Hey guys, I am an senior EE student and my senior design project involves making an MPPT controller. I am the only one working on it so I am a bit overwhelmed. I have poured over countless DIY projects and papers but not one of them has had much of what I neeeded.

Its tough cause I have never had to do something as practical as this and I am also not very good at programming. So far, what i need to do is design something that can charge a 12V Deepcycle battery. I'm am looking at a 50W 12V panel that delivers 2.8Amps max. Since I am not very good at C++ and atm we are not allowed to use Arduino, I have the NI myRio with LabView.

For me I am not too concerned about making it suppply 4A or more. Right now I still don't know where to start. There are some simulations I have but they are based off of one group that didn't actually build it.

Where do I need to start? And if anybody has code they can share, that would mean a lot.

 

I shouldn't do this. But I will.

http://ww1.microchip.com/downloads/en/AppNotes/00001521A.pdf

 

I shouldn't do this. But I will.

http://ww1.microchip.com/downloads/en/AppNotes/00001521A.pdf

Thanks. just read it and read over it again. I had seen that before but I guess I glanced over it. Couple of questions:

What is the AN1521?
The circuit that was provided there didnt seem to specify the specs of the system such as current and voltage treshold.
The schematic doesn't explain the components/why they are picked and how the circuit performs when it comes to trackign and supply power to a load. There actually is no mention of a load.
It uses AN1467 for what it calls inverse SEPIC but it doesn't seem to be availble for purchase. I also have never heard of an inverse SEPIC circuit before
The circuit provided uses a PIC micontroller so I guess the code provided is based on C/C++ code
It appears the code has a way to check for load conditions if a battery is involved but it doesnt appear that on the hardware side, there is circuit to enable even substandard charging (as in preventing reverse current and the controller supplying too much voltage)

These are some quick things that stuck out to me but I need to reread and make sure I didn't miss anything

 

AN 1521, application note # 1521. To impliment this application would be a verry large task; might consider the less presise implimentation in paragraph, " FRACTIONAL OPEN CIRCUIT VOLTAGE" using a small slave panel for open ckt. V . You need to know solar panel, SP, open ckt V & short ckt current., & battery capacity. Would be a good idea to plot I vs V for your SP. Basic hardware required: SP. slave SP, SSP, voltage controlled buck converter, battery, & misc. comparators, OP amps pots etc. I used a similar approach with a one million BTU solar thermal sustem.

 

AN 1521, application note # 1521. To impliment this application would be a verry large task; might consider the less presise implimentation in paragraph, " FRACTIONAL OPEN CIRCUIT VOLTAGE" using a small slave panel for open ckt. V . You need to know solar panel, SP, open ckt V & short ckt current., & battery capacity. Would be a good idea to plot I vs V for your SP. Basic hardware required: SP. slave SP, SSP, voltage controlled buck converter, battery, & misc. comparators, OP amps pots etc. I used a similar approach with a one million BTU solar thermal sustem.

What do you mean a slave SP and how would I for testing purpose vary the load of the circuit?

 

Take a look at page 4. This is the easiest to implement (I think), and can be done with a buck/boost converter and some code.
Then you need to worry about the battery charge controller so you don't over charge it. If the paper is blank you can pick a battery and solar panel where this won't ever happen making the whole thing a lot easier. Do you get points for high efficiency?

 

Take a look at page 4. This is the easiest to implement (I think), and can be done with a buck/boost converter and some code.
Then you need to worry about the battery charge controller so you don't over charge it. If the paper is blank you can pick a battery and solar panel where this won't ever happen making the whole thing a lot easier. Do you get points for high efficiency?

what do you mean where the paper is blank? And no I don't get points for high efficiency

 

If you match the solar panel to the battery you can pick a solar panel that cannot give to much current to your battery. Then you don't have to build a circuit to limit the current.

http://batteryuniversity.com/learn/article/charging_the_lead_acid_battery

 

Blank paper is how a project starts.

Once you have a design your paper is no longer blank.

 

If you match the solar panel to the battery you can pick a solar panel that cannot give to much current to your battery. Then you don't have to build a circuit to limit the current.

I doubt the diode between a small solar cell and a battery would count as a MPPT controller suitable for a senior project.

 

If you match the solar panel to the battery you can pick a solar panel that cannot give to much current to your battery. Then you don't have to build a circuit to limit the current.

http://batteryuniversity.com/learn/article/charging_the_lead_acid_battery

Thanks I will look into that lead acid link.
My problem now is starting the thing. Despite all the research, I still simply do not know where to start

 

No, I doubt it.

