MPPT Using Boost Converter

[9651]

Hi,
Sorry if this comes off as a very obvious/dumb question.
Im trying to implement a complete grid tied PV system.
I'm facing some trouble in the Boost Converter part. I want to implement the Perturb and Observe method and that I can do. My issue is that I cannot understand how I am supposed to decide the values for the Inductor, Capacitor and Resistor in the boost converter.

All the equations I have seen involve the duty ratio, D. But with the P&O method D keeps changing so how do I decide the values best for my circuit?

 

[BradtheRad]

It is reasonable to start with a frequency of 20 kHz. This is above the range of human hearing, so it is less likely to make distracting noise.

The Henry value, together with the resistance in the first half of the cycle,
will create a time constant (L/R). You need to choose values so that amperage can reach maximum in the amount of 'On' time in the first half of the cycle.

A Henry value of 1mH is commonly available, and not too expensive. Maybe 500 uH. The value is flexible, since your controller will change frequency and duty cycle, to reach optimum performance.

Notice that the more current a coil carries, the larger it will be and more expensive it will be. A smaller Henry value becomes attractive in those cases.

 

[9651]

Thanks! It makes sense, but then what about the resistor?

 

[D.A.(Tony)Stewart]

MPPT profiles often have very flat peaks so the optimal point is not critical, but if you introduce error or hysteresis, opt towards lower current and higher voltage or towards the optimal fixed threshold estimate of 76% Voc.

There will be more losses to worry about with surface reception angle error, dust absorption, surface reflection and self heating.

 

[nikhilmahasvar]

Do you mean the shunt resistor ?
Have you made the frame work to do the grid synchronization since that is the toughest part in grid tied systems.

 

[BradtheRad]

Thanks! It makes sense, but then what about the resistor?

Perhaps you mean the current-sensing resistor? This develops a voltage across it, in proportion to the current going through it.

It should be as low an ohm value, as will give you a usable volt reading (maybe 1/10 or 1/20 V).

Example, if current is 10 A, then a 0.01 ohm resistor will yield a 1/10 V reading. It will dissipate 1 W of power as heat.

If the current is high, then you might as well look into hall effect sensors, rather than resistors.