Author Topic: Windmill Data Logger (Read 6331 times)

gizmo · « **on:** August 05, 2006, 10:48:16 PM »

Hi All.

I built this simple data logger a few weeks ago and put the construction details on my web site. I know there are some people in the US who would like one, so here it is. I called it "PicLog".

The "PicLog" uses a 08M chip and a handfull of components. The circuit is simple and could be made on veroboard. Parts should cost less than $15, including the chip.

By itself it can work as a simple shunt regulator, but with the addition of the PC software it can record battery volts, amps, wind speed and/or windmill rpm. This data is sent from the PicLog via a 3 wire serial cable.

The PC software displays the current readings, refreashed every 3 seconds, and logs the data into text files, html page ( for viewing live data over the web ), and a Excel file.

There is also a basic graphing function.

The full details are on my web site at

http://www.thebackshed.com/Windmill/PicLog.asp

This project requires some understanding of Visual Basic, PicAxe basic and circuit board assembly. All the source code is included in the download. There is still room for improvement in the PC software, its an on going project, but as a windmill logger its been very useful. If anyone would like to improve on the PicLog, or supply circuit boards/programmed chips, please let me know.

Glenn

richhagen · « **Reply #1 on:** August 06, 2006, 02:25:47 AM »

That's pretty neat! I'm not quite ready for this project yet, but I am going to bookmark this page so that when I am I can use it. I have a small weather station, and It would be neat to have both for performance evaluations. Rich

amiklic1 · « **Reply #2 on:** August 06, 2006, 02:27:21 AM »

Great stuff. I still have no experience with microcontroller, but I'll try to use this one for sure.

I am still wandering, could it be done that the data is recorded in a kind of a memory card instead directly onto the computer. That would be extremely handy, battery operated (for remote sites)....

commanda · « **Reply #3 on:** August 06, 2006, 03:25:39 AM »

I have a suggestion; accumulated daily kilowatthours.

The program isn't perfect, and may/will crash from time to time.

Stop blaming your program. This is just an undocumented feature of Windows.

Amanda

gizmo · « **Reply #4 on:** August 06, 2006, 07:02:06 AM »

I did consider on board data logging using a 8 pin eeprom, but for now I wanted to keep it simple and cheap. One problem with the PicAxe chips is accurate time recording, it just cant do it. You could add a dallas chip to provide a accurate time source, but again this adds to the cost of the project. Using a PC for the logging solves this problem.

Glenn

gizmo · « **Reply #5 on:** August 06, 2006, 07:10:44 AM »

Displaying watt hours is a good idea. I might add that to the software in the next release. I added the Excel export because someone suggested it would be handy for those fluent in Excel to create their own graphs and reports ( watthours ).

The logging is based on average readings. The PicLog sends data every 3 seconds, and if you set the logger to record every 60 seconds, it will average the 20 odd readings recieved in the 60 seconds since the last data save. This means you may not record any peaks in power, but using a average reading gives better real world figures.

Glenn

willib · « **Reply #6 on:** August 06, 2006, 07:50:44 AM »

its simple to access the registers that holds the time and day on the pc, and just attach them to the incomming data.

it would go something like ,

get time , get volts , get current , ect..

get new time , get volts , get current .

willib · « **Reply #7 on:** August 06, 2006, 09:02:45 AM »

i've been to your site before , its very nice

how is the double F&P comming along? i noticed that you offset the second rotor one tooth/armeture ,that should cut down on the cogging alot .

jimovonz · « **Reply #8 on:** August 06, 2006, 01:46:26 PM »

Great! I have done something similar a couple of times - though not nearly so 'polished'. I am always amazed how cheap/versitile the 08m is. In my case I drive the dump load directly with some fets and pwm to maintain a more constant voltage. My current measurement is via a hall effect sensor for increased resolution/less loss. I hope to have one reporting online soon. Am archiving to MySQL and just brushing up on mt PHP to produce some nice graphics for the webpages.

wdyasq · « **Reply #9 on:** August 06, 2006, 10:26:00 PM »

Glen,

The least expensive time and data device may be the "Palm Pilot". If one could port the information to one of those it has the time data and I think they can download to a memory chip. One of my old "Visors" has an 8mb memory card that is propriatary. I think new ones may take other chips.

