Real Wind MPPT - my fourth patent?

[MountainMan]

The idea is, as an old girlfriend of mine used to describe quantum physics, "complexly simple".

First some background on my recently acquired feeble understanding of why we care about TSR.  My understanding is that any set of blades has some specific TSR at which it will perform at it's theoretical best.  The TSR at which that set of blades will generate the most power it can.  

A "sweet spot", so to speak, and this sweet spot is NOT the speed that set of blades would achieve if it were free-wheeling.  The free-wheeling speed is the max rotational speed for a given wind speed, but not the rotational speed that gathers the largest possible amount of power from a given wind.  When you apply "back torque" to the blades with an alternator, you are slowing it down from its free-wheeling speed towards its "sweet spot" on the TSR curve.  Happily, your alternator is soaking up the power from that back torque, so all is good.  If you apply too much back torque, you move the "system" past its sweet spot and back down the other side of the TSR vs. power curve towards the ultimate slowdown of a stall condition.

To make our set of blades hit their sweet spot with one wind speed requires a particular amount of back torque from the alternator.  To accomplish this same sweet spot at a different wind speed, requires a different amount of back torque.  There is no single "perfect" amount of back torque, as it varies with wind speed.  Normally one looks for a prevailing wind/blade/alternator/battery/load combination that produces an acceptable compromise.  Too "big" of an alternator and your cut-in speed is too high or the blade stalls in a modest wind.  Too "small" of an alternator and you never get as much out of your blades as they are capable of. (some of that last bit may be off, but you get the idea).  You have to reach a happy compromise.

If one were to hook an MPPT controller that is designed for solar power up to a windmill and charge batteries with it, it would solve some of the problems of mating alternator and blades, but not all of them.  Such a controller will try to take whatever the alternator is putting out and maximize its power transfer into the battery by trading off current and voltage.  The load presented to the mill is left to chance, and will not be optimum at all wind speeds.

If, on the other hand, one were to make a controller that knew how much load to apply for a given wind speed to get just the right amount of back-torque on the blades to make them run at their best TSR, one might achieve the other half of the solution, which is to apply a load that guarantees the best possible power efficiency of the blades for whatever wind speed is presently available.

It seems what is needed is a combination of these two principles into a single windmill MPPT controller which will use an anemometer input to gather  current wind speed data, compute based on previously learned data what would be the ideal load at that windspeed.  Such a load would result in these blades running at a speed that puts out the largest potential power at that windspeed.  Our alternator could supply this max power only if we give it precisely the right amount of load.  So the load on the alternator (which is supplied by our MPPT controller) needs to be adjusted to match this "best load".

So, our blades are capable of a particular amount of max power at this wind speed, and through the alternator they produce this power with a particular "best load".  We then need to adjust the algorithm of the MPPT controller's DC to DC converters to shoot for delivering exactly that same power level to the battery.  Make it adjust the charging voltage until that voltage times the resulting charging current equals our preferred power level, and voila!  Our MPPT controller is pushing just enough power to the battery to create our "ideal load" on the alternator we have connected.  This in turn provides precisely the back torque required to make our blades spin with whatever we have empirically determined is their ideal TSR.

So, in review, our special MPPT is adjusting the charging voltage/current so as to take whatever alternator we have in our system and use it to move our blades into their sweet spot on the TSR vs. power curve and keep them in that sweet spot as the wind speed varies over a wide range.

At the same time we get other benefits such as making our system less dependent on precise alternator design, getting the lowest possible cut-in speed, and not wasting power at high wind speeds by running our blades at an inefficient TSR.  And finally ( a question for the experts) perhaps by decoupling some of our divergent design criteria such as cut-in speed vs. highest max power, perhaps we can develop a set of blades with a higher efficiency without sacrificing cut-in speed.

See, just like the lady said, complexly simple.

[windstuffnow]

  The idea is sound, and it would be quite interesting to say the least.   Also, it would be nice to have a true wind turbine controller.  I believe Whisper and Bergey use something similar in their controllers.   For the most part though, keeping it simple works the best even if it tends to be a little inefficient.  I tend to build machines that work the best or most efficient in the average wind of my area ( that would be, in my mind, the "sweet spot").  If it blows harder thats a bonus but if it's chugging along making my 200-400 watts most all the time then it really doesn't matter if I get the 800+ it will do because my batteries are most likely going to be charged.  

