 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Electronics : picaxe 08M_Boost_maximizer
| Author | Message | ||||
GWatPE Senior Member  Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
The illusive, matching of wind energy to electrical energy using a permanent magnet alternator[PMA] and into a battery load is now a reality. The purpose of this exercise was to use this little chip, a picaxe-08M, to perform a difficult task. I am surprised that the companies with serious money and the resources available to them have not been able to solve this for the small scale, in a commercial way. It could be possible for my code to be re-engineered into probably any micro. The code for the picaxe-08M, a relatively cheap and small micro controller, can be written in Basic and there is no feedback control loop. There is only direct control using the output voltage to fine control the inductor voltage gain, and the output current to control the power transfer level. I have found a way using voltage controlled frequency modulation and current controlled pulse width modulation to load a PMA at very close to the optimum maximum power the blade could extract from the available wind. To confirm the results so far obtained I will correlate the test data with a seperate windspeed v power measurement. I have already determined the DC-DC conversion efficiency. I only have to finish an anemometer that will produce sufficient pulses v windspeed to be sampled at up to 20 readings/sec to be used in conjunction with a data logging power meter at the same rate. The final loading profile that has to be programmed into the 256 bytes of the picaxe-08M is highly non linear. I have also been able to provide an output regulation function. I do not need a diversion load and my windmill shuts down when the load is removed or the battery charge exceeds 29.5V, or a control toggle switch is opened. The boost arrangement I have used has a nominal cutin voltage gain of 3.5. As the power level increases the voltage gain reduces. At full power, the voltage gain is 1. At this point the switching electronics play no part. This is exactly the opposite for a buck system, as at maximum power, the inductor and switching components are performing at a maximum. This system will work probably equally well on any system where the windmill is optimized for the cutin to be close to the minimum battery voltage. At generated voltages below the battery voltage, the electronic cct calculates the voltage gain and the power and sets an appropriate output power control amount. A cct that will work on a 24V battery system of approx 500W capacity occupies a pcb I have designed and had made of 4"x3". This has a single SMD Mosfet and 2x SMD Shottky diodes, no heatsinks. This pcb utilizes mainly smd, [sugar grain size components], so was difficult to populate with the components. The most expensive components would be the 6-36V input to 5V logic power supply and the Alegro 50A current sensor. The cct consumes about 10mA in operation. The potcore has about 25 turns of 1.5mm enameled wire and no airgap. This was a std sized component from Neosid in Brisbane. The pcb has to be double sided, or multi layered, incorporating a groundplane. This would also be the case if a buck design was made. A limiting factor of the picaxe, that means I will not go into any production type future is the low resolution of the PWM function, at acceptable modulator switching speeds. I would consider this not a DIY project. This is not a magic bullet, one design will do all type solution, but the code allows for good matching within about a factor of 1:4 input to output voltage. The system voltage will determine inductor turns, capacitor and mosfet and diode voltage ratings. The cct is being tested tonight. We have gale force winds, probably till tomorrow arvo. The last 3 hours have produced 30Ah @ 27.6V. The biggest loss is the potcore. This has a combination of ferrite and copper losses. All the current has to pass through this device. The maximum power so far is 380W, so this is down a little on my analogue design. I do not really think a maximum is much of a guide. The most benefit is the ability to put current into a battery if the windmill is turning, and in doing so get as much battery current as possible from that wind energy. Please do not ask for cct layout or code. I am having a holiday and will come back and decide what I intend to do. The cct has to demonstrate survivability before I go further. If all goes well, I will talk to Gizmo about adding a topic to the projects section. .. .. Gordon. become more energy aware |
||||
oztules Guru  Joined: 26/07/2007 Location: AustraliaPosts: 1686 |
Gordon. Seems like a sensible approach. I like the part time boost method. From your previous posts the reaction time is good, you've ditched the dog chasing it's tail feedback loop...... you look like your good to go. Getting an anemometer positioned so as to measure what the mill is seeing will be the trick to testing the validity of your results. This will guide you on whether you are maximising a badly loaded mill, or maximising the mills wind performance loading to gain better/best output. Hopefully it is the latter. Have a good holiday ..........oztules Village idiot...or... just another hack out of his depth |
||||
| brucedownunder2 Guru Joined: 14/09/2005 Location: AustraliaPosts: 1548 |
Hi Gordon,,nice work ,,, you certainly have demonstrated some deep thinking .. Have a slow,safe trip up to the Gold Coast ,,the weather is perfect .. See you when you arrive,take care.. Bruce& Ilda. Bushboy |
||||
Gill Senior Member Joined: 11/11/2006 Location: AustraliaPosts: 669 |
Gordon, Great work. Looking forward to reports the unit is performing to expectation over trials. Happy holidays. was working fine... til the smoke got out. Cheers Gill _Cairns, FNQ |
||||
| domwild Guru Joined: 16/12/2005 Location: AustraliaPosts: 873 |
Gordon, Have a great holiday! I believe you use Lakota blades. Are they any good and do you mind telling me how much they were? Taxation as a means of achieving prosperity is like a man standing inside a bucket trying to lift himself up. Winston Churchill |
||||
Gizmo Admin Group  Joined: 05/06/2004 Location: AustraliaPosts: 5019 |
