# I keep reading about MPPT... for wind generators. Here's an alternate idea.

Registered Users Posts: 13
For those of you who are into electronics, hi-fi, radio, etc, you're all aware of load matching for any amplifier. For those who don't quite understand, let me demonstrate: If you'd like to charge a 12v battery from 115VAC, you need to rectify it, which gives you a peak DC voltage of actually quite a bit more than 115V. In reality, the absolute peaks are more like 160v. If you just use a resistor to limit the current to a 14v (charged lead acid) battery, you're going to need something along line of about 10 ohms to limit the current to closer to 15 amps. Except that you're going to dissipate 1500 watts across the resistor, and 150 watts into the battery - roughly. This is extremely inefficient.

MPPT controllers attempt to find the PWM load limit - by finding a maximum amperage that doesn't cause a voltage knee - and use a form of voltage step-down to convert high voltage low amperage sources into low voltage high amperage source. ( it's done with high speed switching ) Wind generators are often made by using many turns of finer winding, so that they develop useful voltage at low rpm... and low current. But as the rpm climbs, the voltage doesn't rise, and the current drawn causes much power dissipation in the generator windings - which turn into the equivalent of the resistor in series with the battery.

However, the PMA's produce AC, and PMG's produce DC, and the one thing we CAN measure, and will remain constant, is their apparent impedance. If we think of a wind generator as an amplifier or transmitter, we can see how we could use a switching power supply to match whatever voltage / current ratio the generator is wired to produce. For the most part, this ratio will remain mostly unchanged from "startup" to "full power". Ergo, if we match the load to the characteristics of the generator, the best match and the best and most efficient power generation will occur - at any speed. If this were an AC alternator, putting AC into a transformer, we could use a variable ratio transformer to keep the generator and constant voltage load well matched. Of course, that's not really a viable option, since automating a variable transformer system would be, well.. A challenge not worth fighting.

However... we have an easier option. And it works like this: If we use a PWM load, by simply using a high speed switching load, we can full wave rectify the output (the more phases the better), smooth it with capacitance, and then use a PWM load which is "tuned" by use of percentage of "on" vs "off" in the square wave, to approximate the appropriate voltage / current load, and by use of a very wide range switching voltage regulator on the output of that PWM load, we can then have approximately a 90-95% efficient conversion to the battery bank's required voltage. There is no need to "track" the generator's or alternator's output, we merely need to know it's electrical characteristics at operating speeds (mostly accomplished by measuring internal impedance) and tune our converter to match it, and then tune the output voltage to the highest voltage our load can handle.

Well, this is my theory, and I'm sticking to it. I haven't the means of testing such a theory... some of you might be BSEE's or even MSEE's and might find a fly in the ointment, but it sure seems sound to me. Any of you techie type DIY'ers find this interesting?

Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.

Mark, I am not quite following you... But there are MPPT type charge controllers that can support Wind Turbines too. So at high RPM you can get much higher voltage and still keep the current down (and the windings cool). (MPPT controllers typically use buck mode switchers for efficient down conversion of voltage).

One vendor, Midnite Solar, makes an entire kit (with programmable MPPT function vs RPM and load bank+control). And another person here uses MPPT charge controllers from MorningStar with his hydro-turbine system.

During testing, "halfcrazy" (works at Midnite) said that they saw upwards of 3x the amount of energy gathered running a wind turbine with their MPPT controller vs the standard Turbine/rectifier/battery connection (as I recall).

So--If you are looking for load curve matching--It is available now (not cheap). Midnite has their own forum if you want to research their wind turbine systems (I am sure they have more there than at our site here).

-Bill
Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
• Registered Users Posts: 13
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.

Hmmm... Ok, let me say it simply... an MPPT controller has to be either matched to one specific device, or it has a mechanism inside that attempts to guess at what it should be doing. While this, if done with perfect algorithms (and it can't be, when guessing what it should do), is more efficient, my idea has the advantage of being much simpler, where you adjust the the apparent impedance of the load manually, and the nominal output voltage, and it operates in that mode, using what are basically commodity electronics, doing what are common functions, and it needs no programming, and is not complex to build, and is a set once and forget device.

It does require that you figure out what your load should be, either by experimenting (run your generator with a motor and simply adjust to an efficient load) or else measure it's electrical characteristics and adjust. It would work with just about any wind generator now being produced and for things like many of the PMA's now being sold cheap, will have a very serious increase in their efficiency. And it could be a generic device, made for real cheap.