  You might need to incorporate some type of overspeed control into the system as well.  What would happen in the controller when the batteries are charged?   Will it automatically dump the extra power?  Sounds like it could get complicated fairly quick....

.

[scottsAI]

This was posted under: MPPT charger with windmill/genny 23nov05 (changed here).

MPPT in your example is a constant voltage out, so it's varying the output current.

(Battery as a load is nearly a constant voltage device)

The MPPT would work the same for wind and solar. Most MPPT work in a cycle. Each cycle it increases the current, looks to see if the current did increase. If not then it reduces current next cycle. The controller is changing things, always looking to max out current. Based on a given wind speed and generator, the MPPT would be doing the same thing. Not necessary to know the wind speed. Just load it, see if the current went up or down. Might need a larger delay time then a solar system would need, due to the mass of the blades. For a short time you can steal power due to the mass of the blade.

MPPT would work fine, one big problem.

Lets say you have a nominal 1Kw system,  high winds even more...

The MPPT uses an inductor to transform the energy up or down. The inductor can handle about 10x from min to max.

In low winds you want it to start at 50watts, then it's maxed out at 500w. The system is only half way there. Several inductors could be used, system cost and control more complex.

Solar modules, die quickly, 10x is fine for solar MPPT.

I do believe the wind gen efficiency could be improved with MPPT, but the reality is the wind machine is cheaper than the electronics needed to improve it.

(Reports here of $301 for a 1kw system materials including tower, no labor).

If you have low winds often, then an MPPT may be of value, just to get some power out.

My suggestion would be to build a low speed wind system. Forgetting the power in the higher speeds. Bunch of small low speed 50w wind gen, cheaper than building MPPT.

Star / delta switching helps, not hard to do. So maybe 100w systems. 10 or so.

OK one big one, just can't pass up the higher speed winds!

Lastly the blades for a given profile can be improved for low speed operation by changing them.

The MPPT for a 2Kw system would be around the cost of an inverter of the same power. For the MPPT is nothing more than an inverter. Considering these things, my best guess would be: build a low speed wind gen, and high speed or have changing pitch blades.

Last idea would be to have a dual gen or dual winding gen, one winding for low speeds, other for high speeds. Series / parallel windings may work out.

Have fun,

Scott.

[Flux]

The idea is fine and one day when the Chinese are churning the things out for next to nothing everyone will use it.

At present it will only be for the few people who can make their own. Ed has hit the nail on the head, the time when it makes most difference is in high winds and for most people the batteries are already charged. Within reason you can arrange your needs to do things on those windy days and use the extra power but it is not realistic to have bigger batteries.

For larger machines where the heat is worthwhile you can make a scheme that keeps the batteries charged and produce maximum heating at the same time. Heating control is simpler where you are not tied to constant voltage.

I think you may be right about a new form of prop design, I find that the type of props I make are happy to run fast in low winds and slow in high winds and this is one factor that makes the gain from a tracker less than it should be.

The gain from a tracker in low winds is small if the alternator is chosen for low wind performance anyway, the gain becomes significant above about 15 mph so it largely depends on the wind area you are in and your ability to use the extra power on high wind days.

With battery systems it is a good way to increase the output of a small machine if for some reason you can't do it any other way, but a small increase in prop size will have the same effect in high winds and you will also gain in low winds where the tracker will not help. Having made the prop bigger there is then the tendency to want to add the tracker to do better still. If you are a manufacturer where you need impressive figures for a given size of machine then it is attractive if the cost is right. For home build it is attractive for those who can do it but for the rest who want to produce power reliably then the KISS approach is likely to remain with us for some time.

The thing when operating properly is very nice and is much better than star/ delta or series/ parallel switching and it does also have the nice factor that when the batteries are charged you can phase the converter forward and stall the prop. The size of shunt controller is then reduced drastically.

Flux

[MountainMan]

Hi Ed,

Yeah, you would definitely need to have a dump load also.  On my property that is likely to be an electric water heater that exchanges water with my 10,000 gallon fire water tank.  In the winter I can use the tank with a heat pump to keep the house warm.  That way the dump load isn't wasted energy.  Who knows, with enough VAWT's spinning on the property, maybe I will be able to get that tank warm enough to just use the heat directly in the radiant floor heat system without a heat pump.

