Electrical matching - source and load.

[dyslexicbloke]

Has anyone any experience or knowledge to offer inn respect of any of the following?

Given that a high voltage energy source, alternator or PV array, is likely to suffer fewer losses due to conductor heating than a high current system of the same capacity I find myself wondering if there is anything in actively matching supply characteristics to the desired load.

This seems particularly pertinent with respect to wind driven alternators, which as I understand it, are usually a compromise.

Please correct me if I am wrong....

1. If an alternator has sufficient windings to produce a usable voltage at low speed then it will also have a high impedance which will prevent it from delivering that voltage at higher currents when energy is potentially available.
This would also result in overspend in any wind turbine that had no other form of limiting.

2.  An alternator with fewer and bigger windings will make use of high available energy situations but fail to produce any usable power, due to lack of voltage, at lower speeds.

If both the above are assumptions are correct then the ideal would be to have a low current system all the time but vary the voltage depending on the power available, the problem being that is the exact opposite of what is required for just about any useful load I can think of.

Whilst musing on this point and thinking about switchable alternator configurations it occurred to me that it should be possible to create a variable voltage and current converter. The ideal situation would be to take a constant current from the source whilst allowing the voltage to float and deliver a controlled, lower, voltage to the load whilst allowing the current to float based on the impedance of the load. (The latter half being essentially a standard charge controller)

I am not suggesting that anything would ‘make' power, in fact any such circuit or device would have losses and therefore use power and could operate only if some arbitrary level of energy was present, but there are potential options I think.
Anyone care to comment on any of the following:-

Transformer with multiple primary tapings.
Variable autotransformer (Variac)
Buck converter circuit.
Buck/Boost converter circuit.
Conversion using capacitors and high frequency PWM
Chopping power supply with wide input voltage range and variable output.

I had a quick look around the web and found two interesting things …
A well designed buck circuit can be very efficient, 95% or better.
MPPT modules exist to maximise solar panel performance, these must use a comparable strategy.

Lastly, and far more simply I might add, how about an alternator stator divided into at least 4 parts which could be switched from series to series/parallel to parallel.
This would give 3 potential load and speed curves that a machine could utilise based on the current wind condition (Level of energy available)

Thoughts anyone …. Has anyone tried any of this?
Al

[wooferhound]

I am confused as you seem to be referring to the battery as a load, and you are referring to the items that you are powering as a load?

It's hard to control the voltage going into a battery as the battery will hold the entire system down to it's voltage, so you can't actually measure or control the charging voltage, you can only control the amps for charging. The battery is a power soak, if you try to crank up the volts that are going into it, the voltage will not rise but the amps will. This is what charges the battery and as the battery charges the voltage of the entire system will rise very slowly.

- Here is my standard explanation of this effect -

The battery will always try to keep the voltage going into it, at the voltage level of the battery itself. The battery will be sitting at 13 volts, and your Wind/Solar power will be 22 volts open circuit. As soon as you connect your open circuit Wind/Solar power to the battery, POOF ,everything is now running at 13 volts. It pretty much doesn't matter what voltage you want to charge the battery with, Your genny may be making 65 volts open, but connect it to the battery and then everything is at the battery voltage.

What happens to all that extra voltage? It's converted to amps and that is where you start to worry about your stator burning up. Once your wind genny reaches Cut-In (battery voltage) the measured volts will be the battery voltage which should be increasing slowly as it is charging. As your genny speeds up past cut-in the voltage doesn't increase, but the amps do. Under most circumstances you don't have to worry about the voltage of the device you are connecting to the battery, your worry will be with the ability of the charging device to deliver the Amps without burning up.

When Charging a battery try to keep the current going in to less then 10% of the capacity of the battery. For Example: a 100 amphour battery should be charging at 10 amps or less.

[dnix71]

You are talking about "impedance matching." MPPT's do it by brute force, trial and error. Multiply amps times volts over the range of the panel's output and choose the highest number.

That's simple enough. The only thing better would be to charge a super cap and convert the power to something more efficient to charge the batteries, if such exists.

Higher frequencies than the common line 50/60 Hz give better lumen/watt output in some flourescent tubes so it pays to make a ballast that operates that way.

I don't know if anyone has tried it, but maybe high frequency dc pulses would be more efficient at charging a battery. High frequency high voltages pulses are used by some desulfators. But the batteries are connected to appliances and those might be damaged by ringing them with high voltage/frequency pulses, so it may just be safer to just do it the usual MPPT way.