Yes, "switching voltage regulators (or converters)" would be buck mode switching voltage regulators. They exist now as commodities all over. And the PWM load, with variable percentage is also a very "generic" thing to construct or produce in quantity. The only consideration is that the voltage range for input would be very wide, by comparison to most commodity power supplies.
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.

The Midnite options run from 150/200/250 VDC maximum input (in theory, it is Vmax+Vbattery before damage would occur). And they accept programming (RPM vs whatever--I have not researched their product as I have no need for one) so it will match the turbine it is connected to.

Xantrex has a 600 VDC max input version (don't know if it would work with a wind turbine or not--probably was not designed for that function).

In any case, need to control for maximum RPM and shutdown conditions--Cannot just "unload" the electrical load in high wind or when the battery is "full"--For horizontal axis wind turbines, they will self destruct when unloaded (unless secondary brakes/feathering/furling/etc. is active).

-Bill
Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
• Registered Users Posts: 13
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.

If you have the money, and if you want to spend thousands, there are certainly great options. But that wasn't the point of this. The point of this was to see if we can come up with a generic device that makes our homebuilt, or DIY or cheaply constructed devices produce a lot more power, without spending all that money.
• Solar Expert Posts: 995 ✭✭✭✭
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.

Mark,

May I suggest that you also post your idea to one or two of the DIY wind turbine and wind MPPT forums and see what
they have to say about it. There are very few DIY wind people here and I think you may have a bit more technical
discussion on one of these other forums.

http://www.fieldlines.com/index.php

http://www.anotherpower.com/forum/

Thanks !!
boB

PS... Could you maybe draw a small diagram of your basic hookup ?? i.e. What is connected to where and how your PWM is connected into the system ?
• Registered Users Posts: 13
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.

Well, let me expand a little and see if it doesn't get a little clearer. And, as for a diagram, it's simply a device that goes between your kill switch (if you have one) and your batteries. It's not a charge controller, and not meant to be one. Nor a dump load controller (though it could be added).

As we all know, the biggest headache with wind generators, is that to make useable power at slow speeds, the windings have been built for low current, high voltage operation. That way when your wind is low, it does more than just spin idly. Everyone knows that for PM designs, the finer and more windings you have, the higher the voltage you get at a given rpm. If you build a generator that produces sufficient voltage to be useable at 8mph (like 15v) and say 125 rpm, when the wind hits 25, the voltage isn't 14 anymore, it's more like 80... But if you tie it to a 14v battery system, the generator is still TRYING to produce 80 volts, but since the batteries are limiting the voltage, the windings will be dissipating 66 volts! That's right, HEAT production. Now, somewhere, not too far below 80 volts no load, will be the maximum power production. That point, in terms of percentage of no load voltage is... the same, for all intents and purposes, across the rpm range. There is no real need for an actual MPPT controller - the MPPT controler attempts to figure out what that load is, by making educated guesses about the device producing power. But a generator doesn't need educated guesses, it merely needs a proper load, and little else.

By using a PWM (pulse width modulation) load, you can mimic a variable resistance load. That is, you can adjust it to a pulse width percentage of "on" that will mimic a load that matches the windings in your generator. So, this device will use a high frequency on and off cycle, combined with some inductance and capacitance, to appear to the windings in the generator, to simply be a matching impedance load. So, your best generators will be high voltage, low current. And by high voltage, I mean 100 to 200 volts at max rpm, more if you're going to use industrial strength generation and wiring and insultation. All the load matching device needs to know is: What is the maximum power output, and at what voltage does that occur, and what is the desired battery voltage? From those three bits of information, you now have a device that can extract MUCH more power from a wind generator than it would otherwise, and it will run incredibly cooler doing so.

This device would consist of a tiny microcontroller that reads the input and output voltage and adjusts the maximum pulse width up to much higher values as the generator slows to near the point where the generated voltage matches that of the battery system, and is programmed to determine the default pulse width via the two inputs.

Years ago, the odd folks at hydrogen appliances built a cheap wind generator and a high voltage PMA that went with it. It would begin providing postitive current at a few mph breeze, but not even in a gale would it produce much more. The reason? All the power the generator was making was simply being converted to heat inside the windings of the PMA.

