How to combine Wind and PV generators?

[falexandru]

A. Scope

My project consists in one PV generator and one Vertical Axis wind turbine generator, combined to charge the same battery.

The combination serves to demonstration purposes - to show kids how can we convert the renewable energy into electricity.

The use of electronics is limited to diodes and bipolar transistors of general use. No logic circuits, no integrated circuits, no ready-to-mount modules.

+++
B. General characteristics

The battery is made up 3 units of AAA NiMH 1.2 V, 800 mAh, of general purpose.

The PV panel is a 6 V one.

The vertical axis wind turbine is to be made, design is based on SAvonius, Darrieus, giromill - H-type- all to be decided.

The (combined) generator has to charge the battery and the battery is to light an white LED.

+++
C. Status of work

• The PV charger is completed, tested in field - OK.
• The Wind turbine : only the Joule thief circuitry is tested on prototype desk - OK.
  The schematics attached is not yet tested on desk prototype

+++

D. Combine the 2 chargers

This is the part I would very much appreciate your help and kind patience, which were of crucial relevance for me to get to this point.

Scratching head questions:

a) Since the wind generator can supply only very little voltage (up to 5 V, my guess) and since the Joule thief needs a constant voltage, that the Wind generator cant offer, is the solution I imagine (see the picture) a feasible one?

The 1 regular Diode is plan to offer (more or less) constant 0.6 V, while the 2 in series are expected to offer 1.2 V. The next Zener is 2.1 V (if I can find one) and the next one is 3.1 V.

The manual (at this stage) switch shall select the right diode path according to the wind speed (and provided voltage).

Are the concept and circuit correct/feasible?

b) The 2 circuits are separate at this stage since they are to be developed separately.

Can some problems show up in the moment the 2 circuits will power the same battery?

+++
At later stage, one 3.3 V Arduino board is planned to control and comand the combined generator. In fact, the Arduino will operate the SWs that are now manual, according to a code and info gathered by sensors.

[falexandru]

Parts:

I. Vertical axis wind turbine generator (under development)

1. C1 = 10 uF/25V
2. C2 = 1 uF/ 16 V
3. D1 and D8 = general purpose Schottky diodes
4. D2, D3, D4 = general purpose diodes
5. D5 = 2.1 V Zener diode
6. D6 = 3.1 V Zener diode
7. D7 = 4.7 V/1.3 W Zener Diode
8. R1= 1kOhm/0.25W
9. R2 = 100 Ohm/0.5W
10. T1 = 2N3904
11. H1, H2 = manual torroid 13 mm diameter, 30 cm 2 pieces of 0.5 mm wires
12. D9 = white LED
13. SW1 - 2 postions Switch ON-OFF-ON
14. SW2 - 4 positions switch (selector) - or 3 cascade mounted 2 position switches
15. LED socket
16. Connectors to the generator (DC barrel or any other 2 poles connector)
17. veroboard (stripboard)
18. wires (various)

II. Photovoltaic (PV) charger (this is completed and field tested - OK)
Parts:

1. PV panel 6 V 384 mA 1 pc
2. 4.7 V/1.3 W Zener 1 pc
3. 1 N 5819 Shotky 1 pc
4. R1 resistor 10 Ohm - 1pc
5. 1.2V AAA NiMHd 800 mAh units in series - 1 pc
6. case for 3 x AAA batteries - 1 pc
7. 560 Ohm R2 Resistor - 1 pc
8. SW1 ON-OFF-ON switch - 1 pc
9. SW2 ON-OF switch - 1 pc
10. 5mm LED white - 1 pc
11. Case for 5 mm LED - 1 PC
12. morfets 2 connections - 2 pcs
13. barrel connector to board - 1 pc
14. veroboard (stripboard) - 1 pc
15. wires (solid core) - various

One PV generator and one Vertical Axis wind turbine generator, combined to charge the same battery.

The combination serves to demonstration purposes - to show kids how can we convert the renewable energy into electricity.

[mauried]

You really need to provide 1 simple piece of information.
What voltage does the wind turbine produce for a given rpm , and how will you spin the turbine at that rpm.

[falexandru]

@mauried

From the experiments I have done so far the charging circuit starts to provide above 2.4 V from the input voltage of 0.85 V- in most cases and 0.56V if the circuit started at higher voltage.

The Joule thief needs constant voltage but the Wind generator cant provide constant voltage, because it varies with the wind speed.

