Show Posts
Pages: 1 [2]
16  Using Arduino / Project Guidance / Re: Wind turbine charge controller prototype/shield on: December 31, 2012, 04:00:48 pm
@rockwallaby ... transistors just eat voltage. only thing they are good for is controlling the current draw rate.  This guy is planning a wind turbine which will obviously put out variable voltage (unless its very smart).  Transistors won't help match the voltage.  There needs to be a buck converter i.e. a coil and a high speed diode to drop the voltage if it gets high. That is if he is using dc.  A wind turbine needs mppt where a pv array doesn't.

If he is using ac, he has the option of doing a hybrid rectifier as I stated earlier.  A 1p/2p hybrid rectifier voltage doubles if the voltage is low. As the current flow rises, the capacitors empty and it converts gradually into a bridge rectifier that doesn't voltage double.  So this will create a voltage range of around 2.5x where the current draw steadily rises as the unit spins faster.  I like it, but the capacitor size has to be tuned to the application, and I'm not sure how to make that kind of hybrid rectifier for 3 phase yet.  Getting a single simple equation written up could solve this.  An mppt buck converter is probably harder to design, but it's closer to the class of one design fits all.

There is one more issue that you didn't notice here, and I forgot to mention.  If we use automobile relays to both charge a battery and drive a load, these two conditions are very different. Relays have max. voltage capability. For a 12 volt system it's never an issue, but for a 24 or 48 volt system it is an issue.  When we are charging the battery, we are really only switching the excess voltage, so a 12 volt relay can charge a 48 volt battery no problem.  If the relay needs to switch off, the battery helps to extinguish the arc.  If we are driving a load however, the relay has to be rated for the full power flow (voltage) because the load does not help to extinguish the arc.  OK, so I discovered that the only good option for better relays is on aliexpress.com  There are 120a spdt relays for around $11 and 80 amp dpdt relays for around $14.   This relay business is tricky. To run a 48 volt system we still have to put one of these both at the bottom and the top of the supply because they are only rated for 28 volts wheras automobile relays crash just after 14 volts.  So with two 28 volt relays in the circuit, we should be able to extinguish the 48 volt arc if they are activated simultaneously.

A flexible system has to be able to account for all of these combinatorics in a fashion that doesn't force people to do a lot of thinking as they build their system.  This begs the concept of modularity principles... I.e. a main controller setup that allows plugin of multiple modules.  Now, to manage all of that stuff in code, we are going to need genuine top quality code.  That means using state machines. 

!!!!!!!!!!!!!!!!!!!!!!
It's stupid for the public to have to redesign an open source charge controller every time their application changes. If this is how we are doing it, then we are just DIYers, not open source.  Yeah, the code is open source, but might as well not be , cause when you get it, it doesn't work because you are set up differently, now you have to redesign.  The way to solve this is to see the system as modular. I.e. you can swap in a relay charging module or a mppt charging module depending on need, and your module declares itself to the controller and the controller knows how to handle it.  You can have 12v, 24v, 48v, and +-24 volt (24 volt 2-pole, by far the best option) This requires more I/O from the controller, which bumps the display off of the controller onto another controller.  Costs a little more, ($17 on ebay wow ... this is essentially nothing)  but isn't it worth it to:
a:not have to rewrite the controller for every purpose
b:have option for a higher quality display and space for a charge controller shield without stacking shields
c:be able to upgrade your system without redoing everything by adding or swapping modules
d:get code done for a generic vt100 terminal emulator which can be used in an automation/scada system as well ( as you are talking about), therefore combining your scada project with the charge controller project and all automation projects that beg a generic control panel (which is pretty much all of them)  Later on, the vt100 panel could be expanded with additional standards to do more cool stuff :-)
!!!!!!!!!!!!!!!!!!!!!