You're right though, it does complicate the system.  I'm a proponent of KISS principle, but often forget about it when I have an idea forming in my head.  Still, to misquote Einstein, you should always look for an answer that is as simple as possible...but no simpler.

thanks,

jp

[MountainMan]

Hi Scott,

You're right, I knew there were earlier mentions of solar MPPT with wind power.  The patent subtitle was a bit of a joke.  I don't think there is anything here novel enough to bother with a patent.  A competitor could always accomplish the same thing with a slightly different mechanism.

Not knowing much about MPPT for solar, I assumed that it would have an algorithm based on a nearly instantaneously reacting solar panel, with just milliseconds worth of load juggling to find the panel's sweet spot.  I also don't know if it's solar based algorithm would be completely hosed by the non-linear response a wind mill/alternator would provide compared to the very linear response of a solar panel.  Perhaps a really long delay with a standard MPPT would work without the complication of knowing wind speed.  Perhaps not.  I think that the wind speed will vary frequently enough in unpredictable ways that it would be difficult for a simple non-averaging algorithm to be of much use here.  I think a learning algorithm with instantaneous wind speed readings would wind up beating a simple "change it and try it" algorithm hands down.

Anyway, the primary purpose of this mental experiment was for me to test my understanding of the whole system, and I think I've convinced myself that I understand it well enough now to get busy actually designing a system.  Still lots to learn, but the basics are in place now.

best,

jp

[jimovonz]

Would not a true MPPT find the 'sweet spot' in any case - without any special consideration of the blades/TSR? By adjusting the voltage/current to obtain the max power point the blades would automatically be in the 'sweet spot' because this is where max power is made?

[jimovonz]

Scott, as I understand it the role of the inductor in a SMPS (MPPT) is to store energy so that it can be metered out in a fashion more suited to the load. i.e energy is dumped into the inductor quickly and released more slowly over a longer period or vice-versa. If we restrict our scope to step-down converters (I hope I'm right, but I believe that most of the turbines we see here on the board would require stepping down after cutin to reach the MPP - the situation would be different for an alt that required stepping up in order to make power at a useful cutin speed) Your typical buck converter draws energy in pulses and the output is filtered by the inductor to produce a somewhat more regular output that is basically an average of the input. Given that the batteries in our systems are quite happy to absorb energy in pulses and infact act as an effective filter themselves, would a MPPT have to utilise an inductor at all? Would simple PWM not achieve the same goal? If the alt sees the same pulsed output, and the battery doesn't care about pulsed input then we can achieve the same energy transfer as compared to a true buck converter.

I have put togeter a simple PWM circuit configured as a MPPT for my latest project (smartdrive conversion used to power a remote wireless repeater) and seem get good results - certainly better than the smartdrive alt alone. I have to pull down the turbine soon to put it on a decent pole so plan to do some comprehensive testing then. My MPPT (if I may call it that - maybe it does not track the Max power point, but it does certainly track a Better power point - BPPT cost me all of about $40 to put together. With out putting the output on a scope, its hard to say what the output into the battery looks like but surely the inductance of the alt itself has a role to play as well.

[MountainMan]

Yeah, that was part of what Scott was saying.

It would, but only if its method of discovering the maximum power point was compatible with a wind system.  Unlike a solar panel, which would have an instantaneous response to changes in the loading, a windmill/alternator would have a much slower response due mostly to inertia of the blades.  That, coupled with the frequent changes in windspeed might easily confuse the heck out of an MPPT designed for use with solar.

It is, however, easy to imagine that such a controller with a proper algorithm would probably work well enough to make the additional complexity of having a wind speed input an unnecessary complexity.

jp

[Chiron]

Good idea but it's already being used on large turbines, usualy refered to as tourque control in the US. G.E. has several patents for these systems and I'd suspect the systems Vestas and others use is patented also. Though I don't know of any using DC output gennys.

A charge controller that could sense the current and RPM of the gen. and adjust the load accordingly would be the simplest. RPM on a PM genny is easy, just use one of the phases before rectification to generate pulses....

