Hello Mike,
That sounds like the simplest solution yet! Thank you very much for your help. I'll let you know how I get on.
Meanwhile, just to give you a glimpse of the project so far......

Attached Thumbnails

         

   

 

How tall is your tower? At least 15 meters my guess.
Also, did you make the turbine yourself or did you purchase it?

 

Hi Fernando,

The mast is just under 12 metres tall. Here in Europe there's a limit of 12 metres on the height of the mast. If it's over 12m you need to get planning permission from the local authorities and also permission from the regional authorities because the generator will create "an impact on the environment". If the thing is under 12m you can avoid all this administration.

The generator is 3KW and drives a series of 20 batteries of 200AH. There's a controller that monitors the charge, the wind speed and direction. The wind speed and direction signals are used to control a servomotor in the generator and keep it pointing towards the wind. The circuit will also sense when the windspeed is too high and turn the generator through 90° thus acting as a kind of brake.

There are solar panels to augment the wind turbine output but as yet these haven't been installed (I need to repair the roof on the barn!) The system was purchased from a company called Kingship in China: turbine, mast, blades, batteries, controller, inverter, solar panels, and delivery $10,500

I live in France and I had to pay customs on the import. I have since put the system together myself. A large foundation was dug out for the mast (18 metres cube), a battery box, underground cables, a crane to hoist up the mast, etc. There were several initial problems getting the thing running because the blades weren't properly balanced, and then one had a leak and picked up water causing a further imbalance. Now it seems to work OK and the batteries supply enough power to run the house for four or five days from full charge when there's no wind. The noise is negligible, around 50-55dB.

 

This is a circuit I came up with some time ago. Its great for doing just what you are trying to do. It gives you a high set/low reset window.
I use it on my grid tie inverters to tell them when to connect and disconect based on input voltages.
The opto isolator can be used to trigger stuff or just be ommited if you just want to run the relay in its place.
its not so fussy about its power source voltage either. All you would need to do is put the right resistor voltage divider together for the voltage referance input.

Attached Thumbnails

 

__________________
"Issue a general safety warning. Then look the other way and allow stupidity the chance to eliminate itself." -- tcmtech
"Those who can, Will. Those who can't, will achieve positions of power over those who can and then promptly stop them." -- tcmtech
"Your impossibility may just be my day to day routine." -- tcmtech

 

Hello Again,

Still not much luck I'm afraid. I followed your suggestion Mike and I connected the circuit to a 12v transformer instead of using the wind turbine controller power supply. I then connected the negative side of the battery bank to the ground of the 555 circuit. However, when I connected the positive side of the battery bank to the circuit there was a large spark and quite a bit of smoke. It looks like the voltage division resistors have got cooked, probably not surprising since they were connected directly between the two battery bank terminals.

Thanks too for your circuit tcmtech, it was something like my original idea of a window comparitor driving a SCR, but after what's just happened this afternoon I have a feeling the same problem will occur again in your design. The battery input consists of 20 batteries (200Ah) connected in series that will be shorted across the two trim pots in your schematic. Yes, the voltage will be there alright, but there are far too many amps for my liking and I see the thing exploding again. I'm certain both of these circuits would work fine, but how do I tap off a voltage reading to regulate the chip without all that power getting in the way?

As I mentioned in an earlier reply, neither of the battery bank cables are tied to the wind turbine controller ground; there's about 100 volts difference. However, I have tried putting a pair of voltage dividing resistors between ground and the terminal marked 'L'. When I first tried this the 100v dropped to almost zero; I don't know why, but the system continued to function, nothing got hot or started to smoke. The right resistors will have to be selected to obtain a 4v or 8v output but it might be possible to get the circuit working like this. I'm just guessing of course, I'm really out of my depth so if someone's got some sound advice I'd be glad to hear it!

Thanks everyone for your consideration and helpful ideas.

 

I reran the simulation and checked the power dissipation in R1 and R3. It is only 120mW worst case, so even a 1/4w resistor should have been ok. I suspect that the resistors you used for R1 and R3 could not stand ~240V, so here is my suggestion: Put three or four 1/4W metal film resistors in series, such that R1 totals 630K and R3 totals 590K. This will keep the voltage across each resistor to less than a hundred volts.Here is a spec sheet that shows the max working voltage for a typical Metal Film resistor. What style of resistors did you use for R1 and R3?