A "Palm Pilot" also has a very low current draw. I think the older ones are inexpensive enough used where an actual machine could just be switched or downloaded at intervals.

Ron

kitno455 · « **Reply #10 on:** August 08, 2006, 11:58:28 AM »

you guys might want to take a look at RRDTool. it is a really neat graphing program for multiple streams of time based data, runs on linux, by the same guy that made MRTG.

allan

mattbrush · « **Reply #11 on:** August 08, 2006, 09:26:42 PM »

If anyone cares, dallas/maxim make real-time clock chips with integrated quartz crystals, lithium battery, and ample non-volatile (battery-backed) memory.

Interfacing to them with a microcontroller is quite easy.

They also make a data logger chip that has a programmable sampling interval, real-time clock, analog to digital converter, and memory. It basically just sits there logging data into it's memory until you tell it otherwise.

The big bonus is you can sample their stuff for free.

My $0.02

Matthew Brush

domwild · « **Reply #12 on:** August 09, 2006, 02:22:39 AM »

Great job, once again! Just what i wanted to do! Assuming I use seven coils in parallel on the F&P, would there be seven pulses per revolution to be measured by the Picaxe count command??

Commanda suggested splitting the 14 coils into two lots of seven and paralleled, this is where the seven comes from.

Have spent big on a 28X and dream of using it also for the star/delta switching.

commanda · « **Reply #13 on:** August 09, 2006, 03:35:03 AM »

Dom,

I answered this question for you recently. Maybe with my email problems you didn't get it, in which case I apologise. (My isp, unwired, started blocking port 25 without prior notice. It took me 5 days to realise none of my emails were getting out.)

The number of pulses per revolution is based upon the number of magnets, not the number of coils. I've got the rpm/frequency info somewhere.

Amanda

gizmo · « **Reply #14 on:** August 09, 2006, 05:06:37 AM »

I had problems with the double stator. The plastic hubs would flex, and interfer with the stator. Down under we are looking at replacing the plastic hubs with cast aluminum units fitted with neo magnets, still under developement, but this will make double ( or more ) stator windmills feasible.

Glenn

gizmo · « **Reply #15 on:** August 09, 2006, 05:11:34 AM »

Yeah Dom, Amanda is right.

I made the same mistake myself, thinking you would divied the number of pulses per revolution by 14 ( number of poles per phase ), but in fact its divided by the number of magnets, 56 for the F&P. 56 magnets gives 56 sine waves per revolution, but I "Count" the positive cycle, so divide 56 by 2 = 28 pulses per revolution.

Glenn

gizmo · « **Reply #16 on:** August 09, 2006, 05:13:26 AM »

Sorry thats wrong...

"56 magnets gives 56 sine waves per revolution"

56 magnets gives 28 sine waves.

domwild · « **Reply #17 on:** August 09, 2006, 09:03:20 AM »

Thanks for that, Amanda and Glenn. Your (Amanda's) email did not get through.

willib · « **Reply #18 on:** August 09, 2006, 09:11:46 PM »

I cant make out your data ?

Your average RPM and average KPH dont match.

what i mean is line 4 & 5 , your average max air speed for the particular data set shown (20 kph ,12.4 mph ),does not correspond with your average max RPM line nine.

It may sound like i'm nit picking ,but not at all. i just think that the numbers should correspond , and it would bug the crap out of me till they did match.

if you sampled the wind speed more often that may help a bit.

also you need a better match between the alt and the blade set , because your mill spends most of its time not producing any power .

if it could produce power in the 6 mph range(9KPH), that would help out a lot , it may be as simple as making your blades a little smaller ( more rpm) , or adding another three coils (one per phase) , not the easier of the two choices..

have you got a photo of the blades per chance?

i am very jealous of what you have done so far , keep up the good work , some day i hope to have a similar data logger for my blades , but i need one more piece of data , torque output of the blades vs wind speed and rpm.

windy · « **Reply #19 on:** August 09, 2006, 09:17:46 PM »

gizmo

I would like to try and build your datalogger but I'm not sure if it would work on my windmill. I built a 3 hp 3 phase motor conversion that I plan on using to heat hot water without using batteries. The generator puts out 58 volts at 2 amps per phase turning at 300 rpm to a maximum of 117 volts at 5 amps turning at 600 rpm.