hehe  ebay offers a fabulous set of d class amplifier boards that will run sweetly and very efficiently with no need for a power supply if you can come up with a +-24 volt supply ... There are also lower power 12/24 volt ones, and high power 50 volt ones that would require only a cheap buck converter because 48 volt system tops out at about 56-57 volts. just one example of the advantage of a flexible system.  With a two pole 24 volt supply, you can efficiently move your power 100 feet or more, directly power 24 volt light bulbs as opposed to having to wire them in series, use cheap, simple circuitry to power 3 phase 24/48 volt motors, use the most efficient 48 volt inverters and need them the least.  On ebay you can even find a 48 volt dc refrigerator compressor.  You can run 48 volt computer power supplies that are more efficient. I have about five of them. They don't even need a fan because bucking from 48 down to 12 is a lot more efficient than bucking from 150 or 300 down to 12.  They are more efficient, they are smaller, lighter,  they should be able to cost less, and they are silent.  We are talking about a cheaper, better way of life on the other side of an open systems initiative. 

If we can as a community put together a bunch of kick-ass solutions to dc power distribution, we can cut the cost of solar power systems very legitimately in half by doing nothing but redesigning the system. No need to chop solar panel prices down to nothing.   We are talking about a freaking giant scaled revolution because the system cost will begin to be considerably lower than grid power.  You can shut down half of the f-ing coal power plants, murderous nuclear power scammery where we pretend we are doing it for the power but really doing it for the bombs and actually providing our enemies a fabulous target to blow up in our own back yard so it can shower us with deadly radiation like Japan and Russia. 

The magnitude of potential advantage available by taking on this problem and doing it right is mindboggling, and we don't have to rely on any free energy charlatans, any government regulators, any corporations. This is a golden opportunity for the open source community to literally begin to take control of gargantuan new turfs and show corporation minded nutcases who has the real "power" and that collective intelligence is superior to top down control.  Do it once, do it right, do it modular and flexible and move on to the next project. We can start designing more dc power stuff and put it on the market.  When people realise which plan is best, the old sh** will start fading away.  Solar power prices have dropped radically.  If we get this stuff under control, we can change the world.

let's see ... I am also connected with a 1700 member facebook group and a smaller group of engineers operating at http://egrouphub.com/collab  So our plans and strategies are actually bigger and more complex than this. (and will have a lot more uses for arduino boards) It seems that the biggest job to do here is to help people see how they can accomplish things that are way way bigger by just adjusting their behavior to successfully connect into a larger resonance.

thanks guys for the links,  the timnolan mppt code will be a good starting point.   :-)

17  Community / Gigs and Collaborations / Re: Looking for help to set up an open source charge controller shield + vt100 emula on: December 30, 2012, 08:12:17 pm
interesting ... the vt100 codes were included into later standards .. of course I knew that, but wasn't aware of the related ansi standard.  That's no surprise that this would end up going that direction. It would probably be good to implement the entire ansi standard after getting the vt100 standard in place ... Kinda down the road, but a vt100 solution would be a great start for that loftier goal.   :-)  Unless of course, someone knew of a codebase that basically had that already mostly solved.   I need to look around the open source community for c++ implementations of these standards.
18  Community / Gigs and Collaborations / Re: Looking for help to set up an open source charge controller shield + vt100 emula on: December 30, 2012, 08:03:09 pm
I bet there is some open source code kicking around that just needs some adjustments to port to arduino
19  Community / Gigs and Collaborations / Re: Looking for help to set up an open source charge controller shield + vt100 emula on: December 30, 2012, 06:38:09 pm
vt100 is an excellent means of creating a simple machine control panel that is more reliable than computers. In the automation industry, reliability is super important because when something doesn't work, it can cost thousands, even hundreds of thousands of dollars, so we tend to gravitate towards simpler stuff that doesn't crash on bootup or tangle itself all up into a mess.  B&R automation is doing this (very popular german line of plcs) and I think it is a very good idea.   On the other hand, I think a good open source control panel needs to support some other standards as well.
20  Community / Gigs and Collaborations / Re: Looking for help to set up an open source charge controller shield + vt100 emula on: December 28, 2012, 06:36:03 pm
Oh, as an additional perk, you will have the opportunity to work with me. I am an experienced physicist, EE, automation engineer, systems theorist.  I am professor grade material, and newbs will have the opportunity to pick up a ton of valuable clues as we do this project. In fact, this project could change your life and dramatically elevate your entire world view.  :-)
21  Community / Gigs and Collaborations / Looking for help to set up an open source charge controller shield + vt100 emula on: December 28, 2012, 06:31:44 pm
Hi, I am working with another open source group, and we are looking for people interested in helping out with a high quality project creating a modular, dc power distribution system controlled by arduino. We will open source the code and just sell shields, or whatever level of the product people want to purchase, and distribute explanation/promo documents. Module development for the controller will also be done openly so that we have a complete open source dc power distribution system with the purpose of dropping the cost of solar/wind power created on site. (to promote independence)  I have already thought out the project conceptually (lots of homework on that) and done a dc power distribution writeup that is posted on another site. (still growing).  Cost of solar power is dropping.  We want to cut it in half again because of some very interesting factors. I can't explain all of them in a paragraph, see the writeup.  But in any case, I have figured out how. If we can get the proper system to fly, there will be a global solar power avalanche that is currently just waiting to happen.  So If anyone would be interested in taking part in a historical project that could make money by selling arduino shields and various other innovative and related dc power products please let me know.