What I mean is if he picks a solar panel with 2.8 amps maximum current and a 28 amp hour battery he doesn't need to build in any current limit in the charge controller. Makes it easier.

 

No, I doubt it.

What I mean is if he picks a solar panel with 2.8 amps maximum current and a 28 amp hour battery he doesn't need to build in any current limit in the charge controller. Makes it easier.

Thats actually what I picked. Except instead of 28, I went for a 35AH. What did you mean by a current limit?

 

If you look at the stuff on charging batteries they will tell you that you should not charge them with higher than C/10 so if you have a 35 amp hour battery you should not charge it with more than 3.5 amps. Since the most your panel can put out in full sun is 2.8 amps you don't need to worry about to much current. So what you need to do is move the output voltage up and down while measuring the current into the battery. If you move it up lets say and the current goes up, move it up some more. If it goes down move it down some more. You will find the maximum power point for the solar panel and bounce around it.

 

Ah right sorry I meant what circuit would provide the current limit? And do you know of a system I could use to do the simplified charger you guys have been suggesting? Circuit wise and controlwise. The AN attached to this forum is very unclear on its hardware circuit and it just lays the P&O code, no mention of how it is compiled

 

And i guess what I still need to figure out is how to build this thing step by step, building each component code and hardware and testing it. Despite all the googling and references, still have a hard time figuring it out. Any help in this regard will be very appreaciated

 

what circuit would provide the current limit?

Googling 'current limit schematic' and 'battery charger schematic' should find you plenty.

 

I think you are making it to hard.
Start with a block diagram with the panel on the left and the battery on the right. Then see what is needed in the middle.
Forget copying someone else's and just pick your favorite mico and language.

 

This is a SP project, battery is just a load & ronv has pretty well put the battery question to rest; just need to disconnect battery when V = 13.8.
For a load to test SP, a healthy rheostat would be handy, 100 ohms rated for 3 A, lacking one believe I would make an electronic load with N ch FET or NPN darlington transistor with large heat sink. Measurements can be quick so load does not need to be connected continusly. Maybe one V steps up to 12 V, 1/2 V steps up to OC V.
The SSP is just a small panel ( 6 V OC +- ) with same charistics as main SP & attached to SP. Using equation 1, V-MPP = k SSP- OC. k = ratio of SP to SSP OC's times .71. The .71 from SP plot of P vs V.
Now we connect a pot across SP- OC & adjust tap to = SSP- OC V, say 6 V; with a load on SP OC V no longer available, but is now available by measuring SSP * ratio. Add a pot across SSP , tap @ .71 OC V = 4.26 V. If we monitor SP pot tap & start loading SP, when V falls to 4.26 V, SP should be operating @ MPP.
A cloud passes over, 4.26 V from SSP drops , loading needs to be readjusted to re-establish balance.
Connecting SP to battery via voltage controlled boost-buck converter allows implimitation of automatic control. ???

 

This is a SP project, battery is just a load & ronv has pretty well put the battery question to rest; just need to disconnect battery when V = 13.8.
For a load to test SP, a healthy rheostat would be handy, 100 ohms rated for 3 A, lacking one believe I would make an electronic load with N ch FET or NPN darlington transistor with large heat sink. Measurements can be quick so load does not need to be connected continusly. Maybe one V steps up to 12 V, 1/2 V steps up to OC V.
The SSP is just a small panel ( 6 V OC +- ) with same charistics as main SP & attached to SP. Using equation 1, V-MPP = k SSP- OC. k = ratio of SP to SSP OC's times .71. The .71 from SP plot of P vs V.
Now we connect a pot across SP- OC & adjust tap to = SSP- OC V, say 6 V; with a load on SP OC V no longer available, but is now available by measuring SSP * ratio. Add a pot across SSP , tap @ .71 OC V = 4.26 V. If we monitor SP pot tap & start loading SP, when V falls to 4.26 V, SP should be operating @ MPP.
A cloud passes over, 4.26 V from SSP drops , loading needs to be readjusted to re-establish balance.
Connecting SP to battery via voltage controlled boost-buck converter allows implimitation of automatic control. ???

Man that sounds like a lot of good details but its hard for me to follow it.
For starters, what is SP and SSP?

 

Defined way back in post # 4. SP = solar panel, SSP slave SP. Re post # 17, just take some of this junk and make a block diagram & we'll go from there.