48 volt version of "Michael Davis 555 controller"

[pguillem]

Hi All.

I'm new at the forum and taught it was a nice place to discuss this topic.

Many of you probably know the award winnng 555 design of Michael Davis at:
http://www.mdpub.com/555Controller/

Honestly i had never seen a 555 being used in such a clever way. Which is why i choose to try his model before getting into the RossW approach, which i also find very interesting.

This is the Schematic:
http://www.mdpub.com/555Controller/images/555ChargeController.jpg

Since i will be building a 48 volt version of the 555 controller circuit, it would be great to hear some expert opinions on the additional calculations i did in order to make it work:

1) The voltage regulator i will be using will be a LM2575, which can take arbitrary DC inputs from 7V to 60V. This solves the problem with powering the circuit and works within the ranges handled by both the turbine and my battery array.

2) In order to set the charge/dump thresholds i will set R1 to 34.3K and R2 to 18K. Meaning start charge at 54 V and Dump at 58 V.

3) Obviously, the relay will not be a standard car (12V), but a 48V version.

4) I think the rectifier bridge at the 3-phase input will be enought protection against "reverse voltage", but somehow i also have my doubts.

5) The system´s GROUND reference will not be -12VDC, but -48 VDC from the negative poles of the batteries and rectifiers. I have no idea if i should add some more ohms before hitting ground at all appropiate places. Any thoughts?.

I´m calculating this with the stuff i remember from university... and i´m a software architect, not an electronics engineer hahaha... hence the warning and posibly silly questions.

Aside from my calculations and perspective, would any of you have any important remarks or ideas?

I have also written to Mike asking for his toughts, i will share them here if he answers (let´s hope he does).

Thanks!
Pedro

PS: I´m a 30 year old kid building a turbine with another 60 year old kid (my dad) LOL.. and we are really enjoying this. So far our 2 kilowatt LIFT VAWT reaches 60 Volts AC @ 150 RPM without an issue ... all we need now is a working charge controller... we tried a HAFEI PWM unit but sadly it is designed to ask for a $#|+load of torque from the turbine, rendering it useless. So we taught of a "less sofisticated" approach to just charge what´s really available.

[rossw]

Which is why i choose to try his model before getting into the RossW approach, which i also find very interesting.

Not sure which "RossW approach" you're referring to, but if it's the circuit that's been bashed about by a number of people that I scrawled out 5+ years ago - that was never INTENDED to be a working controller!

It was a discussion point of how it COULD be done. It has so many shortcomings and limitations, so many poor design failings that I would never PROMOTE that circuit as a solution.

OK, a number of people built it. They're happy with it. It works as advertised. Fine. But it's NOT something I would ever make or use myself.

[David HK]

I wonder if Ghurd will comment on this thread?

Ghurd's controller can work at 24 volts DC with suitable modification but when you get up into the 48 volt DC range you're dealing with serious power.

Reliance on a 12 volt DC DIL chip for dump load control on a 48 volt DC machine is something I would not embark on.

New comments will be read with interest.

David in Hong Kong

[pguillem]

Hi Ross!

LOL, yup, that´s it.

Thanks for taking the time to read my post. I fully understand your remarks on your circuit and also saw all your comments on it from 2007. Don´t worry, we are just experimenting here and mostly trying to have fun ... 

I taught your diagram was very nice proposal worth evaluating and analyzing.  You were even kind enought to share it with the community, which i consider very valuable. I´m sorry to hear some people reacted in an overly-critic manner...

To tell you the truth i started researching all this controller part just today... as we have been setting up a turbine since august last year and got preety much RIPPED OFF by a Chinese company called QINGDAO RICHUAN MACHINERY.

The 2 KW lift VAWT turbine they sold us spins and produces voltage, but only  some 150 watts (3 amps) with random 700 watt spikes that last under a second...

they sent us a PWM driven controller which we think was not designed for that turbine... the spikes (we think) are peak ampere bursts produced by the inertia of the spinning generator itself, not being able to sustain the required torque when the controller sets itself to allow more current to flow.

So our logical solution is to avoid using that junk and instead charging the batteries with "what the generator can give"... without complicated poorly designed PWM low charging functions.

Would you consider this a viable solution?

Cheers!
Pete

[David HK]

Pete,

Can you tell us which country you are in, and the nearest town.

Before getting into this area of interest did you ever study your local wind characteristics, to work out if you are in a good or bad wind area?

Dave in HK

[pguillem]

Reliance on a 12 volt DC DIL chip for dump load control on a 48 volt DC machine is something I would not embark on.

Well it is indeed serious power. 40 amps at 54 volts DC is worth being delicate with. Furthermore, we have a 10KW inverter and 12 batteries (3 series of 4 batteries each) to power the whole house.. with air conditioning included ...