All the generators you can buy, cheap or otherwise, are electrically inefficient... or are wind inefficient, because they produce power that requires a variable voltage load, and we statically load it with a fixed voltage load. By creating a matching load, rather than a fixed voltage load, most generator systems could probably produce anywhere from 2 to 5 times as much power as they do. You can even compensate for oversized generators and undersized blades, by simply artificially reducing the apparent load impedance, allowing the generator to have relaxed torque requirements so the blades aren't kept at stall speed or less.

This device would be, electrically, into 3 modules... 1 would be a PWM load, that has a square wave output. 2 would be a voltage to current PWM type converter on the output stage. This would convert the square wave, variable voltage output of module 1 into a current source to charge batteries, and provide a consistent load to make the 1st module work correctly. And 3, a microcontroller that accepts inputs from the user for maximum voltage and power ratings for the generator (or a guess), and the desired nominal output voltage for your battery bank.

The "software" would be very simplistic. I can write it, for that matter, and I'm not a programmer. This device would be "generic" meaning as long as the maximum currents and voltages for the electronic parts are not exceeded, it will work with any generator ( either dc generator or rectified alternator) wind or water.
• Solar Expert Posts: 995 ✭✭✭✭
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.

Markk !! This is GREAT stuff !!! I think you've hit the nail right on the head !!

Sounds pretty darn simple actually.. Just build it up and test it with a hydro or wind turbine.

I can't wait to see how much more power it can extract from all that wasted head in the windings !

boB
• Solar Expert Posts: 3,009 ✭✭✭✭
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.

It's been a long day for me, so perhaps I'm missing something - - but I'm looking for where, or how, the PWM device takes all that extra voltage, and converts it into extra amps going to the battery, like a MPPT controller would.
• Solar Expert Posts: 995 ✭✭✭✭
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.
It's been a long day for me, so perhaps I'm missing something - - but I'm looking for where, or how, the PWM device takes all that extra voltage, and converts it into extra amps going to the battery, like a MPPT controller would.

Well, there is a sort of "poor man's MPPT" for wind where you add a fixed value power resistor in-between the turbine and battery.

It semi unloads the turbine to allow it to spin closer to the proper TSR. Not as good as a real MPPT but is a compromise.

boB
• Registered Users Posts: 13
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.
It's been a long day for me, so perhaps I'm missing something - - but I'm looking for where, or how, the PWM device takes all that extra voltage, and converts it into extra amps going to the battery, like a MPPT controller would.

I can't write you a good description off the top of my head, but if you use google to research "switching voltage regulator" or "switching power supply". First, the "load" on the generator will consist of a very rapid on-off switch - several thousand times per second. Imagine you have a dc motor with a flywheel, and you can switch it on and off very, very fast, the motor will simply appear to run at a steady speed... a speed which will be relative to the ratio of "on" vs "off" time, and will act as if we've converted the source voltage to a lower one. The average voltage and current will be less than if it were just left on. By using a several khz switching rate, the system acts as if it can actually transform voltages. In reality, it IS, with an efficiency of well over 90%. This, by the way, can work in both directions. This switching system can actually RAISE the voltage, as well.

How does it convert high voltage / low current to low voltage / high current? Well, if you have a source of 100 volts and 5 amps, you have 500 watts. When this source is drawn upon by a switching regulator, the source charges a capacitor, which then discharges at very, very high amperage, for a brief period. This charges another capacitor, from which the load is drawn. By using extremely small percentages of "on", the system keeps the supply at a very close to steady voltage. But the current flowing into from the 100V source is very small - only 5 amps. The output is 10 volts, and it draws 50 amps - which is 500 watts. The same amount of ENERGY is going in as coming out, but because of how current flows and inductors / capacitors charge and discharge, it has the effect of converting dc voltages with well over 90% efficiency. In reality, if 500 is going in, likely about 475 watts or so will be coming out, with 25 watts of heat being generated by resistance of the devices carrying current, be it wires or semiconductors.
• Solar Expert Posts: 995 ✭✭✭✭
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.
markk wrote: »
I can't write you a good description off the top of my head, but

Sounds to me like you are trying to describe a basic "buck converter" which most MPPT controllers are made of.

Including wind MPPT controllers.
• Solar Expert Posts: 3,009 ✭✭✭✭
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.

• Solar Expert Posts: 995 ✭✭✭✭
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.