oh, that writeup is here:  http://egrouphub.com/wiki/index.php/DC_Power_Distribution 

:-)
22  Using Arduino / Project Guidance / Re: Wind turbine charge controller prototype/shield on: December 28, 2012, 06:18:07 pm
sorry, I am kinda lost, could you post a link to "gigs/collaborations"
23  Using Arduino / Project Guidance / Re: Wind turbine charge controller prototype/shield on: December 28, 2012, 06:13:02 pm
arduino has a gigs/collaboration forum section?  I'll check it out.
24  Using Arduino / Project Guidance / Re: Wind turbine charge controller prototype/shield on: December 28, 2012, 06:11:51 pm
Oh, I forgot ... there is another issue ... lead acid batteries differ in their allowable charge current. Gel cell batteries have current limits. High currents can cause bubbles that damage the gel.  Flooded batteries generally don't have any meaningful charge current limits.  Gel cells have become rather rare, but keep your eyes open.  Sealed, no maintenence batteries are also less tolerant of high currents than flooded ones.  I have an extensive dc power distribution writeup (that is growing) here: 
http://egrouphub.com/wiki/index.php/DC_Power_Distribution  we will be doing a lot of open source projects, including arduino projects on that site. 
25  Using Arduino / Project Guidance / Re: Wind turbine charge controller prototype/shield on: December 28, 2012, 06:04:41 pm
so anyone who wants to get involved in this project and help me to do it right, please feel free to let me know  :-)
26  Using Arduino / Project Guidance / Re: Wind turbine charge controller prototype/shield on: December 28, 2012, 06:03:12 pm
BTW, there is some stuff bouncing around here that is not making proper sense.

First of all.... a 24 volt motor rating?   Don't we understand what all of the related dynamics are here guys?  Motors that act as generators output voltage according to their speed.  The 24 volts will pop up if you run the unit at its rated speed, and you have some control over its speed by means of gear ratios and wind mill protection schemes. Next thing is that whatever the battery is at, it will draw the motors output down to the battery voltage. So if it is trying to output 20 volts, but your battery is at 15, the thing will pump a lot of current into the battery at 14-15 and risk burning out.  It will not pop your inverter because the battery will hold the voltage down.  The real problem here is that your battery will draw nothing from the unit until it goes over 12.5 volts, then at 14-15 volts output from the motor, it will be drawing like mad, so the power flow curve will not respect the power available curve, and you could call this an impedence mismatch.  The impedence mismatch is handled by doing something like the above mentioned hybrid 2-pole/bridge rectifier for an ac output generator or an mppt controller which carefully watches the current flow and draws current optimally while it converts extra voltage down.  So, there is another solution to this and its called using a rotor current scheme for voltage regulation like is done with automobile alternators, but having to supply rotor current for a wind generator could really suck. Automobile alternators do not make good wind generators excepting in very carefully designed circumstances. Basically the only circumstance that works well for this I have ever heard of is if you use a large wind mill that uses a serpentine belt that spins several automobile alternators.  Then you have a controller that turns on the rotor coils one at a time with the proper voltage to make it output the voltage you are after.  So the alternators turn on something like an audio level meter, and as they spin faster, the controller diminishes the rotor coil voltages... or, as another option to diminishing the rotor coil voltage, the hybrid rectifier strategy could be used.   
27  Using Arduino / Project Guidance / Re: Wind turbine charge controller prototype/shield on: December 28, 2012, 05:44:17 pm
Hi, hope there is someone still watching this thread.  I have been cracking on a more complex version of this issue for some time.  I am an  EE as well as a software engineer, and I would really be pleased if I could get some help from anyone building a charge controller like this. I have done most of the homework conceptually and just need to work out details.  First of all, looking at this schematic, I see some weaknesses.  For example, a blocking diode is only necessary if the panel remains connected to the batteries continually rather than just when the voltage is high enough to charge. This may or may not be necessary for a wind generator. That blocking diode is going to eat around 4% of your power on a 12 volt system. If you measure current in stead and switch off as soon as it isn't positive, then the problem is solved.   Second of all, we need a current flow measurement on the battery to really see what is going on.  You are basically only using a single analog in to measure the voltage on the battery.  This won't work well to tell you the charge state because the voltage on the battery will change depending on how much current you are driving into it. (a major issue for a wind power system) You don't really want to overcharge the battery as this can be destructive.  Breakers are of course preferred over fuses.

