, it has to implement that.

The "simplest" algorithm is "float charging" which would be used in say an exit sign.

The solar charger needs to maximize battery life and charge the batteries before the sun goes away.

 

[NorthGuy]

The diagram looks measleading to me. It makes it appear that MPPT is a certain hardware block that power goes through. MPPT is rather a method of controlling DC-DC converter. Also DC-AC followed by AC-DC converter is usually not the most efficient way to convert the power.

Look at the buck DC-DC converter. They have a decent explanation on how it works. Now, instead of the input voltage source, connect a solar panel and a capacitor in parallel. Connect the battery as the load. When the battery is connected, the output capacitor is not really needed, so you may remove it.

The MPPT controller is concerned about two entities - Vs - voltage at the panel (and capacitor), and Vb - voltage at the battery.

See what happens when the switch is off. Panel is charging the capacitor, so Vs increases (and will increase until it reaches panel open cirtcuit voltage - Voc). The current into pattery (if there was any stored in the inductor) decreases and as a result Vb decreases.

Now when you turn the switch on. Battery strat dischaging the capacitor, so Vs decreases, but the current through the inductor into the battery increases and then Vb increases too.

Battery is usually charged in four stages.

1. Bulk stage occurs at the beginning when battery is discharged.

The goal of the MPPT controller is to get as much energy as possible. To do this, it must keep Vs as close to the optimum MP point as possible. We will set aside how this point is found (In the simplest form, you just use the panel specification). The MP point occurs at certain voltage called Vmp. Total power is equal to V x I. Vs higher than Vmp will decrease production because of lower current. Vs lower than Vmp will decrease production because of lower voltage.

So, the goal of MPPT controller is to keep Vs as close to Vmp as possible. This can be done by manipulating the switch. When Vs > Vmp, you turn the switch on. When Vs < Vmp, you turn the switch off. Of course, you use some hysteresis.

2. Absorption stage occurs later when battery voltage reaches pre-set Vabs voltage.

The goal of the MPPT controller is not to let Vb rise above Vabs, which would overcharge, or could even destroy the battery. This can be done by manipulating the same switch. When Vb < Vabs, you turn the switch on. When Vb > Vabs, you turn the swith off. Hysteresis is needed too. While you're doing this, Vs naturally slides up above Vmp limiting the production to what is needed.

3. Float stage occurs when battery is fully charged.

You could turn the charger off completely, but there could be loads on the batteries (inverter, lights etc.), and it's better to keep the charger on to support these loads instead of discharging the batteries. This is done by maintaining batteries at a constant Vfloat voltage, which is slightly higher than resting voltage for fully charged batteries.

The goal of MPPT controller is to keep Vb as close to Vf as it can. When Vb < Vfloat, you turn the switch on. When Vb > Vfloat, you turn the swith off. Hysteresis ...

4. Off stage occurs during the night.

Panels must be disconnected from the controller to avoid draining batteries through the panels. Usually a separate relay is used.

Combining stages.

The three day stages can work using the same algorithm. You have a maximum voltage Vmax, which is set to Vabs during bulk and absorption stages, and to Vfloat during the float stage. If ((Vs > Vmp) AND (Vb < Vmax)) you turn the switch on, otherwise, you turn it off. In the simplest form, this can be done with two comparators and a logic gate.

Of course, there are more complicated MPPT controllers, but the concept is the same.

 

[MrAl]

Hi there,

You certainly came to the right place to ask these questions and i'll tell you why in a minute <smile+wink>.

The main idea of the text is about the max power point tracking, not what is the best converter choice. There are various converter choices which are generalized in the text using a rather complex scheme which of course does not have to actually be implemented. It depends on various factors like array voltage and battery voltage.

It should be obvious that you can not charge a 12v battery with a 20v source unless the source is current limited and the maximum voltage of the battery is observed. The function of the converter (regardless of type) is to supply a voltage that more closely matches the battery voltage so that maximum power can be drawn from the solar array in order to charge the battery most efficiently.

If the solar array max power point voltage is 20v for example, then we want to draw power with the array at 20v, not at 12v. We could probably get away with connecting a 12v battery to the array directly if it can handle the max current of the array, but then what would happen is we would be drawing more current from the array then we should in order to keep the 20v array voltage, so we would end up getting reduced power from the array, and that is considered inefficient. Ideally, we want to draw power at 20v but still charge the battery at 12v (or closer to 14v actually), and we want to be able to transfer as much energy as possible.

For example, if we had a 1 square meter array that was 100 percent efficient and received 1kw of energy from the sun on it's surface and put out 1kw at 20v, then at 12v it might only be able to put out 500 watts so we'd be loosing 50 percent of the power available from the sun with that surface area. So this is where max power point tracking comes in.