There is no reason a good MPPT connected to a turbine wouldn't raise efficiency charging batteries, but home brew turbines put out wild a/c and sometimes high voltage so the MPPT to control that won't be inexpensive.

[jimovonz]

The mis-match you describe has been covered here quite extensively over the years. A soultion would not only improve electrical efficiency but potentially allow a better match between the power available from the blades and the load presented by the alt/battery.  Check out Flux's 'Matching the Load' diary here for a most enlightening discussion: http://fieldlines.com/board/index.php?topic=127288.0

[JeffD]

In 2008 I went the boost converter route to improve the load matching on one of my first little wind turbines.  The boost converter works from 2 m/s to about 6.5m/s.  The stator was wound for a cut in at 6m/s so that the main rectifiers kick in at about 6m/s and are on their own around 6.5m/s.  I used Flux's circuit that he provided back in 2006 but used a KA7500C PWM controller (same as tl494) running at 25Khz. all the parts I used for the boost converter came from an old PSU that I found at the dump.  Output current was used to control pwm duty cycle and works well once you tune the gain and offset of the error op amp.  An Atmel AVR atmega328 (was using 168 but recently upgraded) micro-controller monitors the boost converter and a number of the wind turbine parameters ie rpm, wind speed, output current, battery voltage, temperature and other stuff.



In Dec 2009 I built a synchronous buck converter for one of my other little wind turbines.  Instead of a dedicated PWM controller I used an Atmel AVR atmega328 micro-controller to directly control the buck converter.    I went with a micro-controller to control the buck converter because I wanted to experiment with a number of different algorithms for mppt for the wind turbine and be able to log the data for future analysis.  The circuit I used was based on the solar MPPT circuit I was using which was based on Tim Nolan's work: http://www.timnolan.com/index.php?page=arduino-ppt-solar-charger.  I didn't use the ir2104 high voltage half-bridge n mosfet driver for the wind turbine buck converter but used a FAN73832N driver.  I used a FDP39N20 n-mosfet for both the high and low side.  All components on the input side of the buck converter were rated for 200v or better.  The alternator output at the diodes is about 80vdc just before furling (13m/s) but I have seen it jump to about 120vdc (on the input capacitor bank to the buck converter) when a very strong gust hits (15-18m/s).

In having used both schemes for a while now, I am sticking with using a buck converter for now.  Efficiency is higher with the buck converter in the very low wind speeds and at the top end which is what I wanted.  There are a lot of pluses for using the boost converter though.  The boost converter was much easier to set up initially and handles a considerably smaller amount of the wind turbine power than the buck converter so possibly less prone to failure.  The max voltage on the transmission line from the turbine alt to the controller is much lower for the boost converter than the buck converter.  And if the boost converter stops working then you still have the main rectifiers for the mid to high winds.  The down side to the boost converter is the very big impact the rectifiers and transmission line impedance have on output efficiency in low winds especially when charging into a 12v battery system.  

The big plus I found with the buck converter and the micro-controller marriage is the greater flexibility in the control of the power output.  The main advantage in using the buck converter (over the boost converter) is that the rectifiers, stator impedance, and transmission impedance have a much lower impact on power transfer efficiency for a 12v battery system.

I haven't tried the other schemes yet for trying to match the load.  I went with boost and buck converter methods because I wanted to learn more about how they worked and how they could be used to improve a wind turbine's power output over a wide range of wind speeds.

[scoraigwind]

As for controllers, Midnite solar have the classic and Outback power have the flexmax (although they do not recommend this for wind yet).  Flux has built several such converters and reported them on this board. 

Switching series/parallel tends to be a bit drastic rather than the smooth transition that you need.  Star/delta s popular but again a bit drastic.

A company here in the UK called marlec used transformer tappings since a lot of years ago (fm1800 was the product). It worked OK although sometimes the analog circuit was a bit unreliable.

[dyslexicbloke]

Great info folks thanks ....
I haven't decided yet weather to go with home build or off the shelf modules but I expect, realistically, that off the shelf will be far better than I could build.

I an confident that I could program a micro, PIC or similar, to do what is required its the analogue bits that worry me.

is there a buck converter out there which is programmable or will accept an external reference voltage to set its output voltage?
Am I wrong in assuming that a buck converter, wide input voltage, will do the job if its output voltage is controlled based on an MPPT algorithm?
Al