To have a constant voltage starting from as low as 0.6 V and still be able to operate at higher voltage, I designed the 4-paths voltage "regulator" - in the red square.

I do not worry to much at this stage about the precise value of the voltage provided by the turbine generator, because I can easily add more Zener paths to the circuit and/or remove existing paths.

From 3.1V I can find regular Zeners. The point is that I need constant voltage bellow the 3.1 V.

+++

In the next days I am going to use the portable oscilloscope to get more accurate data about the relation between the input voltage and the output voltage in the Joule thief circuit in schematics.

By doing these measurements I want to find out at which input voltage the Joule thief circuit is able to provide 4.7 V, which is the required voltage to charge the batteries.

[wvmarle]

I like this project, sounds very useful and interesting to me.

For your voltage, you may want to look at reference diodes; 2.495V is a very common value for those. They don't carry much voltage, though: a few mA usually. I suppose you may be able to use one of those to control a BJT's base level and that way syphon off excess voltage - which I understand is what you're doing.

That joule thief is basically a simple boost converter, indeed the output voltage is directly dependent on the input voltage. But does that really matter when charging a battery? Or when powering an LED for that matter? Both need a certain amount of current, and a voltage that's big enough to create such a current, especially for NiCd or NiMH type batteries.

It does sound kinda inefficient, though, with quite some current leaking away through the transistor. Makes me wonder if some AC coupling could help to increase this efficiency (after all it's 50 kHz and not that much current).

[wvmarle]

Reading a bit on using a joule thief as battery charger I came up with this circuit. Should work fine for any voltage from 0.5-1V up. The instructable adds a zener to the base of the transistor, I left that out as I don't understand the circuit well enough yet.

There should be no need to worry about the input voltage. It's current that matters most here. The diode and cap (1µF is a guess, may be too small) stabilise the voltage and smooth the current for the battery charger. For powering a diode, you may use a constant current circuit (can be built easily around two NPN transistors). If the voltage is high enough you may even have two or three diodes in series.

[wvmarle]

For the solar circuit, is that zener D10 really necessary? Same for R3, it just makes you lose energy. I don't see it in any similar circuits, just that diode D11 to stop current from flowing back into the panels - particularly for small, 1-5W panels that don't need power management. Otherwise an LM317 regulator can be used as current limiter to charge the batteries.

Safest is probably the good old NiCd batteries, as they're the most forgiving when it comes to charging.

The wind part could be a great "maker" style project. Have kids try and design the best wind turbine to drive the generator.

[falexandru]

@Wvmarle

Some fast answers:

• D10 in solar circuit is for cases when the Sun brights so much that it goes up to more than 6V - resulting in perhaps damaging the NiMH units; also good in delivering a constant voltage to the batteries which - as far as I understod - may shorten the life of the NiMH units.

• similar protection role plays the R3 which is very small anyway (10 Ohm); I suspect the "regular" PV projects assume no significant variation in voltage or just dont care about that (this is my supposition, no evidence though); however, the first prototype delivered safely enough current to light the LED without the R3, which I added later (no test in superbright sunlight however).

• Cd units are banned in European Union (or at least no new units are allowed)

• mounting a small capacitor in parallel to the batteries seems to be well-spread - but it is really necessary? The point is that I have to explain to the kids why it is there, I am not sure if I can :-). Something like "this is like a bucket that takes the surplus or pour it to the batteries when the voltage instantly decreases - I guess it may be intuitive.

+++

In the schematics you posted the 1k resistor is between the coils the transistors is somehow "bare-mounted" - I think the voltage between the collector and the base is now higher than in the first schematics, because the resistor is in AC regime. Not sure, however.

+++

I fully agree the circuits are not efficient, I dump a lot of power along the way. I am afraid that is the price of keeping things simple enough for everybody to deeply understand.

Thanks for the brilliant idea to let kids play around the wind turbine design! I am very curious how they will came up.

+++

I cant pick your point in the AC- 50 Hz added part. What do you mean, please?

+++

Thank you for your support and appreciation to the project! That means a lot to me - really.

[wvmarle]

falexandru:

• mounting a small capacitor in parallel to the batteries seems to be well-spread - but it is really necessary? The point is that I have to explain to the kids why it is there, I am not sure if I can :-). Something like "this is like a bucket that takes the surplus or pour it to the batteries when the voltage instantly decreases - I guess it may be intuitive.