One method is to perturb the output power. With the output powering say a load at 10 amps and 14v, that means we are drawing 140 watts from the array (assuming a 100 percent converter efficiency for simplicity). We then try to increase the output current (so we can charge faster and thus obtain more energy for storage in the battery). Now if the output current increases, that means we got more power so we were not originally at the max power point yet. But if the current goes down, that means the input to the converter dropped and so we might be beyond the max power point, so then we go back to the original current setting. If the current did go up then we try again and again, until the current goes down, and then go back to the previous current set point, and that's the max power point (MPP).
Another way to measure this is to measure the actual array voltage and current. By perturbing the output power we measure the voltage of the array and the current out of the array and simply multiply the two together and that gives us the power the array is putting out. If the calculated power goes up then we are not yet there, but if it goes down then we are beyond the MPP so we have to back up.
One disadvantage is that the output power varies somewhat while we are testing for the MPP. Another disadvantage is changing cloud cover can easily interfere with the readings we obtain because a cloud moving overhead can change the output current just like our perturbation did, so the circuit may get mixed up and cause a temporary brownout or worse oscillation between brownouts and regular operation.

Another sometimes preferred method is to use a dedicated secondary solar cell to act as a solar reference cell. This cell is used only to measure the available sunlight power. The system relies on information from this cell alone to set the MPP and it does so by correlating the cell's readings to the MPP of the bigger power array. This is preferred especially when output oscillation is unacceptable.

So what is inside that first block is a measurement and control system that measures the solar panel parameters and changes the output power to help keep the panel at the MPP, or else it uses a reference cell to help determine the best output power to keep the panel at the MPP.

Another simplifed view of the max power point tracking and power converter is that of a variable transformer (except it is DC) that can match the input to the output perfectly for any level of sunlight.

In the beginning i mentioned that you came to the right place to ask this question. That's because i happen to be one of the people who designed and redesigned max power point tracking circuits, both analog and digital types when i worked in the industry. At one time we used a huge million dollar array panel to power line tied synthesized sine wave converters. It was interesting at the time.
Eventually Sandia Labs bought a lot of these converters and seemed to favor the reference solar cell method for max power point tracking over the perturbed method.

 

[KeepItSimpleStupid]

I worked in the industry too, but the solar cells never made it to the array stage. Pure basic research. MPPT is affected by cloud cover as was stated, but there is also another MPPT tracking system which has to do with pure position. The MPPT point does vary with intensity.

One presentation I went too about >1 MW systems was interesting because these can be used for local power factor correction, so the inverters could supply current just to correct the power factor.

 

[PG1995]

Thank you, KISS, ronsimpson, NG, MrAl, for your help.

Q1:

A 12 V battery will need 50A to reach 20V.
1) The panel can not make that much power.
2) This will result in the battery boiling over and exploding.

Do you mean to say that a 12 V battery would need 50 A of current to reach 20 V if it was directly connected to 20 V panel which can only supply 8 A of current.

Could someone please help me with Q2 and Q3?

Please try to keep it simple because I need to understand all this soon. Thanks a lot.

Regards
PG

 

[KeepItSimpleStupid]

Q2: The service to your house is 200 A split phase; your toaster, the only load, is using 15A , where is the rest of the power going? Your car battery can supply 400 A @ 12 V, where is the rest of the power going?

The lake is full of 100,000 of water, I'm using 1 gal/min, where is the other water going?

Think!

 

[PG1995]

But my Q2 is not about the current, it's about voltage drop. In your example of toaster, the toaster will limit the flow of current to only 15 A and the voltage drop it will depend on your line voltage.

Please someone also help me with other queries. Thanks.

Regards
PG

 

[NorthGuy]

Q2. There's no voltage drop. If you connect panel directly to the battery, they, obviously, will be at the same voltage. The problem is that you could've used 20V and would get nearly the same current from the panel. But you used 14V, so you unrealized (8x20) - (8-14) = 48W.

Q3. You got it right. However DC->AC->DC path may not be the most efficient way. DC->DC is very common.

 

[PG1995]

Thanks, NG.

Q2. There's no voltage drop. If you connect panel directly to the battery, they, obviously, will be at the same voltage. The problem is that you could've used 20V and would get nearly the same current from the panel. But you used 14V, so you unrealized (8x20) - (8-14) = 48W.

My apologies if I'm sounding too dumb. But many a time I might sound dumber than this! If I have connected 20 V, 8 A panel directly to the battery then why are only 14 V being used? Suppose, 4 ohm resistor (it needs 4V/A) is connected to 20 V then it can ideally draw 5A current. But if the source can only supply 2A then the power is 20x2=40W. Where am I going wrong? Thanks.

Regards
PG

 

[KeepItSimpleStupid]

Q2: The operating point of the panel shifts or the panel can supply more power but the load (charging the battery) requires a certain amount of power. just like the toaster only needs 15*120 and the service can supply 200*240 W. The utility shifts the delivered power.

Neuclear provides a nearly constant amount of power. Wind, you take what you can get. Fossel fuel plants act like a regulator. Solar can be used to supply power or locally regulate the power factor. Hydro you can also throttle, just not as fast.

Some "free" power you can store. Pump water to a higher elevation or batteries.

Demand side management - the power company turning off your AC, alows the power company to delay starting another plant or negates the need to build one.