The 12 volt regulator on the design (LM7805) would be replaced by a suitable LM2575, which will output the same +5 volts to power the small "logic" circuit.

Given that the serious power will not be touching anything different to the RELAY (RLY1), i don´t see why it would be an issue.

I´m also all ears, for i´m almost sure i´m missing something here!

Cheers!
Pete

[pguillem]

Hi Dave.

Yes. We are in "Santa Veronica, Colombia". A little town close to the city of Barranquilla. We have a beach house here since the year 2000 and have been studying the wind behavior for 10 years.

Average wind speed (all day and night): 10 m/s to 13 m/s.
The house is on the edge of a cliff at 30 meters above sea level.

- Turbine is a 5 blade, lift VAWT.
- Each blade is 2.60 meters high and 35 cms long. (NACA airfoil unknown)
- The combined diameter is 1.70 meters.

- Generator outs 3-phase AC.
- System spins steady at 102 - 150 RPM.

I can provide more data if required. I even have some cool videos.

Cheers
Pete

[oztules]

Not sure if I would go this way.

I would prefer to drive the batteries directly from the mill and bleed the excess off with a Ghurd controller (or similar) from the batteries into a traditional dump load.
Your power in is fairly trivial at 150 - 800 watts, and I would find it unlikely you would have to drain off much excess if your using it to power a big inverter.

Even a 2807  fet ( dime a dozen now) is good for 75v 80 odd amps. A few in parallel would give you room to spare (13milli ohms rds on).
I'm not a fan of switching higher voltage higher current (55-57v@ <20A) with relays if I can avoid it.

Simple op amps or a 48v Ghurd to suit the voltage would do the job I think very well.

This way, any power from the mill automatically goes to the batteries, and in the event you get to float voltage, you drain away the excess.

Just my take on it. ( I've never been a fan of 555 timers for reasons I dont understand in this duty... just never seen the need for them in this application I guess.)

I suspect you will get a better result from a Ghurd type controller, than the circuit envisage at this time.... I see relay chatter in the mix with the proposed circuit, as I don't see the hysteresis in the circuit (wasn't looking too hard either I admit )

With a Ghurd style dump, the hysteresis is unimportant, as the pulsing is without physical interruption of the current  via a relay, but a quasi pwm .



.............oztules

[OperaHouse]

NOTHING with a 555 is award winning. I refuse to use ome for anything.

OK, you will use it anyway.  This is known as the IKEA syndrome.  At least consider the following.

Your high voltage regulator is destined to fail.  Tap off the battery at 12V.  It wont draw much.  If drawn from 48V, isolate with a resistor into the regulator. protect max voltage into reg to 18V with zener and large cap.

Relays have a lower current spec for NC contacts than NO contacts.  Even if rated the same the NO contacts will last a lot longer.  Forget the NC contacts and just use a diode.  I repair a friends construction lifts.  They have a 24V 75A relay.  He says he only gets a couple months use before they fail.  I drill them open and reburnish the contacts.  Just a bad idea.

A quad op amp could control four FET for a progressive turn on.  This would limit massive turn on spikes.

If you are a computer guy you could do this with a small micro and have complete control of turj on voltages.

[ghurd]

I have working modifications to the Ghurd controller for 48V.

That 555 controller is about as bad as it gets.
There is no way to calculate a suitable dump load.  Either it will be effectively shorted, or running away.  Neither is a decent option.
Plus everything they said.
G-

[hydrosun]

I used a Chris Geason controller from Homepower that used a 723  and 555  IC for morw than 20 years. They work great to control dump loads. If you do a 48 volt system the voltages will get over 60 volts so you need components at least 75 volts or higher, or protected from higher voltages. Ghurds controllers take all of that into consideration.
Chris

[pguillem]

It seems clear to me that the 555 won´t be helping much and could become a potential headache.

After carefully reading all your posts i´m sure i´ll go with the Guhrd controller.

Where can i find the suitable design for 48v?.

Best!

Pete

[zvizdic]

I use Ghurd  48V controller on 10 ' for years now .
4 paralleld IRFP150N  driving 2.5 ohm  heating  element.
I like so much Ghurd controller if you ofer me C60 for it I refuse .

[oztules]

"After carefully reading all your posts i´m sure i´ll go with the Guhrd controller.

Where can i find the suitable design for 48v?."

Then I would PM Ghurd, and buy it direct, not worth finding the bits and making a board...  you also get sane support.
If you really want to do it the hard way, his site has the circuit and explanations etc...but for the price, I'd just buy it direct and save the headaches.

Try this http://www.ghurd.info/


........................oztules

[zvizdic]

It is very easy  to assemble if you buy a kit from Ghurd.
I absolutely trust it, it is  simple ,reliable and cheap !
 Buy single high current mosfet 100V or up with big heat sink in place of BUK455.


