This is my vacation time,thought of building a solar night lamp.I used arduino for mppt tracking (P&O method).Now that i am half way through,i burnt my arduino and i think investing on a new arduino is waste of money,i decided to implement mppt without it.I just wanna get an idea,how it can be done and the difficulty involved.Solar panel-20W,Battery -6v,4.5AH.

 

So you want to do Maximum Power Point Tracking without an MCU? Perturb and Observe method?

You want to do something fairly technical without an idea of how it can be done, and prefer the limitation of working without an MCU, which is very possible but not trivial or necessarily a logical move.

I doubt you would enjoy building and tuning a fully analog implementation. (It can make grown men cry.) There are reasons why digital is becoming more common every day, and a reason why MCUs are becoming a default part of so many digital solutions. Software versatility beats complex hardware.

I would counter that you want to get a stock of Atmel chips and program them all with the Arduino bootloader.

In most cases your arduino can recover from being burnt by replacing the MCU (if it is a socketed UNO model) or the Voltage Regulator.

To protect you Arduino board in the future, just program a chip and protoboard it.

 

Yeah! fully analog implementation would be a wild goose chase,but how about using ADCs to convert voltage and current measurements to digital and multiply them using a digital multiplier and convert back to analog using a DAC,depending on which duty cycle is selected for the MOSFET? Am i on the right track or this can be made simpler?

 

No need to do that, you can easily muliply in the analog domain. But the question is what you want to do with the result.

 

Like i said before,decide the duty cycle for the MOSFET that i am gonna use in the DC-DC converter (in this case buck converter)

 

Well, obviously. But how exactly are you going to decide which way and how much you want the duty cycle adjusted?

 

I would take samples[VnxIn=Pn] at regular intervals and compare it with the previous samples[P(n-1)] to decide whether duty cycle has to increased or decreased,this cycle continues until maximum power is obtained.

Expounded by the attached flowchart!

 

You seem clear on how the duty cycle will be modified. Bonus points for using a state diagram.

But...

I would take samples[VnxIn=Pn] at regular intervals and compare it with the previous samples[P(n-1)] to decide whether duty cycle has to increased or decreased,this cycle continues until maximum power is obtained.

The description of the ways that you intend to do this without an MCU just convince me more that you really want to use an MCU.

You started out using an Arduino and that was a smart choice. This is a learning exercise but you ran into the standard problem of finding things you need to know about before you can learn the thing you wanted to learn. That is what I get from this.

Can I ask how you burnt the Arduino?

Was it the connection to the sense inputs?
Was it mismatched power and ground connections?

I really think that is where you should focus first. It is bad practice to leave any problems and breaks unexamined. There are sayings about doomed to repeat. Wiring problems can happen in any circuit and, with or without an MCU, they can be just as frustrating.

hidden text
If it is an issue of the cost of the Arduino, then you can switch to cheaper MCU's. Instead of replacing the Arduino you might want to get a PICKit and a bunch of cheap chips.

 

Is there some compelling reason this thread was moved to an area to discuss techniques the OP is specifically asking not be used?

...i think investing on a new arduino is waste of money,i decided to implement mppt without it.

 

He doesn't need to use an Arduino but the sampling and tracking described defenitely sounds like an MCU job. Any MCU.

Using a discrete ADC and some type of buffer chip for a memory - it can be done without an MCU.

Using a sample and hold with OP amps and going fully analog can be done.

It will work more easily with an MCU.

 

Sorry for the late reply!
Ironically.....i bought an arduino mega for my project,but i would like to implement it with a sequential circuit specifically designed for tracking mppt.

Can the duty cycle be decided on the irradiance level?

Afterall with all the capabilities that a microprocessor is designed for,using it only for tracking mppt is definitely under utilization of hardware.

 

for,using it only for tracking mppt is definitely under utilization of hardware.

There is something just completely wrong with that statement.

 

A microprocessor is the cheapest and simplest way to do the MPPT function. Trying to do it any other way may be an interesting academic exercise but is not a practical approach and a waste of effort.

 

Afterall with all the capabilities that a microprocessor is designed for,using it only for tracking mppt is definitely under utilization of hardware.

Who told you that to utilize a micro you need to use ALL the features it has? Most of the time this is not even possible.
Microcontrollers are made to be verstaile enough so that you dont need to look different chip for each project - sometimes you use the ADC, sometimes you use all of the I/O ports. You want to use the same chip all the time, one that has everything you would ever need, and not worry about chosing one that has the combination of peripherals you need for todays project.

Yes for professional use you would want to find a cheapset thing that can do what you want, but for a hobbyist there is no shame in using the same (even though it might be overkill) processor for anything that you encounter, as long as it fits on the board.

 

If i were not clear,i meant that i used only two analog ports and a single digital port out of the 40 digital ports and 10 analog ports that an arduino mega has.

Anyways,this thread has been posted in the aim of understanding how mppt can be implemented without a microprocessor.(Ofcourse i wanna LEARN how it can be done,not on any PRACTICAL APPLICATION grounds)

 

Look up EDN, December 5, 2008, page 51, " Solar-array controller needs no multiplier to maximize power". It is still there- just looked.

 

If i were not clear,i meant that i used only two analog ports and a single digital port out of the 40 digital ports and 10 analog ports that an arduino mega has.

Anyways,this thread has been posted in the aim of understanding how mppt can be implemented without a microprocessor.(Ofcourse i wanna LEARN how it can be done,not on any PRACTICAL APPLICATION grounds)

Then perhaps you with to choose a different device then an Arduino for the final device. Until then, there is much good work you can accomplish using that platform.

 

For simple battery charging, just buy a panel that has an open-circuit voltage about twice the nominal battery voltage. Done!

The "reduction" in charging time by MPPT tracking to control Pulse Mode charging is just not worth the effort.

 

Just use a SMPS (buck etc) set to regulate INPUT voltage.

The solar panels MPP is always at the same input voltage, generally about 17v for a 12v panel.

You don't need any peturb/observe or other fancy system, just adjust the PWM to keep the input voltage at the correct amount. You can do it with any standard bcuk regulator IC and a transistor to invert the voltage Fb pin.

 

Never thought about that.