 

[NorthGuy]

If you look at the I-V curve, you'll see that panels are current sources - they produce roughly the same current under wide variety of voltages. Batteries, on the other hand, are mostly voltage sources - they maintain roughly the same voltage under wide variety of currents. So, when you connect them together you get the panel's current and battery's voltage.

When you connect battery to the panel, panel starts producing current. If battery is discharged, then the current pumped into the battery elevates its voltage only by a little. As the battery is nearing the full charge, the same current will elevate battery voltage more, so, at some point, you will have to limit it to protect the battery.

 

[KeepItSimpleStupid]

You can't connect a 20V 8A panel direct to the battery. The battery could explode without a charge controller.

in any event, there needs to be a blocking diode.

PG, why did you switch from apples (controlled system) to oranges (uncontrolled)?

 

[NorthGuy]

You can't connect a 20V 8A panel direct to the battery. The battery could explode without a charge controller.

It depends on the battery.

For 12V battery that can handle 8A, 20V/8A panel will not be harmful. Moreover, 1000V/8A solar array will have exactly the same effect on 12V battery as 20V/8A panel. Solar array cannot produce more than its rated Isc current even if you short it. And the voltage will come down as dictated by the load.

If you connect a 125 Ohm resistor to 1000V/8A array, you get 1000V and 8A (8kW), which has a good chance to evaporate the resistor quickly.

Direct short will produce the same 8A, but the voltage may be in uV - you will not even get any reasonable heat.

 

[KeepItSimpleStupid]

NG:

Describing a solar cell as a current source is absoluely correct. That's what the model is. A current source in parallel with a high value resistor. That in series with a low value resistor.

I should know better.

 

[MrAl]

Thank you, KISS, ronsimpson, NG, MrAl, for your help.

Q1:


Do you mean to say that a 12 V battery would need 50 A of current to reach 20 V if it was directly connected to 20 V panel which can only supply 8 A of current.

Could someone please help me with Q2 and Q3?

Please try to keep it simple because I need to understand all this soon. Thanks a lot.

Regards
PG

Hi,

You do not connect a 12v battery to a 20v source unless it is current limited. A 20v source would keep charging the battery and eventually destroy it. But a photovoltaic array is not a voltage source, it is more like a current source actually (but not exactly because the voltage varies too). It is a special kind of source.
It's not a battery, and it's not like the line voltage when you plug something in. It's different than that.

One very simplified view would be a current source in parallel with a resistance. When we draw no current from it the voltage is maximum because V=I*R and I is constant and R is constant, so V comes from those two, and the voltage here is the open circuit voltage. When we connect a load however even if that load is a battery, that draws some current and so the equation now looks more like this:
V=(I-iLoad)*R

So you can see that when we draw load current we take current away from the array and so the voltage across the array goes down. Keep in mind that this is a very rough linear approximation to the array but it works very similar to this in principle.

If that is still a little hard to grasp, then think of it as a voltage source with a series resistance. As we draw current from this non ideal source the series resistance drops some voltage so the external voltage measured across the source (after the resistor) is lower than before. The effect can be drastic because for example say we have a 10 ohm series resistor and a 10 ohm load, then that means the external voltage measured will only be one half what it was before we connected the load.

When we connect a battery of lower voltage than the open circuit voltage of the array a similar thing happens. We end up forcing the voltage of the array to go lower than it was before. If the array voltage was 20 volts and we connect a 12v high current battery then the array gets pulled down to 12 volts, and the current is whatever the array can put out at 12v. But if it's MPP voltage is at 15 volts, then we are not getting the full power that is available from the array. If the MPP voltage is 12v though then we are getting the max power that we can get with that particular level of sunlight at the time.

So to determine if the load is absorbing all the power that the array can possibly put out we need to know the max power point voltage for that level of sunlight, and if the load forces the array to that voltage then we are getting the maximum power available. If it is not at that voltage, then we are not getting the full available power from the array.

Another way of putting it is that we want to LOAD the array with the best possible load in order to get the maximum power available from the array with that given level of sunlight.

 

[NorthGuy]

One very simplified view would be a current source in parallel with a resistance. When we draw no current from it the voltage is maximum because V=I*R and I is constant and R is constant, so V comes from those two, and the voltage here is the open circuit voltage. When we connect a load however even if that load is a battery, that draws some current and so the equation now looks more like this:
V=(I-iLoad)*R

This equation describes a straight line with negative slope. If you look at an actual I-V curve for a solar panel, for example the one presented by OP in the first post, it doesn't really look as a straight falling line. Rather, it goes absolutely flat until it reaches relatively high voltages, then drops down rather quickly in a very curvy way.

 

[PG1995]

Thanks a lot, KISS, NG, MrAl, for your help.

In next of couple of days I will read about MPPT whenever I get free time to get general understanding of its working.

One last query for now. Suppose, I'm using constant voltage algorithm for a buck MPPT. Then, I decide to use another algorithm such as perturb and observe, would this mean that I need to make changes to the circuit too? Or can I just load a new software into the microcontroller and keep using previous circuit. Thanks.

Regards
PG