The "transformer" used to power this circuit should be a mains-powered plug-in charger or power supply that puts out 15V to 20V DC. That goes into the input of the LM7812 voltage regulator to make the regulated 12V to power the 555. The accuracy of the trip points is directly a function of the voltage at the 555 pin 8. If this voltage is 3% low, then the trip points will also be 3% low, etc.

 

OK Mike, I follow your what you're saying. The resistors I used were all 1/8 watt 5% and must have been far too weak for the circuit. I'll have to look around to see what else I've got, otherwise I'll have to order some sturdier parts.
The transformer was a mains powered device giving out just over 13 vdc; it was connected to the regulator as in your design.
I'll continue in the morning and keep you up to date as I go along.

 

Ok, I think I understand what is going on. Neither the battery bank, nor the generator, nor the voltage sensing circuit inside the windmill controller is tied to earth ground. The entire system is "floating". The 100V difference between one end of the battery bank and the "ground" on the controller that mysteriously went away when you were probing around is probably due to some very high-impedance leakage path somewhere inside the battery bank.

I'll bet that if you measured the voltage from the controller ground to the negative side of the battery bank, you might see -100V, while from ground to the positive side of the battery you might see +140V, such that the sum of these two readings would come close to the measurement from Bank- to Bank+ of 240V.

This is somewhat of a surprise to me. I thought that either Bank+ or Bank- would be connected to earth ground. However, this does not matter if the 555 voltage sensing circuit is totally isolated in its own right.

A caution: The secondary winding in the small transformer that you are using to power the 555 circuit will be tied to Bank-. If Bank- is not at earth potential, then whatever voltage Bank- sits at will appear between the primary and secondary windings on your transformer. This should be ok, because most of those consumer grade power supplies (wall-warts) are tested to something like 1000V isolation between the windings, but I just thought I would mention this to you.

 

If I had to buy them, I would order some physically larger resistors, like 1W size. They seem to stand off higher voltages than 1/8 or 1/4W ones do. I would also break R1 (630K) into two separate resistors, say 130K and 499K. Mount the 130K on your circuit board, but mount the 499K off the board right on the Bank+ terminal (use heatshrink tubing on the floating connection off that resistor) That way, you do not have the 240V on your handwired board (only 55V). Even if you slipped with a meter probe, the 499K resistor between Bus+ and your board would limit the current to a safe value. Do the same for the R3 line.

I would put the trim pots at the 555 end of the R1/R3 chains. That way you are not apt to hit the 240V while you are tweaking them.

This raises another issue. The LM7812 regulator has a dropout voltage of 2V, so to get the rated 12V out, the input voltage must be 14V or higher. If the voltage sags below 14V, the 7812 might be dropping out of regulation.
If it is close, try putting a larger electrolytic filter capacitor (say 1000uF at 20V) across the input to the LM7812. There might already be a capacitor inside your Wall-Wart, but if there isn't, adding one might help get the voltage a volt or so higher.

 

Hi Mike,

You're right about the battery voltage measurements. Accross the bank terminals you'll usually get a reading of around 240V, but measuring from the positive terminal to the controller ground the figure is around 100v and from negative to ground its closer to 120v. Why the two figures don't even add to 240v is beyond me but that's what I keep getting from the voltmeter.

I went through the 555 circuit today and rebuilt it. I dug out my most powerful resistors and chained them all in series as you suggested; I was able to obtain exactly 590k and 630k to divide the incoming signal. I didn't add any trim pots at this time because I don't know their power ratings. I don't have a 7812 but I am using a transformer running through a LM317 circuit that puts out exactly 12v.

I took the revised hardware out to the wind turbine controller a few minutes ago and tried wiring it up. First attempt ended in another big spark and a loud bang! However, after the smoke had cleared I was able to see what had happened; it was my own fault. The battery bank is driven directly from the turbine generator and as a result the voltage has an ac ripple. I had included a large electrolytic capacitor across the input to the PCB to try to eliminate this fluctuation; the bang and the smoke were due to this cap exploding. As nothing else seemed to be damaged (except one terminal that had been completely melted off the end of input lead!) I removed the remains of the cap and tried reconnecting for a second time..... ....all seemed to work OK! No sparks or smoke. The LED glows from the pin 3 output, and yes, it looks like it's working! Now I'll need to add those trim pots to try and check the trip levels.