In position ZD1 = 8.2V
                  RY   = 2.2K
                  RX   = 24K
C1 parallel with 5.1 ZD

[ghurd]

Thanks guys.

quasi pwm

I never throughly understood the obsession with "true" PWM at a fixed frequency and time related dump load.
I do NOT care what time it is.
Or how long the pulse is active, within reason.
Or what the frequency is, within reason.

I only care if the battery or PMA is being ruined from using bad controllers.
G-

[pguillem]

Excellent.
 
I really thank every one of you for contributing and expressing your opinions. I hope i can also stay contributing on the forum.

I did not know the parts were for sale on the site. I will gladly buy one of the kits to try it out (i hope you can ship international orders).

If you want to see the system running as it is right now, you can try this links:

Turbine starting up (13 m/s --- 150 RPM):
http://youtu.be/Hb6cRzzO-r4

Controller output:
http://youtu.be/kBNKMH3nD4o


This last data is what is pissing me off. Take note on how the watts and amps spike to 500 for a brief moment every 30 seconds or so... as if the controller  demanded more torque than the turbine can give in order to mantain 500 watts. This happens with both the batteries depleted or almost full, no difference. The outcome is similar with slower winds.


Any toughts?


Remarks:
NEVER... EVER... buy from this chinese suppliers:

1) QUINGDAO RICHUAN MACHINERY CO.
2) HEFEI WIND POWER

Best
Pete

[zvizdic]

If I got this right battery at 55V wind at 52.3 .
That is not right,wind side voltage should be higher then battery side to charge .
How big is your battery bank?

[pguillem]

Dear Zvizdic.

Yes, you are right. In normal conditions it should be above battery voltage, BUT:

The charge controller has a "Low Voltage charging function" built in. It uses PWM to take whatever input voltage is available from the turbine to create a constant 58 volt output to the batteries.

The Volt/Amps ratio however is proportionally affected... if the turbine is producing 8 Volts and 5 amps, the PWM mechanism will use idle amps to compensate and raise the output voltage, and so on.

As the input voltage increases and almost "match" the battery voltage, the real idle amps should be almost a 1:1 relation from what is being produced and what is being fed into the batteries.

And this is the part where is not working... you saw the turbine spinning at 130 RPM.. according to its table, it should be generating at least 1500 watts but the damn thing only hits those little spikes few times per minute.

I believe what happens is the controller handles the power by switching over different circuits depending on the idle power, and the torque that would be required to sustain generation on final circuit is simply too much for the turbine... falling back to the previous circuit where it can only drain 20 watts.

The controller has 4 variable parameters:

- Floating voltage: 58,8 volts
- Discharge Current: 40 Amps
- Start charging voltage: 8V to 60V
- Admitance (in a factor of Siemens): 1 to 15

By the way, the lower the admitance value we set the less load will be demanded from the turbine... but less amps will pass.. so far we have tried every possible combination of admitance vs. start charge voltage and the outcome is the same..it either spins faster and produces less amps, or spins slower and produces more amps but less voltage.

Am i making sense? have you guys seen anything like this before??

Here is the link to the controller:
http://www.win-power.cn/en/product_display.asp?id=1004


Pete

[pguillem]

Continuing the previous post....By the way...

We have a 10KW inverter and 12 gel deep cycle batteries (150 Amp/h).
The 12 batteries builds an array of 48 volt in 3 series of 4 batteries each.

A 2KW turbine and a 13m/s average wind should charge the batteries in 30 hours.

All we built so far is a HUGE UPS... the inverter has a charging function as well and we have been using it while we figure this out, but the turbine has not been able to charge more than 130 constant watts.

[ghurd]

Hi Pete,

1st-  Did you get my reply to your email?  And did you reply to it?
(I may be having some email issues again, which happens every so often with yahoo)

"Am i making sense? have you guys seen anything like this before??"
Yes, but in a different context.
I have seen it in DIY controllers that switch the turbine output to a load (like the 555 controller above, and why in reply #9 I said 'effectively shorted'),
or in DIY or Oatly controllers that short the turbine as a form of 'charge control'.
During regulation, both of those make power into the battery go to 0W.


"So our logical solution is to avoid using that junk and instead charging the batteries with "what the generator can give"... without complicated poorly designed PWM low charging functions.

Would you consider this a viable solution?"

Yes.  IF the turbine's power output and Volt per RPM is suited to a 48V battery bank.

You will have to determine if the turbine is suited to direct 48V battery charging.

If the 3-phase is rectified before the controller, maybe try connecting the rectified DC to the battery bank directly (without the Chinese thing).
I would suggest doing that in lower wind speeds, and being very careful about having the turbine unloaded.  Unloaded makes high RPM.  High RPM and unloaded makes VERY high voltages.

I hope I worded that well enough to make sense.
G-