If the power curve of the blades were known, Ideal RPM for a given wind speed, and the voltage VS current the genny produces at any given RPM the ideal power curve could be derived and programmed into the controller keeping the blades in thier "sweet spot". I wrote about this in another diary http://www.fieldlines.com/story/2005/12/6/173732/267

Large turbines use an anemometer to measure wind speeds to adjust for gusts and lulls in the wind since the lag between gust and increase in gen output does have some lagtime with all that mass and inertia of the blades etc.  Cutting power during a short lull preserves the momentum of the blades so they don't have to come back up to speed again. I don't think this would be a big issue with a smaller turbine.

Anemometer is a bad word IMO considering how many turbines I had to climb to replace them.

The large turbines use variable pitch blades to maintain an ideal profile to the wind at higher speeds without overlaoding the gen with good low speed performance.

I've thought of a couple of different apraoches to do the same thing but they add levels of complexity and I'm a big believer in KISS.

[Flux]

The anemometer is not needed on a small machine, I would avoid it except possibly as a means of setting up the characteristic originally.

Rpm is good enough to control the characteristic although I find it easier to use the battery current.

I agree about keeping it simple if it is your source of power. I do it as a hobby and if something fails then it is not a big issue.

Variable pitch does make it much simpler as far as the electronics goes, but variable pitch is not very suitable for home manufacture, to work reliably it needs to be extremely well engineered, it is not worth the effort on any small machine.

Flux

[paradigmdesign]

Here is why, I think making a actual MPPT would be so hard.  There are many variables in finding the Max Power Point.  First there is the efficency that the blades are operating at, which can vary from +- 5% on an average variable speed rotor design.  That is +- 5% out of a max of arround 30%, so if you use the 30% as the max blade output, than the efficency of the blades from there closeness to there optimum tsr as a function of max blade output actually varries +/- 15% or a total of 30%.  Alternators efficencies are usually between 75%-92%, so a 1% loss in genny efficency is better than a 1% loss in blade performance.  Also you have to figure in the losses incorperated in whatever form of regulation that you are doing (including transmission loss differences).  

What makes all of that so difficult is the fact that all thoes variables change at a different curve.  In "theory" the blade efficency if kept in line with there TSR should remain fairly constant, while the genny's and transmission efficencies get better as rpm's increase due to the higher voltatge.  But it takes raising the resistance to decrease the loads from the coils, which decreases efficency.    

Since figuring out a formula for all of thoes varibles, and comming up with some sort of way to keep it perfectly in line is so difficult and expensive, I think I'll make up the loss of not controlling with a couple extra inches of blade, or some more copper & mags.  Also, alot of thoes controls can add "sharp" loads when switching between resistances, which can't be good for the genny.  Plus the time most properly made gennys will be missing out on is the low wind speed times anyways, if you want it to start up a little earlier match slightly larger and thinner blades to the same sized genny.  Like 60% of the farmable energy in wind is  in a MPH range of 5-6 MPH

Though a "perfect" genny would surely have controlls like that, I think I'll go with a "properly proportioned" larger genny for the same cost.

$300-$500 can buy a lot of extra blade and genny componets.  If you ask me, it all comes down to the amount of power you get at the end of the day.

[jimovonz]

Why does it need to be so complicated? The MPPT doesn't have to calculate the best loading - it just has to find it. The MPPT adjusts the current into the load up, checks the power output. If the power output goes up then it bumps the current up and checks again. If the power output goes down then it drops the current and checks to see if that improves power. Two inputs: voltage and current. No need to measure windspeed, RPM or anything else. As long as the interval over which you take your measurements is suited to the characteristics of the turbine it should find and track the MPP (alternating by one power level increment either side). By this method you should find the 'sweet spot' for the system as a whole, rather than for just the blades or alt. This method would have trouble if the power curve of a turbine exibited more than one local maxima - but as far as I have seen there is only one.

[paradigmdesign]

So at the very least you would need; a variable load, a voltmeter, a ammeter, and some sort of processing unit.  All to grab that .05% of the wind power that exists 9mph or less zone.  Unless your talking about building a fractional Megawatt or larger, I don't really see the point.

I know it bugs you too see your genny standing still on the calm days, but if you ask me, a MPPT that costs more than 1 or 2 cents per watt is useless.