Here is an old graph of power output improvements from a fairly small turbine connected
directly to battery, (bottom curve), with the resistor inbetween (middle graph "R matched),
and real MPPT (top curve).

This data is from Flux on the otherpower forum from years ago.

boB

• Registered Users Posts: 13
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.
boB wrote: »
Sounds to me like you are trying to describe a basic "buck converter" which most MPPT controllers are made of.

Including wind MPPT controllers.

The "T" at the end of MPPT means "tracking". It consists of a mechanism to search for the optimum point. Necessary for solar arrays, becuase they tend to have a constant voltage characteristic. But a wind generator really doesn't need "tracking", since it is a relatively linear device. It just needs to know what the load should look like, and then convert it to the right voltage. Solar panels are not linear devices. They produce different ratios of current to voltage, but not a generator that converts mechanical power to electrical.

By getting rid of the "tracking" and instead, simply making a reasonable guess at what the voltage could be, you pretty much get the same as a very complex MPPT tracker, for a fraction of the complexity. Tried selling Nando and several others on the idea a few years ago, but you c an't tell the experts anything, they know it all.
• Solar Expert Posts: 10,300 ✭✭✭✭
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.

sorry mark, but wind turbines are far from being linear because the source of the power (the wind) is not linear.
• Registered Users Posts: 3
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.
markk wrote: »
...If we think of a wind generator as an amplifier or transmitter, we can see how we could use a switching power supply to match whatever voltage / current ratio the generator is wired to produce. For the most part, this ratio will remain mostly unchanged from "startup" to "full power". Ergo, if we match the load to the characteristics of the generator, the best match and the best and most efficient power generation will occur - at any speed....

However... we have an easier option. And it works like this: If we use a PWM load, by simply using a high speed switching load, we can full wave rectify the output (the more phases the better), smooth it with capacitance, and then use a PWM load which is "tuned" by use of percentage of "on" vs "off" in the square wave, to approximate the appropriate voltage / current load, and by use of a very wide range switching voltage regulator on the output of that PWM load, we can then have approximately a 90-95% efficient conversion to the battery bank's required voltage. There is no need to "track" the generator's or alternator's output, we merely need to know it's electrical characteristics at operating speeds (mostly accomplished by measuring internal impedance) and tune our converter to match it, and then tune the output voltage to the highest voltage our load can handle.

Well, this is my theory, and I'm sticking to it. I haven't the means of testing such a theory... some of you might be BSEE's or even MSEE's and might find a fly in the ointment, but it sure seems sound to me. Any of you techie type DIY'ers find this interesting?

Hi Mark, interesting concept you have there. I think what you're describing though is basically what an MPPT controller - be it for wind or for solar - does and that is effectively impedance transformation. Think of a buck converter; on the input say you have 48v at 10 A and it's converting that down to 24 v at 20 A (assuming no losses). R = V/I, that's effectively transforming 4.8 ohms into 1.2 ohms. As you probably know, the max power transfer occurs when the source impedance matches the load impedance and in our case, the source is either wind or solar and the load is the battery being charged. This is the same concept as matching the load to an amplifier or an antenna to a radio.

However, a wind turbine driving an alternator is not a fixed impedance at all speeds, and this is where a real MPPT tracker would get the power gain from. As I recall, the voltage out of an alternator is more or less proportional to its speed; but when tied directly to the battery, the voltage is held down to whatever the battery's voltage is and the current will then increase as power output continues to go up. This is a change in impedance as far as the alternator is concerned, so it's no longer running at max efficiency.

So I guess I don't see the difference between what you're describing and what a commercial MPPT controller does. You mention there is no need to track the alternator's output, but that's the only way you're going to get any real gain. Without the tracking, it's just a static impedance transform, which (aside from being able to run the alternator at a higher voltage and save some line loss) could just as well be accomplished by winding the alternator to match your battery bank voltage in the first place. The practical way to do this tracking is to monitor the voltage and current coming from the alternator and adjust the switching rate of your converter to maximize their product.
• Registered Users Posts: 13
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.

So I guess I don't see the difference between what you're describing and what a commercial MPPT controller does. You mention there is no need to track the alternator's output, but that's the only way you're going to get any real gain. Without the tracking, it's just a static impedance transform, which (aside from being able to run the alternator at a higher voltage and save some line loss) could just as well be accomplished by winding the alternator to match your battery bank voltage in the first place. The practical way to do this tracking is to monitor the voltage and current coming from the alternator and adjust the switching rate of your converter to maximize their product.

