MPPT for photovoltaic application (using microcontroller)

Dear All,

I am planning to implement an MPPT algorithm for solar powered battery (Lead Acid, 12V, 7Ah) charger. The solar panel is a 50W one. At peak power it can give 18V and 2.78A. It has a Voc of of 21V and Isc of 3A. I have some querries regarding the design I plan to do, which are:

1. Which converter topology is suitable for this design? I was thinking of using non-inverting Buck-Boost topology as it can boost panel voltage when it is less than 14V and step it down when around 18V. However, I am not sure if at such low voltages (<14V), the panel can provide sufficient current output. Hence, would a buck converter be sufficient?

2. Since I am planning to charge a battery which has specific charging requirements like CC/CV/Float charge, I need to maintain certain parameters like voltage and current at the output. The only variable I have control on is the voltage at the O/P which depends upon the Duty cycle of the controller. For Eg: for a buck Vout = D*Vin. If I increase the voltage, more current is drawn by the battery and hence more power. Hence my control strategy would be focused on maintaining the voltage requirement for the battery. If this is the case, then how do I implement MPPT algorithm which also calls for a duty cycle adjustment based on the O/P power of the solar panel? Wouldn't there be 2 conflicting loops inside my micro-controller which would try to reduce and increase the Duty cycle?

3. How do I achieve Constant Current(CC) mode in a DC-DC converter (For Eg: Buck)? Is it by monitoring the output current and turning off the power switch when the current reaches a set threshold?

Thank You.

Regards,
Ashare

Your design will have to combine these two functions, while making them cooperate:

a) the MPPT circuitry, whose purpose is to push maximum current into the battery during bulk charge,

and

b) the charge control circuit, which tapers off the current into the battery as its volt level rises.

1. Which converter topology is suitable for this design? I was thinking of using non-inverting Buck-Boost topology as it can boost panel voltage when it is less than 14V and step it down when around 18V. However, I am not sure if at such low voltages (<14V), the panel can provide sufficient current output. Hence, would a buck converter be sufficient?

Click to expand...

Yes, a buck converter.

You probably will not benefit from a boost function. You would only get a few mA at light levels that are too dim to cause your PV panel to produce more than your battery V.

BradtheRad said:

Your design will have to combine these two functions, while making them cooperate:

a) the MPPT circuitry, whose purpose is to push maximum current into the battery during bulk charge,

and

b) the charge control circuit, which tapers off the current into the battery as its volt level rises.

Click to expand...

Thank you for your reply. With regard to the above response I have a few queries. Please find them below:

a) In the context of the bulk charge, what would be the internal impedance of the battery? [low, isn't it?] If low, does this mean I need to program the micro controller to limit the current to a safe value? This safe value of current will determine the upper limit of my max power drawn, isn't it?

b) how can I measure the battery's terminal voltage when my DC-DC converter is connected to it. Do I need a mechanism to poll the battery voltage at intervals disconnecting the converter O/P?

Thank You.

Regards,
ashare

a) In the context of the bulk charge, what would be the internal impedance of the battery? [low, isn't it?] If low, does this mean I need to program the micro controller to limit the current to a safe value? This safe value of current will determine the upper limit of my max power drawn, isn't it?

Click to expand...

If a battery is healthy then its internal impedance will be so low it is difficult to measure. On the order of a few hundredths of an ohm.

As to what is a safe charge rate, common advice says 1/10 of your A-Hr rating. Battery health will not suffer if charging continues after it is full. (But be careful with some varieties such as gel cells, they are not so tolerant.)

Your battery has 7 A-hrs capacity. A C/10 rate would come from a PV panel producing 0.7 A. Your battery would (theoretically) charge in 10 hours. This might be a day's charging session, if you get sufficient sunlight.

Your PV panel puts out 2.78 A, which is C/2.5. It will take just 2.5 hours (sunny day) to fully charge your 7 A-Hr. battery. A healthy lead-acid battery can absorb this rate, although you should monitor its condition so that the electrolyte does not start to bubble away (if you were to let battery volt level rise to 15 or 16V). Your controller needs to start tapering the current as the battery reaches full.

b) how can I measure the battery's terminal voltage when my DC-DC converter is connected to it. Do I need a mechanism to poll the battery voltage at intervals disconnecting the converter O/P?

Click to expand...

Some chargers use that method, by disconnecting current periodically, in order to measure battery voltage by itself. This is not so necessary with a slow charge rate, however.

The faster the charge rate, the more it tends to elevate the battery voltage.

Thank you for the elaborate answer. Kindly let me know if I need to implement any special constant current circuits or can I do current control by controlling the duty cycle of the controller?

Regards,
ashare

Kindly let me know if I need to implement any special constant current circuits or can I do current control by controlling the duty cycle of the controller?

Click to expand...

You can go either way.

I used the second method when building a charge controller for my PV system.

The first method restricts current flow to the battery. There would be some component heating up. The taper charge would require that I make a lot of tedious adjustments.

So I thought it is easier, and more efficient, to:

1) switch charge current on full,
2) until the battery reaches 14.4 V,
3) then stop charging, and
4) let the battery drift back down a volt,
5) then charge it again at full current to 14.4V.
Etc.