[jimovonz]

I think the potential benefits go beyond low wind gain. Check out this thread if you haven't already: http://www.fieldlines.com/story/2005/11/24/184750/52

Flux posted a graph showing the effect a suitably configured buck converter has on a turbine output. In my case am am looking to optimise an F&P smartdrive conversions so I don't have the advantage of being able to alter the alt design to suit the application.  

If the alt is essentially free and a MPPT turns it into a viable wind turbine alt then I think I can afford to spend a bit on it. I don't expect the components to cost more than $50. I also have and electronic 'decog' circuit to aid startup (pulses each of the phases in turn to start the blades turning - overcomes the startup torque) that I will incorporate (the configuration of which also lends itself to synchronous rectification, but I haven't gone there yet...)

[scottsAI]

Well MountainMan, and jimovonz,

I'm glad you posted this question MountainMan.

It does keep coming up, and there is a reason. It makes sense.

Also, it has caused me to look into the subject of matching blades to generator. What the MPPT would do. I spent most of Saturday and Sunday working on it. I ran some simulations and have found, it's impossible to match, without some help. Wind speed is proportional to RPM as is generator (PM) output voltage. Wind power is a cubic function. The desired goal of low cut in speed. With a direct connect to the battery you  have to give up something. What I see is you loose the high end of the generator.

I discovered if you size the generator for good output at the higher wind speeds, then the generator is overloading the blades at lower speeds. Lastly if you want a low cut in speed, then the generator outputs a high voltage with the result most of the power is waisted in the generator! If you size the generator for the lower, then you get very little for the higher. Linear function does not match up with a  cubic function.

Jimovonz,

PWM control, could you explain how your doing it? I could guess.

You are right in the gen has inductance. Unfortunately it's not good for reducing the voltage but good at boosting it. Which brings up, sizing the gen for the right voltage at the desired wind speed, use the inductance in the gen to boost the voltage at the low cut in speed. This would be a rather simple controller. More complex then a direct connection but, now you can prevent overloading the generator etc. Complexity is minimized by boosting the low end, limited power being controlled, using the inductance of the gen simplifies the circuit. The low voltage will impact efficiency, but using a FET rectifier may make it work out. The cubic nature of the wind power, probably requires a micro to determine the boost level.

I have noticed the wind turbines output vs wind speed, the power goes up linearly. During my simulations I found out why. Should be cubic.

My write up and the simulation results are quite large, I have not done a diary entry, think I will post them there, next weekend. Thanks,

Have fun,

Scott.

[Flux]

Yes you have 3 options, match the generator at cut in and use a converter to increase volts at higher winds. Match the generator for high winds and boost the low end or choose a much higher generator voltage and use a converter over the whole range.

The 3 rd option is the only one that entirely solves the problem of line resistance.

The first option makes the line resistance easier but it is still an issue.

The second option is simple, uses a small and cheap converter and works very well at 24v or above, but for success it is essential to keep the line loss as low as practically possible. This is its major snag if the machine is a long way from the batteries but at 24v it is quite practical for runs of about 50M.

If the alternator has inductance then in theory you can use it as part of the boost converter, may work for the F & P, but it means chopping at a low frequency ( 1 TO 5 KhZ),  it also means passing the total current through the converter diodes.

Air gap alternators will have negligible inductance. I have not tried using the alternator inductance but even with an iron cored alternator I would still chop at a much higher frequency (30 kHz) and use a small ferrite cored inductor in the converter. Using the alternator inductance will almost certainly be associated wit extra iron loss. Also a self contained converter will not be associated with problems of voltage spikes that may damage the FET, long straggly wiring layout is a serious problem.

If you can make a simple FET rectifier work let me know, my attempts have been too complicated to consider it.

I think that the whole thing needs to be kept simple, micro controllers, anemometers etc may be fine for the youngsters wanting to exercise their digital skills but I just do it analogue ( I don't know how to do it any other way)

For the boost converter approach, design the alternator for a cut in of about 15 mph and keep its resistance low. You will have a perfectly good track of the prop output up to furling speed and you can manage about 80% if you can keep the line losses down.

If you boost the output from about 7 mph with a simple converter you will stall at about 10 mph, so you need to reduce the slope of the characteristic to a line that more or less fits the low wind power out. A simple 2 line approach is plenty good enough as the prop has a fair tolerance of tsr anyway. Props are not ideal and don't work as predicted so you have to live with that.