I think C3 is a good idea to keep the voltage and with it the current through the batteries constant. This circuit produces pulses at a high frequency and high voltage, the cap smooths this. Just like the caps used in an AC to DC power supply.

In the schematics you posted the 1k resistor is between the coils the transistors is somehow "bare-mounted" - I think the voltage between the collector and the base is now higher than in the first schematics, because the resistor is in AC regime.

This doesn't seem to matter. I see this resistor on both sides of the coils.

Over the past hour I've done some more research on joule thieves, and noticed a VERY important thing: the coils of the transformer must be connected in reverse phase. So that part of my circuit is wrong, connected as sketched it won't work. One of the sides of the transformer has to be reversed in direction. That's key to the whole principle of the joule thief.

The cap C2 (C3 in my circuit) you placed at the wrong spot. It's from the connection of R1 and the coil to GND, as its purpose is to filter peaks. This cap supposedly makes the circuit much more efficient. I've seen this another article as well as the instructable, interestingly both used a film cap rather than a ceramic one but did not mention this in the description. Usually ceramics work much better for such filtering jobs.

I cant pick your point in the AC- 50 Hz added part. What do you mean, please?

50 kHz, you missed a k here. Reportedly the addition of C2 brings that even towards 300 kHz.

AC coupling can be a way of stopping DC current, while letting AC through. A cap acts as resistor (it's called "reluctance" or "impedance") in an AC circuit but it does not dissipate any energy: the energy stored is returned to the circuit during the opposite phase. Well, that's an ideal capacitor, in the real world there's of course some resistance in the wires and the capacitor plates, plus some stray inductance.

At these high frequencies a quite small capacitor has very low impedance. A 1µF cap at 300 kHz has an impedance of just over 0.5Ω, 3.2Ω at 50 kHz. A 2µ2 or 4µ7 cap has negligible impedance at those frequencies - just beware to use a non-polarised one, preferably ceramic as it has much lower stray inductance than film.

Possible points to insert these caps would be at the collector and the base of Q1.

Updated circuit with comments added:

[wvmarle]

Interesting to see there are quite a few wind kits out there, such as this one. Even small hydropower generators.

What do you use for generator? Just a DC motor?

[falexandru]

First attempts:

a) regular toy motor - can barely rises voltage to 0.3V -measured by a multimeter and also not able to light an white led while mounted directly to it

But they are many types of "toy motor" - perhaps I shall keep searching.

b) computer fans

That is very confusing - a friend of mine lighted an white LED using one of them at relatively low speed (blowing on it :))). While another one, on 5V can light the white LED only by blowing a monster hair dryer into it

Perhaps going for low speed PC fans (the silent type ones) would do the job, but I couldnt find one so far.

Most exiting way to go: brushless motor. I brought one - but I have to build a sort of "convertor" to DC - the simpliest one I could find uses 6 friendly Schottkys. I am not familiar with this type of motor, so I have to do some research first.

++++

I dont even bother to place the dots on the coils at this stage, so I got your point about the transformer anyway. The ferrite rings have no DS here, so I must either trial-and-error them or find an alternative way - regular transformers for instance.

I brought a small 220 V transformer, which at least has some labels on it and I can buy more from the vendor. Still no DS available.

Apparently, nobody cares about the inductors

[wvmarle]

I just tried a DC motor (from an air pump, brushed type): turning one direction I easily got 3-5V out of it (I connected it to an electric drill to turn it), the other direction nothing significant. Didn't know they're directional. Of note: it was a sawtooth shaped voltage that came out of it (basically an open circuit, just scope attached).

BLDC motors indeed need six diodes for the three phases. A good cap to smooth the signal may also help the joule thief a lot.

[falexandru]

I think the output voltage depends on the rpm for which the motor is designed and on some other factors I have no idea about.

The only thing that I want from my motor (apart from providing an usable voltage) is to be easy to find.

Now I am thinking that I was wrong to look for motors at low voltage, assuming it might be easier to get usable voltage out of them than from higher voltage ones. At lower speed than the one for which is designed I mean.

+++

I found this (quoted from Charles Cowie in another forum):

"Brushless DC motors are usually motors that have permanent magnet rotors. It would be extremely unusual to find any other type of motor described as a brushless DC motor. All such motors can be used as generators, but some designs are easier to use as generators than others. A major example of a difficult motor is a BLDC fan motor found in a computer. Those have electronic circuitry built into them that must be removed or disconnected in order to use the motor as a generator. You might find some other design described as a BLDC motor that would be difficult to use as a generator, but most of them only require the shaft to be turned to produce AC at the terminals and a rectifier added if you want DC."