Well done Gordon, hope it gives back the results your hoping for. When/if your ready to put together an article about the maximiser, let me know and I'll put something together. Even an article about the trials and tribulations of building the thing would be interresting reading. Glenn The best time to plant a tree was twenty years ago, the second best time is right now. JAQ |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Hi Dom, Yes the Lakota blades are very good. Someone had wind tunnel tests in germany that confirmed this. I hope to have mine a long time. I have recorded 1208 RPM unloaded on my mill. I operate them at around 500 max RPM loaded. These are 2m rotor dia. Not much use on a F&P conversion, unless you used 5 or so. My mill only has blade drag and bearing resistance. The ironless stator with special wire has reduced the circulating eddy current losses that would normally occur in the very strong closed magnetic field in the 3mm magnetic air gap I have used, at high RPM's. I was told that the only way to get a set of Lakota blades was to buy a Lakota mill. I nagged the importer, I think in Melb. I had to wait over 6 months. I would probably replicate a wooden set, slightly longer. It is easier to remake when you can make some templates. I cannot remember the price, at least $500 in 2005. The maximizer survived the gale force winds, and top end output was slightly lower when compared to the maximum from my analogue version. 400W is pretty good still, from a 2m dia mill. The inbuilt output regulation is a winner. No need for a diversion load. The cct turns off [soft regulation] and only the peaks of the wind gusts can get to the battery. If the battery voltage still rises more, then the mill is automatically electronically braked [hard regulation] and kept OFF for 60 seconds. This requires no additional components. I hope to test on Bruces test rig in a couple of weeks. I have thought about some options and I will eventually publish my cct layout, and if anyone wishes to build a unit, I can supply a preprogrammed chip. This would be like having to buy a hall effect sensor, or a mosfet driver etc. I see no point at this stage in publishing the code. The design allows for a windmill to be interfaced to a solar MPPT, or other input regulated DC-DC converter. This allows a high voltage [say 80V] windmill to be connected to batteries at say 24V. The windmill could start producing power at a cutin of 30V. The power could be transmitted as DC at say 80V. The buck cct would need the battery maintenance functions. As the battery became charged, the load would diminish and the windmill would unload. The unloaded RPM may rise to a point where soft regulation ended. As soon as the voltage rose higher, then the windmill would be electronically turned OFF. This system would allow longer transmission distances with thinner wire. I intend continuing testing of these ideas when I return. .. .. Gordon. become more energy aware |
||||
| dazz Regular Member  Joined: 15/04/2008 Location: Posts: 78 |
Hi Gordon, Well done! I know you have spent a lot of time and effort developing this solution. It's very exciting 
You might find an anemometer not suitable for measuring the finer points of wind gusts. they still have quite a bit of inertial lag and drag when compared to the rapid fluctuations during gusts. If this does pose a problem then i suggest a wind force sensor instead. This can be built from alrady available commercial force sensors used in factory automation or you can build your own. Cheers, Daryl |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Hi dazz, I appreciate your comments, but this is not a PHd thesis project. I will correlate averaged windspeed data from my Davis weather station with averaged unloaded windmill data and then to averaged loaded windmill data. There is no point in using instantaneous data, because the windmill will not accelerate fast enough to make full use of it. I have not been not able to use the cct to stop the mill completely, so I will just use a separate voltage sensing switch with some hysterisis. .. Gordon. become more energy aware |
||||
| domwild Guru Joined: 16/12/2005 Location: AustraliaPosts: 873 |
Gordon, 400W with a 2m dia turbine is very good. My cogging F&P needs a 3m PVC prop, two blades, to turn over and then stalls at some unknown top RPM. This is done with a test tower of 3.5m height only. I am looking forward to see what the circuit looks like. * Taxation as a means of achieving prosperity is like a man standing inside a bucket trying to lift himself up. Winston Churchill |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Hi readers, To clarify a little more, what my maximizer does. Most would agree that the staggered rewire allows a F&P stator to have probably 3 useful separate additive cutin speeds. This allows each section of the stator to extract power at the cutin and add this to the sections already producing current. This allows the loading to vary according to windmill rpm. This is the goal of windmill loading. The trouble is, optimum use of the copper in the coils and the space is not achieved. Each of the stator coils has a maximum current output. Each of the coil groups have to be independently rectified, so this has to be done on the millhead to reduce the total wires running down the tower. As has been shown on another thread in the windmill setion, capacitors across eack winding can allow a better matching of the winding to the loading required at a particular rpm. At the end of the day there is a maximum useful number of windings that can be configured this way, and there is no automatic way of tuning the system to the changing load as a result of battery [state of charge]SOC. The maximizer that I have designed, essentially performs this same role with included SOC compensation and by using the winding reigime that gives the maximum current at the maximum rpm from the blades in use before furling. On a F&P, with a 24V battery, with 3m rotor dia, this will probably be an 80SP wired in delta, with 200uF capacitors. I intend to confirm this later this year on my own mill. I intend testing a 100S in delta and an 80x2S7P in star and delta and a 100x7S2P in star. The latter may be the best match. .. .. Gordon. become more energy aware |
||||
 Print this page |