I alter the slope of the boost converter by using the battery current as a feedback to the pwm drive to the FET. altering the gain changes the slope. When the main alternator starts to produce a few amps the converter phases back and comes off line.

If you arrange the feed rectifiers suitably the converter never has to carry more than the low wind current.

As the wind drops and the main alternator output falls the converter phases back in. The whole thing is totally smooth without the bumps and bangs of star/ delta or tap changing.

At 24 v you should maintain 70% efficiency over the range from cut in to furling if you get it right.

Similar performance is available from the buck converter option, which is a better choice for 12v but otherwise is a more complicated more expensive and potentially less reliable option.

I have run boost converters for years, I have only run a buck converter as an experimental set up but so far it has been fine.

Flux

[MountainMan]

Wow, I went away and you guys kept going.

A few minor points I wanted to make.  One of the things that endeared me to this idea when I had it (even if I wasn't the first or even the 42nd guy that thought of it) is that you could build a sort of "one size fits all" controller that would mate any sub-killowatt blade set with any alternator and any stack of batteries and have that magical box "make the most of it".  

In other words, it would remove quite a bit of the current necessity to "start with a combination that somebody already had success with and then make variations from there".  As the current state of the art in DIY RE stands, one can either choose the size of one component and then slave over a calculator or spreadsheet for days trying to find a combination of the many other factors that suits him...or use an existing design.

I like the idea of a magic box that makes it more practical to just take a wild stab at a design and have a reasonable chance of having some useful juice flow out of it at the end.

As an analogy, when I was a  young'n, I was doing electronics design instead of programming.  There was one guy in the group who absolutely loved doing math - the rest of us just tolerated it when there wasn't an easier way.  When we had a prototype circuit we were working on, and we found that a calculated value for a resistor wasn't producing the desired results, Bob would always go straight to his trusty HP41 calculator and get busy.  One of the rest of us would simply use an "analog computer" to solve the problem.  Simply hook up a meter or a scope and insert a variable resistor (pot) in place of the part in question and adjust the pot until we had the results we were looking for on the meter.  Then yank the pot out and measure it.  Bob was always slower than the "analog computer".  We came to refer to pot's as "engineering elements".

There are definitely some folks around this thread who have a lot in common with Bob.  One of them just recently came to the conclusion after lots of calculation and simulation that a magic box that matches the parts of a wind system would be pretty useful.  I agree.

Problem is, I would rather buy one that is mass produced for a reasonable price than try to spend the next several years of my life prototyping the thing and building a company that makes them.  Maybe, just maybe, if "we" keep inventing this thing on a fairly regular basis, some existing MPPT company will notice it and start building them for us.

best,

jp

[Flux]

I think you will be waiting a long time for a commercial fit all unit that is affordable.

The actual component cost is not that high but the design will be expensive and the assembly will be horribly so unless there is a market for 10's of thousands.

The only real interest will be machine manufacturers and they may or may not feel inclined to sell it to someone who will likely run into trouble with it. There is not the same market as for solar and the simple approach I suggested would not work as a fit all scheme, you need to get the components right.

The true tracking micro controlled unit will be more expensive to develop and may still not work or may be damaged if used with unsuitable components.

Maybe that someone will do the development and be prepared to sell as a kit or something but success would not be certain without tight control on how it is used.

Flux

[finnsawyer]

Another possible fly in the ointment is the width of the power curve of the windmill for a particular wind speed.  The graphs of the power curve versus rpm that I have seen (not lately) were quite broad, meaning it was not easy to hit the sweet spot exactly, nor would it really be necessary to.  The result for your scheme could be an inordinate amount of "hunting" with the controller never really finding the best spot.  For a line of manufactured mills it might just be better to monitor the wind speed and adjust the controller by using a best power curve stored in a computer.

[MountainMan]

New half baked idea...

...what could be done with an off the shelf solar MPPT to fake it into doing what's needed for wind?

As I see it, you would most likely need to lengthen it's cycle time between change/re-measure.  Depending on how a particular unit works, this could be really easy (change out a capacitor), or it could be basically impossible, reprogram an undocumented FPLA or a masked ROM micro-controller.