The battery moves between charged and discharged with as little as half a volt of change, so you need to have equations that govern this to get good control of it.  There is a pdf out there that says a ton about the behavior of 12 volt batteries charge state.  I have a copy of it.

Now, here is the big reason why you don't want to turn this into a circuit design for a shield.  This system is not flexible/modular.  To make a charge controller shield that has value, it needs to be prepared to deal with a lot of variations, including adding in stuff like mppt charge control.  PWM charge control .. next thing it needs is an exhaustive document explaining  how to design and plug in modules into the system, configure it, etc. By the time we get that done, you will discover that the board may not have enough I/O.

Also, these cheesy little lcd screens are really weak. Charge controllers can be pretty complex, lots of settings, etc.  The thing really needs a tft and a keypad.  That sucks up more I/O. There is a very interesting way to solve this problem so that you can have/use a tft or not.  It's called use the vt100 terminal emulation standard.  If you set up an additional arduino that acts as a vt100 terminal emulator, several things will happen.  First of all, you will have the option to run a cool display with an additional arduino in communication with the original one.  Second of all, you won't have to use that if you don't want because there are terminal emulator programs that run on computers, so if you want to save the $50, you can run a terminal emulator on a computer to configure the unit in stead.  Next, If we design such a terminal emulator arduino, we will have created a marketable product in and of itself.  All it takes is an arduino mega/tft combination (cheap in some places) that uses multiple communication options to allow multiple types of communications links. 


  When we have wrapped this stuff up, then we have a charge controller shield that you/we can sell on the market. We also have an arduino vt100 terminal emulator system that we can sell on the market as well as use for other automation efforts.  And the system kicks ass because it was done aggressively and properly.  I already have a vt100 client library that I wrote for another controller/project that could be ported to Arduino. That library can be used to set up a settings management system lickety split. 


So, the next issue is .... 12 volts sucks. I'm shooting for a dual-pole 24/48 volt system that balances the batteries, has options for mppt, and options for multiple pv arrays. There is a ton of value available by doing dual pole 24/48 volts. 

1: you can power stuff that wants 2 poles directly like class d audio amplifier boards
2: you can power 24 or 48 volt items
3: the system is more efficient because you will run at higher voltages.

costs: requires more breakers and relays to protect and secure the system.



Lastly, if you are planning on using a wind generator, mppt control will be critical.  because wind generators have a very ficle and complex behavior.

Next issue for a wind generator is that if you have a generator that produces ac, then a hybrid two-pole/bridge rectifier will seriously help manage your charge current without needing an mppt controller.

:-)  anyone who wants to help me build this cool item can be in on it.   Then we can just sell the kick-ass shields, and people can have kick-ass dc power distribution at open systems prices.  This could go a long way towards making home power and independence cheap.

BTW I already have partial parts lists, some code, some arduino megas, tfts etc.
Pages: 1 [2]