I read your last message and by chance the resistors I added are all mounted far away from the 555 at the opposite end of the board. They're large components and there was no room to place them where the original resistors had been situated. I think it'll be fairly safe in this configuration.

I'll now try adding the trim pots, put the relay into the circuit, connect it all up and keep you informed. The project is making progress and I think I can now stop worrying about the sparks and the explosions! Thanks again for all your help.

 

What is the input impedance of the voltmeter you are using to measure the battery terminals? I believe that the reason the two readings do not add to 240V is because the battery bank "ground" is being defined by a very high impedance leakage path to earth ground.

Just the act of connecting your voltmeter between earth ground and Bank- or Bank+ draws enough current(through the meter) to effect each reading. If you owned two identical meters and made the two measurements simultaneously, the two readings would add exactly to the bank voltage. Making the two reading serially with one meter shifts the first reading one way, and the second reading shifts it the other way

If you want to "filter" the voltages presented to the comparators inside the 555, then increase the size of C4 and C5. With the present values of 0.1uF, the -3 db corner frequency above which any ripple on the battery bus is attenuated is about 70Hz. If you increased C4 & C5 to 10uF, for example, the corner frequencies shift down to about 1Hz, providing about about 30db of additional attenuation at a ripple frequency of 100Hz. Look at this:

Attached Thumbnails

 

 

Sorry Mike, but I can't answer your question about the impedance of the meter. The one I've been using is a Fluke digital multimeter 8022A that I've had for many years. I am beginning to suspect that there is some problem coming to the surface in the battery bank because these days it won't hold a charge as long as it did a few months ago; maybe the connections need to be checked out; maybe a battery has gone "bad". This though is another subject and not the 555 controller circuit we've been building.

I'll take your advice and change the capacitors to eliminate the ripple. After that all should be set. I've adjusted the resistor values and installed a couple of trim pots; all are fairly beefy components to handle the power. The indicator LED has been removed and the output from pin 3 now goes to drive the low voltage relay. When the circuit is connected the output is high and so the relay will make contact, this then is connected to a second more powerful relay that switches the house supply between inverter output and the national grid.

I assume that when the pin 3 output is high the batteries are charged up and that the system can be left connected to the inverter/battery bank? When the battery output level falls low (below 240v) the 555 will trigger the two relays to switch the house across to the national grid and allow the batteries to charge up again without any load. When the batteries reach 260v the system will trip again, pin 3 will go high, the relays will change state and the batteries will begin to discharge via the inverter as the house uses up energy.

If this isn't the way things work please point me in the right direction! I'll try wiring it up today and the only way I can really try it out is to wait for low voltages or high wind.
I'll keep you posted.

 

doesn't the 555 have a "push pull" output meaning that the diode would short the out put when it is "off" i may of course be talking out of my hat

__________________
I AM the exeption that disproves the rule in many ways but the rules still apply (unfortunately)

www.simonsphotography.org.uk/ - My other hobby
www.rushdenrotaract.org.uk/ - make a difference and have fun !


Never buy "Trust" products, all mine broke !!!

 

why not use a computer power supply from the 240 V dc of the batteries to output the 12 volt supply ?

__________________
I AM the exeption that disproves the rule in many ways but the rules still apply (unfortunately)

www.simonsphotography.org.uk/ - My other hobby
www.rushdenrotaract.org.uk/ - make a difference and have fun !


Never buy "Trust" products, all mine broke !!!

 

Hello Thunderchild,

If a 120Ω relay coil is connected between pin 3 of the 555 and ground, and pin 3 is HIGH, the 555 sources current from pin 3 through the relay coil to ground, actually about E/R=12/120=0.1A=100mA. If a diode parallels the relay coil such that the cathode end is at pin 3, then the diode is reversed biased by ≈12V, and only reverse leakage current flows though the diode. If pin 3 of the 555 is LOW, there is 0V across the coil and the diode, and no current flows through the diode. Look at the revised schematic I'm posting in my next response to vielle568.

The only time any significant current flows through the diode occurs just as the 555 switches from HIGH to LOW. Without the diode, the inductive spike from the relay coil would try to drive 555 pin 3 to a negative voltage. The diode clamps that voltage to no more than -0.65V (forward conduction voltage of a Si Diode).