Also would want to be able to control a dump load for high wind situations to avoid over-voltage on the input.  Perhaps this could be handled separately with an additional unit, but now we're adding on cost in a hurry and leaving KISS far behind.

As an alternative, what about computer control of a programmable charge controller?  Are there models out there that will communicate/take orders from a pc?  Might be a good use for an old laptop that is inching its way towards the dumpster.

jp

[Nando]

The MPPT controller has the potential capability of "ADJUSTING" the generator to the blades or the blades to the generator.

A good MPPT allows the wind mill to go as high as possible in voltage and of course the wind mill in RPM.

One may be able to attain around 85 % overall efficiency if the generator is not clamped by the battery bank.

So, like for a 48 Volts battery bank, the wind mill may go, as high as, 130-150 Volts ( or even more) and the MPPT controller varies the current up and down to set the mill to its best RPM for maximum power harvesting.

Also, the controller should be converting that higher voltage to the battery bank voltage.

The controller, as well, has to have the capability of regulating the maximum generated voltage to a certain limit and at that limit the controller should behave like a battery, clamping the rise of the voltage -- though at that point the Furling should occur or if with pitch controller the generator should be set at maximum limited power.

Nando

[SamoaPower]

I'm really surprised to see that this thread died out so quickly, given that it addresses what I would consider to be a major issue concerning DIY mills. Mainly, what's the most I can get for the least?

Actually, I'm surprised at the apparent lack of interest in a number of issues on this site that I would deem to be of primary importance to many who are interested in obtaining the most from their investment. Airfoils, component matching, and overall efficiency, to name a few, are touched on, but don't seem to be examined in depth. My congrats to MountainMan, Flux and others for stimulating thought on these issues.

Take airfoils, for example. It would seem that, from the posts, blade carving is very much an approximate issue that only needs to come close to an airfoil shape with little attention needed to detail. Initial lumber size seems to be the primary criteria. Yet, if you look at the performance data, mainly L/D, you will see that small differences in detail can have a large effect. Would you rather have L/D = 30 or 100?

Another area that astounds me is the apparent degree of reluctance to do something a "little" (or a lot) more complicated. Sure, the KISS principle has a lot of merit, but do you really want to go back to the days of the mechanical timer on your microwave and the '55 Chevy?  There's a lot more to be had if you're willing to learn.

[ghurd]

I believe most of those issues are covered in a fairly indepth manner.

'The most I can get for the least'

Don't underestimate what an extra $20 in magnets can make up for.

'Take airfoils'

There is also a balance of what most of us can actually make. Just because I can print out some fancy NASA design doesn't mean I can build it, or want to spend huge amounts of money and time trying. Often 6" of extra lumber can be just as good, if you see what I'm saying.

A '55 Chevy is better than walking. Did you see the recent post about the guy making blades with a chainsaw?  I bet his printer can spit out some very nice airfoils. We do the best we can with what we have to work with.

You gotta do what you gotta do.

G-

[SamoaPower]

Points noted and taken.

I still believe that if the uninitiated were aware of what they are giving up in the name of simplicity that they may have second thoughts. Efficiency should be a watchword in the vocabulary of all alternative energy enthusiasts. Just because the basic energy sources are "free", there are still real costs involved in obtaining that energy for practical use. Why accept 20% overall efficiency when 30% or more is possible? Enough of that - on to the thread.

Here's a link to a MPPT design that many could build for little money.

hppt://www.elecdesign.com/Articles/Print.cfm?ArticleID=6262

This is a low power design, mainly meant for PV, but the author acknowledges suitability to hydro and wind with minor changes. For most wind applications it would need to be up-scaled in power handling capability, which is certainly feasible. It uses a rather interesting analog approach to finding the MPP.

It would seem that we amateurs could benefit by paying more attention to what the "big boys" are doing. Utility class machines invariably use variable pitch rotors. The benefits gained are indisputable yet, we see little use of them in homebrew machines because it seems to be deemed by the gurus to be too complex to be implemented by most of us. Rubbish! If you can build an alternator, you can build a pitch mechanism. Being able to run at a constant TSR is a real asset. To maximize efficiency, large machines  control both the rotor pitch and the load profile. Using MPPT and variable pitch we can do the same.