More MPPT Setup Progress

[s4w2099]

Ok, as I said before I materialized my data logger board that mates with my MPPT controller via the SPI port.

It logs data to an SD Card and the data can be extracted by reading the SD card in a computer or by connecting the data logger to a PC using the USB port.

So far it is logging great but I still have a few rough edges to chip out. I will do so when I get more time. I need to log input voltage and current.

Here are some pics:

Front:

Back:

The complete setup:

Obviously this is not where everything will ultimately go. I need to get a nice box and bolt everything nice and tight to the wall. But right now that is good enough while I finish adding features to the code and complete the debugging.

Meters measuring input current and Output Current:

On Feb 4th I used the data logger to sweep the PWM from 0 to 99% to get the real life VI curves for my complete setup, not just manufacturer's curves that often lie. I have in my roof 4x Harbor Freight panels in series + 4 first solar 50W panels for a total of 260W. As I have commented before my panels are not angled very nicely for this season but I see improvements every day as the sun moves back up to the sky. Here are the curves I got:

Luckily I am getting a bit more power than that now.

Now I show some of the data that I have grabbed starting Feb 23rd - Feb 27th (5 work days).

Battery Voltage:

Current Into Batteries:

Power Into Batteries:

Converter Temperature:

Now the last one is a very interesting one. At the beginning of the week I left the cooling fan disconnected to see if the MPPT controller would make it. It turns out that its thermally safe without a fan running at about 16Amps on the output. Though the temperatures are a bit high for my taste so I decided to put the fan back on by the third day at about mid day. As you can see the temperatures quickly went down to 60C and remained there. There after the temps never went over 60C with the fan on. This week has been very hot, about 28C during the day.

I am adding an excel spreadsheet with the data collected point by point:

Click Here for Spreadsheet

This data logger can do a lot more things than just log into the SD card. It also has an H bridge to control a 5A motor at any speed and direction in case that I want to play with a tracker some time. It also has 8 analog input to measure anything else I want to measure and record!

Those 5 days I got:

A total of 5811Watt/Hours

Max Power: 213.57W

Avg. Power: 40.2W

Max Current: 15.96A

Avg. Current: 3.15A

Max Voltage: 13.55V

Min Voltage: 12.35V

Avg. Voltage: 12.76V

Max Temp: 126.86C

Min Temp: 15.45C

[boB]

This is really neat !  Good job I must say.

Where were you taking the temperature from ?   FET Tabs ?

Inductor wire ?  Core ?  Somewhere else ?

Are you actually MPP tracking or finding the sweet spot by hand right now ?

boB

[Opera House]

I would be interested in seeing a plot of the input voltage during the day.  I've always felt that operating the converter at a fixed input voltage at the power point gets you 90% of the benefits of constantly calculating the power point. Wonder if you think this is true.  My panel installation is 800 miles away and can not do the tests.  This year I am taking the time to get more watts out with a better controller design.  

[ghurd]

Very Cool.

No logging of silicon smoke?  

G-

[s4w2099]

Hi,

I am taking the the temperature right from the mosfet's tab. These mosfets are surface mount ones so I have a SMD thermistor right next to it. The inductor is also right next to the mosfets so if it gets too hot it also triggers the fan.

I am doing full dynamic MPPT using P&O method and some other combined. The full MPPT description can be found here:

http://www.fieldlines.com/story/2009/1/26/11314/0349

I have a jumper in this board to allow me to set the PWM to whatever value I want just like an open loop buck converter. It allows me to test if the MPP algorithm is working at its fullest.

[s4w2099]

Opera House,

Yes you are right. Even though my MPPT is 100% automatic I have noticed that the input voltage remains pretty much constant at MPP throughout the whole day. There are a few exceptions though.

When the ambient temperature goes up the MPP voltage goes down and in a manual setup you would be loosing power if you are not there to adjust the converter. For example, in January my ambient temperatures ranged from -5C to about 20C and the array was operating at about 62V. Now my temperatures have risen and my array is operating at about 55V.

Even when its overcast my array operates at about 62V and yieds quite a few amps into the batteries.

I will be adding new features to the data logger soon. I also want to log input power so I can calculate converter efficiency at different conditions.

Thanks

[s4w2099]

:-(

Unfortunately I did not fitted in a smoke detector.

It could be added via one of the free ADC ports though.

:-D

[Colaman]

Colour me impressed.

Are you thinking of selling these one day, or are you just teasing poor slobs like me who have no access to SMD tools and board-making facilities? :-)

[TomW]

colaman;

He still has to get some out to us Alpha Testers to see if we can release the magic smoke before it can get released to the masses!

It sure is a sweet project that many will find useful.

Been watching his progress over on IRC as he pulled it together. Very impressive.

Tom

[s4w2099]

Lol

I think I will be selling MPPTs but I made the data logger just for play. I would add data logging capabilities to a bigger, meaner version of my MPPT controller though, and that could be sellable.

Thanks for the compliments guys but the data logger could be better. As Ross suggested I should have added a real time clock to keep time more accurate. Also I need more RAM than the available in the PIC16F877A to be able to read/write FAT16 partitions. Unfortunately I noticed that too late, so I ended up logging the data raw in the SD card.

I will be making another version of it when I find all the possible annoyances that this one could have so I end up with a nice clean and robust tool.

[DamonHD]

That's what prototypes are for, unless unlike me you're a clairvoyant supergenius that never makes a mistake!

Never be embarrassed to bin/recycle V1.

Rgds

Damon

[WindJuggler]

Hello,

I have been working on a similar project for the past year. But it is rather going slowly here. I have finished my PCB but there are still some problems. I had the same problem with the SD-card (I used a 18f4550). I haven't started on that code jet.

My controller is allready all-in-one. It has a data logger, it has an MPPT converter on board (sepic) and it can regulate 2 dump-loads.

I made my controller easy configurable from the computer via a simple program.

I still having some prolems with my inductor (50A controller).

If you want some help with your (next) controller, you can count me in. I have had some mistakes with the last one and I have learned a lot from it.

[s4w2099]

Thanks, I really appreciate it.

[Yyrkoon]

Heya s4w2099, I've talked to you on IRC once or twice about your MPPT controller a while back, and I am glad to see that you're progressing nicely.

Perhaps check out rabbit semi and their embedded systems.

http://www.rabbit.com/products/CoreModules/

Now as a disclaimer, I do not work for them, and all that. We just have designed our own solar controller around one of their systems, and while the first iteration worked great, we are now working on a V3 model. As a caveat, they use a non standard version of C. I really prefer ANSI C to dynamic C, but whatever . . . I am after all still a hobbyist programmer.