Tim Nolan's Arduino MPPT - now with LCD

[kibi]

I have taken some interest in Tim Nolan's MPPT solar charge controller based on an Arduino.

Please refer to Tim's website below:

http://www.timnolan.com/index.php?page=arduino-ppt-solar-charger

The one problem I did have with his design was that it didn't support an LCD.
I have rectified this problem and modified his original code to support an LCD. Upon checking his circuit digram and I/O usage, I have determined that there are still enough I/O's on a standard UNO to drive a parallel LCD which saves a lot of cost. Serial or I2C displays are really cool, but come at a monetary cost.
I have got the code going utilising preset resistors to simulate the feedback from a live system to display certain parameters as below.



Now, I ran out of space with a 16x2 LCD. My requirements are to display more information about the charger's status. I have a 20x4 LCD on order, so when it arrives and I figure out the code, I'll post an update.

Unfortunately, I'm reluctant to share the modified code without permission from Tim, so once I've got the code going how I think it should be, I'll request his permission to publicise it.

Also, I don't plan on sacrificing an Arduino board for the final project, so I am in the process of designing a PCB around the ATMAGA328 which can be programmed in the Arduino board an then the MCU can be simply transferred to a dedicated PCB.

[MJLorton]

You are brilliant! Tim Nolan's site and MPPT project are great.

I'll certainly be following your progress with a keen eye as this is one of the Arduino projects I know I'm going to really enjoy.

Thanks for sharing good man.

P.S. the Arduino Nano can be used for putting Arduino projects into "production" without sacrificing your main project board:
http://arduino.cc/en/Main/ArduinoBoardNano

Martin.

[kibi]

OK, so I have made some more progress with this project.
The 20x4 display arrived and I have got it to display more useful information in a more easy to read manner.



I'm quite pleased with the way it works now.
I am now busy designing the PCB, well I have designed it, but looking at it, I've decided its crap. I'm going to have to start again, but that's just the way these things go.
Idea's of networking the charger have popped into my head, but I'm not sure if I have enough I/O's left on the MEGA328. It would be cool to have remote monitoring and data logging ability.

[MJLorton]

OK, so I have made some more progress with this project.
The 20x4 display arrived and I have got it to display more useful information in a more easy to read manner.

I'm quite pleased with the way it works now.
I am now busy designing the PCB, well I have designed it, but looking at it, I've decided its crap. I'm going to have to start again, but that's just the way these things go.
Idea's of networking the charger have popped into my head, but I'm not sure if I have enough I/O's left on the MEGA328. It would be cool to have remote monitoring and data logging ability.

Great update, keep them coming good man!
Remote monitoring and data logging...yes, yes, yes...I'd purchase one!

Cheers,
Martin.

[kibi]

I finished re-designing the PCB and I have fabricated a prototype. Considering that I last made a PCB back in the space year 1996, I am reasonably pleased with the result. I only went single sided because there is less to go wrong. It's not very neat, but it works.
I made a giant error, well two actually, but the first big mistake was that I forgot to include the ICSP header. I have soldered one on temporarily so that the code can be tweaked.
The second big mistake I discovered during testing. The AMPS readout was five times lower than the actual measured current. Like a klutz, I did not read the data sheet for the MAX4173. It turns out that there are three versions of this device. The part number is suffixed with T, F or H. I didn't pay too much attention to this fact when I ordered the chip from Mouser and just clicked on some random one which turned out to be the T version. The T version has a gain of 20, the F has a gain of 50 and the H has a gain of 100.
The design uses a H version with a gain of 100 and my project has ended up with a gain of 20, hence the reason why the readout was 5 times less.
With my bodged on ICSP lead I tweaked the code to get the readings correct.
I have not installed any of the power devices yet until I'm sure that the code is running properly.

Bottom side of the board with the ICSP lead soldered on temporarily.


Top


Top, including the display.


Lessons learned: READ the data sheet! Don't forget to include facilities to alter the code, the code WILL need to be perfected or modified.

[MJLorton]

This is brilliant, thanks so much for sharing the progress AND the mistakes. A great learning experience for all of us getting into this. I chuckled at the comment about making the PCB back in "space year 1996". I'm going to have to go through the same process myself for the multimeter reference check once I move from a prototyping board....I think it was back in 1989 / 1990 when I last did a PCB design on a PC and had the PCB made....

Cheers,
Martin.

[kibi]

I have installed the MOSFETs and the MOSFET driver and powered it up.
Alas, no go. The MCU is trying to switch the MOSFETs on, but they won't turn on.

It turns out that the IR2104 MOSFET driver that I ordered is not an IR2104. It's some TI part that I can't even find any information on.
So now, unfortunately, the project is on hold until the correct parts arrive.

[kibi]

The IR2104 arrived from America this morning, Mouser are really quick.
I have installed it and everything is running well.
In also got the MAX4137H. It is necessary to have this higher gain version because the lower voltage presented to the ADC is too low with the low gain version. This means that the resolution of the measurements is not very good. It not only provides low resolution readings on the display, but the calculations are made with poor quality data.
I'll replace the MAX4173 at the weekend perhaps.
The display updates too quickly to be able to make legible readings. I'll have to figure out how to send updates to the display less frequently.
Otherwise, I'm pleased that it's working.

The sub-assembly with heat sink. For the eagle eyed, the TI part is still on the board at this time.


The fully assembled device.


The unit in operation, charging an actual battery.


After that, I'll start work on getting network or logging or even better, both working for "version 2".
I'd also like to get some more power transfer out of this thing. I'd have to look into higher power MOSFET's with otherwise similar characteristics to the IRFZ44. It'd need a more staunch inductor too. However, there is not point in upgrading the power if I don't even have one solar panel yet

[Tooms]

Hi

Is this MPPT only supporting 12v system or can it be used at 60V or 48V system ?


Thomas

[kibi]

Hi

Is this MPPT only supporting 12v system or can it be used at 60V or 48V system ?


Thomas

Soory for the delayed response.
I don't see any huge problems with operating this at higher voltages. 24v systems will just require the voltage dividers for the voltage sense inputs to be adjusted appropriately and the code modified.
48v and 60v systems are slightly more tricky because the 7805 regulator can't cope. A 5v supply will have to be arranged to power the controller and then the voltage dividers and code can be modified to suit.

[kibi]

It would seem that you need to read the code
At the top of Tim Nolan's code it says that his code is free to share as long as you give mention to him and link to his website.
Tim has been mentioned, so here is his website - www.timnolan.com

Attached is the code that I have modified.

[s3n]

OK, so I have made some more progress with this project.
The 20x4 display arrived and I have got it to display more useful information in a more easy to read manner.



I'm quite pleased with the way it works now.
I am now busy designing the PCB, well I have designed it, but looking at it, I've decided its crap. I'm going to have to start again, but that's just the way these things go.
Idea's of networking the charger have popped into my head, but I'm not sure if I have enough I/O's left on the MEGA328. It would be cool to have remote monitoring and data logging ability.

can u share or give the schematics of the MPPT w/LCD using MEGA328?

[kibi]

[kibi]

I have modified the code so that it runs a bit faster and there is less flicker on the display.

[Neeraj3]

Hi, I am working on this project.I am facing problems.The mosfet is not getting turn ON when i connected the current sensing input to the A/D channel of arduino. It always shows current input as 0.00. I need the help to over come this problem.