Would your circuit work for this or does the voltage have to remain constant? I could use a small 9 volt battery to supply power for the circuit. I'm just not sure if the IN4004 diode would hold up to the 117 volt or higher spikes.

Any comments or suggestions would be greatly appreicated!

windy

gizmo · « **Reply #20 on:** August 09, 2006, 09:59:30 PM »

The data files included are there for testing, and not that accurate. Wind speed and windmill rpm dont realy line up in real life, the load on the windmill is dynamic, depends on battery charge and wind speed, where wind speed is only affected by wind speed. That said, I should also mention the wind speed anomemeter is not next to the windmill where it should be, its actually next to the house about 20 meters away from the windmill.

The windmill used was a Lenz2 design on a F&P motor, and not well matched at all. Again, it was set up to get some basic data into the logger, and not a valid performance test.

The project is by no means a fully fledged data logger, but a working logger all the same. Like all the projects on my site, they are there to get people thinking, cheap to make, and used as a basis for your own designs.

You could use the Excel output to calculate blade torque. I dont know the maths but someone here might be able to help.

Glenn

gizmo · « **Reply #21 on:** August 09, 2006, 10:07:49 PM »

I could be made to work. You would need to change the 22k resistor in the Voltage measurment circuit to 220k. As you probally use AC for your elements, you would need to add a rectifier as well so the input to the PicAxe pin 3 is DC. In the RPM measuring circuit, I would change the 10k resistor to 47k. I would not use the current reading, just leave it not connected.

The circuit draws about 8mA, so a 9 volt battery wont last long.

Glenn

commanda · « **Reply #22 on:** August 10, 2006, 03:50:10 AM »

The windmill used was a Lenz2 design on a F&P motor, and not well matched at all.

Looks like you're several steps ahead of me, again.

Height, diameter, and coupling to the F&P? Direct drive or geared up?

The smoking area outside work is between 2 buildings, and forms a natural wind tunnel. Perfect place for a Lenz turbine. Just need a weldmesh wall to protect the smokers if the turbine flies apart <grin>.

Amanda

gizmo · « **Reply #23 on:** August 10, 2006, 05:50:51 AM »

The Lenz is 1200 high, 1300 diameter. Direct drive to the F&P. I've tried a F&P in standard connection but slightly decoged to reduce noise. Starts charging at 20 rpm ( 12v ). I'm currently running a stator wired in delta, starts charging at 30 rpm. The windmill is in a bad location, only a few feet from ground level in low turbulent winds. I should mention the HAWT near by isn't even running this time of year.

I'm starting on a new Lenz, this one 2400 high, 2050 wide, on a 8 meter mast to replace the HAWT. The fact that the Lenz is making power where the HAWT isn't has convinced me VAWT's are the way to go for my location.

Glenn

domwild · « **Reply #24 on:** August 10, 2006, 09:38:18 PM »

Glenn,

Once again, many thanks for this contribution. Have contacted Michael Lawley at some stage to purchase the specially turned F&P shaft for the double rotor unit. Impossible to buy as Michael is of the opinion that 1000 W can easily be achieved with a single setup anyway. He is no longer pushing doubles for the hydro setups, where they were being marketed for. You have surely seen his hydro setup on the ecoinn web site. I suspect, you are, of course, pursueing your scheme for the simple reason of obviating the cogging. Convoluted English!

Thanks.

willib · « **Reply #25 on:** August 11, 2006, 07:48:10 AM »

i can see the reasons for using cast aluminum , low melting point being one of them.

i am wondering if a cast iron brake drum could be found that will fit around the F&P stator

what is the dia of the F&P stator?

just a thought

tonig123 · « **Reply #26 on:** August 24, 2006, 06:23:49 AM »

Wow! excellent work.

I'm not an electronic expert but I wonder if there is a better way to measure the amps. I feel that the resistance (shunt) at the battery is just a waste of hard-generated power.

News:

Author Topic: Windmill Data Logger (Read 6331 times)