Hi, sorry for the long delays... I run a business by myself these days and often find myself unable to pursue everyhting.

A wind generator IS a fixed impedance device. Not wholly, but enough so that efficiency losses are minimal. There is no need for the "tracking" part at the end, except that with a "tracking" algorithm, in theory you don't need to tell it what electrical characteristics the generator or load has.

Your power output is linear with rpm, and voltage is linear with current, as well. In other words, over changing wind speeds, the power production lines are quite straight, not curved much. the curves are near the bottom, because below a certain voltage there's NO power produced, even though it may be spinning and producing some voltage, there's no current, ergo, no power.

I guess my point in all this, is that we could make use of high voltage, low amperage, cheap wind generators that are mechanicaly sound, but produce horribly little power, due to the fact that we load them with batteries. We tie a contant voltage, variable current load to a variable voltage generator. That's like the least efficient method possible.

All we need is someone with a little electronic device design experience, who has experience to choose the parts and circuitry for the high amperage parts, and we're mostly ready to go. This would be, in my view "open source", meaning a collaborative design, published and free for anyone to duplicate, revise, improve. And, with availability of commodity electronics parts from Asia, cheap and easy to make, for those of us who tend to DIY most things.
• Banned Posts: 17,615 ✭✭
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.

Actually wind turbines are horribly inefficient due to the nature of the beast. The mechanical losses vary with wind speed vs. prop design vs. wiring. A design most efficient at one speed will be far less efficient at any other. Can't be helped; physics. There is no linearity to this either.

You are quite correct about turbines being inefficient when connected to batteries, for the batteries will only take the power they need. The same is true of solar panels. Some time is inevitably spent producing no power because the batteries are full. No way around that either; "opportunity" loads can only be automated to a certain extent. This is why grid tie systems, whether for PV or wind, are more efficient: there is always some place for any power than can be generated to go.
All we need is someone with a little electronic device design experience, who has experience to choose the parts and circuitry for the high amperage parts

Well boB fits that description very well, so I'll take his view on the matter.
• Solar Expert Posts: 995 ✭✭✭✭
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.
markk wrote: »

All we need is someone with a little electronic device design experience, who has experience to choose the parts and circuitry for the high amperage parts, and we're mostly ready to go. This would be, in my view "open source", meaning a collaborative design, published and free for anyone to duplicate, revise, improve. And, with availability of commodity electronics parts from Asia, cheap and easy to make, for those of us who tend to DIY most things.

I mentioned this before but there is already DIY MPPT dialog going on over at these forums...

http://www.fieldlines.com/index.php

http://www.anotherpower.com/board/index.php

http://www.thebackshed.com/forum/home.asp

These are some pretty smart fellows.

boB
• Solar Expert Posts: 995 ✭✭✭✭
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.
markk wrote: »
The "T" at the end of MPPT means "tracking". It consists of a mechanism to search for the optimum point.

In the case of wind turbing mppT, the "T" part has already been done, doesn't change (OK, some with air density etc) and this Pre-Tracking
is remembered by the power curve that the MPPT buck converter follows. The "Tracking" is the "Following" of the curve in this case. You could
pick a different word I suppose.

Turns out that trying to actually find the maximum power point in real time does not work very well for wind turbines for various reasons.

As for finding someone to "choose the parts and circuitry for the high amperage parts", maybe that's the hidden 90% part that makes this
not so simple or as inexpensive as you might think it is to do.

Check out those other sites I listed and see how well they are all doing. There are schematics and ideas there about this very subject.

boB
• Solar Expert Posts: 3,123 ✭✭✭✭
Re: I keep reading about MPPT... for wind generators. Here's an alternate idea.

Something which comes up on the wind-generator threads a lot is this:

If you design the turbine so that it depends on the load of the inverter to prevent runaway in high winds, beware. If the wind reaches the point where the inverter cannot take any more power and shuts down to protect itself (or just fails), all of a sudden there is no more load on the turbine at the worst possible time. That is where the dump load comes into play.
It takes time to turn the turbine out of the wind and mechanical brakes just cannot dissipate enough energy for the time required.

Just designing the electronics for safety is only half the battle. And some imported turbine packages are not "designed" at all, just assembled.
SMA SB 3000, old BP panels.