I cant get the point in "the shaft to be turned" - what shaft?

[wvmarle]

Motor shaft. The thing that sticks out and can turn.

[falexandru]

Ok, I got it now, thanks!

I am going to try the brushless motor - this one:

There is no DS available.

So I will mount the 6 Schottkys and start experiment. I hope I can get some 1-2 V at some bellow 20 rpm (assuming a reductor between the shaft of the motor and the turbine axis).

In parallel I will dig into my boxes and check all toy-like motors. I have one or two that are equipped with 1:300 reductors, but I am a bit concerned about the friction force.

[wvmarle]

That's the trouble of using motors as generators... motors are usually designed to run at high RPM, generators are usually operated at much lower RPM.

[falexandru]

@wvmarle

Indeed. None of my toy motors was able to provide more than 1 V when rotated by hand (which roughly equals the slow wind rpm of a vertical axis turbine).

Although I could not find reliable info about brushless motors used as generators, I am confident the one I brought can work as generator as well (picture in a previous post).

I am going to check:
a) 6 regular diodes 1N4011
b) 6 Schotkys of 1.3 W

to convert AC from the Brushless motor to DC.

The Schotkeys have less voltage drop on forward current than the regular diodes and are therefore apparently better to convert low voltage than regular diodes. But I shall see. The circuits I can find are by regular diodes.

++++

The problem I must solve in the first place is to connect the motor. Its connector is something I cant figure out what it is and therefore I have to mount individual hand-made jumpers to each of the outputs.

[MarkT]

wvmarle:
That's the trouble of using motors as generators... motors are usually designed to run at high RPM, generators are usually operated at much lower RPM.

A standard DC motor (permanent magnet) will run happily as a generator, the voltage depends on the
speed, the torque on the current.

[falexandru]

@MarkT

Yes, from my experiments they do, but supply low voltage at low RPM (ten-to hunderd). I need above 1.4 V to use the current in order to obtain constant voltage downstream.

I found one motor that can supply above 2V, but using a gearbox which is to high torque to spin in wind.

I am currently working on a Brushless motor to convert to supply DC.

+++

If not successful, I still have 2 more plans (C and D :-)):

• plan C: to make my own generator - that is to use electromagnets and not permanent magnets (loose permanent magnets are banned when working with kids) - I am still digging for a tutorial

• plan D: to use an (old-time) bicycle generator

[falexandru]

A BLDC motor is a very tricky device, in my opinion.

I measured the resistance of the one I brought (see picture): there are 4 wires (not 3)- between every pair of them is 1.4 Ohm.

Now, if I supply to the motor even the lowest current from my bench source, chances are to melt it down.

From what I found on the net, the maximum current is only drawn at maximum speed which is fairly high (above 4000 rpm anyway).

The motor I have is in Why-connection (aka star connection). This means that there is a common end of the wires - in this case is the brown wire.

More details here:

https://www.makesea.com/web/claimer/brushless-motor?p_p_id=107&p_p_lifecycle=0&p_p_state=normal&p_p_mode=view&p_p_col_id=column-2&p_p_col_pos=6&p_p_col_count=8&_107_keywords=&_107_advancedSearch=false&_107_andOperator=true&_107_resetCur=false&_107_delta=5

An advantage of the star connection is that it can start at lower torque than the Delta one:

Why is star delta starter preferred with an induction motor?

In consequence, I am not going to supply power to the motor via diode rectifier - as I initially planned.

Instead, I am going to measure the voltage at the output of the rectifier, while spinning the rotor.

The motor is recommended for wind turbines and can supply up to 3 A, maximum voltage being 12 V. I didnt know that by the way :-).

(Scarce) details about the motor here:

https://www.solidrop.net/product/5pcs-4-wire-3-phase-brushless-motor-dc-micro-motor-dia-29mm-for-diy-wind-turbine-generator-water-generator.html

and here (basically the same):

https://savepower.me/product/2pcs-4-wire-3-phase-brushless-motor-dc-micro-motor-dia-29mm-for-diy-wind-turbine-generator-water-generator/

The common end has to be left unconnected, while the other ends have to get into diode rectifier. There are schemes that couple the diodie with Mosfet (parallel) to command the motor. I have no idea whether this can be of any use in the case of wind generator.

More soon.