Simple MPPT solar panel charge controllers

Hi everyone,

I've been looking at the "proper" voltages for charging, and especially at this web site: http://www.solarnavigator.net/battery_charging.htm. If I'm understanding this info correctly, assuming my solar panel is incapable of exceeding the charge rate limitations of my batteries (wouldn't that be a nice problem to have) then I don't need to worry about current, only voltage to the battery. My "bulK" and "absorption" phases look to be the same, and (in my case) should be at a 15.5 volt level, followed by a "float" phase with the voltage at about 13.11 volts.

Two steps, in other words. Kick the battery as hard as I can at 15.5 volts, then when it reaches that voltage drop the input down to 13.11 volts and maintain it there.

Does that sound right to everyone?

BillyDoc

 

Outback say the Power mx 80 is even better at it. But I really can afford this much efficiency and I hat to buy a 60 when I want to buy an 80... Oh, so I wish I could get a even cheaper one for like $99 that would work with higher voltage system even though I would be buying 80 in future. Any solar controller consolers out there.

 

Are you saying that the controlles somehow increases power? That makes no sense. All circuits have power losses. Study some basic high school physics.

 

Gonzo, I'm saying that an MPPT controller "can increase your renewable energy yield by up to 30%," which does indeed have the effect of increasing the available power that actually goes into the battery. In other words, your approach of just hooking the solar panel directly up to the battery will work, but you will be throwing away roughly 30% of the power that solar panel develops, not putting it into the battery.

This is not creating power out of thin air, as you imply, it's a matter of intelligently dealing with the natural characteristics of solar panels. As I pointed out above, they are NOT linear devices. Please take a long hard look at the graph below, which shows typical solar cell current/voltage behavior, which I extracted from the Maxim app note attached above. It might also be useful to go back to the app note itself, and read it. The same goes for the LM5118 controller, which can indeed charge a battery at an output voltage higher than the input voltage. Since you have had that high school physics course it should be easily understandable for you.

As for the high school physics, you have the advantage of me there, I missed it. I was too busy with other things at the time. Sounds like a good idea though!

BillyDoc

 

Can you explain what an increase in energy yield is? If it is not an increase in power then how can it charge the battery more? If you start with a solar panel that can only trickle charge the battery, there isn't any power source to charge more than that. I think you are very emotionally attached to the snake oil they sell. I read their website and it is typical of their kind. There are many claims and no real data.
Also, if you charge the battery to 15.5V and then drop the input ot 13.3, whatever the source is will discharge the battery until it gets to 13.3V. You really need to study basic physics and electricity

 

Gonzo, I started to try and explain . . . but after thinking about it a bit and looking over your previous posts here I'm afraid I have to conclude that there is nothing I could possibly say that you would understand. You just don't have the educational background for this sort of thing. So, I'm afraid the answer is, no I can't explain it for you. You might try the web or reading some of the stuff I've already posted, if you are really interested, but my reason for posting here is not to be teaching anyone basic electricity. My reason for posting is simply to provide a handy circuit for those who can understand it, and use it. It's a take it or leave it kind of thing, and you should clearly leave it.

BillyDoc

 

In other words your ignorance is my fault. Sorry to do that to you.

 

Exactly! On at least four occasions above I tried to point you in a direction that would answer your questions, but in each case you couldn't be bothered to do any reading or try to sort it out yourself even though I provided the necessary links. And in most of your replies you also included snide and personal comments. So, yes, my ignorance of how to help you is exactly your fault. And apology accepted. I'll be moving on now, I've a design to work on.

BillyDoc

 

Gonzo,

Don't be angry over this; lots of folks don't understand basic electronics, and this subject is just a touch above that. If you like to read books, the Radio Shack books on basic electronics and the one on power supplies will bring you up to speed on all this in as little as one week, maybe less if you grasp all the concepts quickly.

The book "Electronics, 2nd edition, a self-teaching guide" by professor Harry Kybett, ISBN 0-471-00916-4, also available at Radio Shack, is about as good a self-teach basic electronics course as exists. Though less than 300 pages, if you learn all contained within, you'll know more about electronics than 99% of the population, and be able to design simple and even moderately complex circuits, select component values, understand semiconductor amplification and regulation and power supplies of all types.


I've been following this thread from day one, I just have not had anything of value to contribute thus far. I'll go back into 'lurk' mode now

Jimbo

P.S.

Billy has been very tactful and polite about your lack of knowledge of the subject matter. You probably know 10X as much as he does about some subjects, just not this one. Read up!

 

Billy,

I haven't had a chance to go over the last schematic you posted in detail, but the general concept of the circuit seems about right. A complete detailed analysis is, admittedly, a bit beyond my field of expertise.

Gonzo,

An MPPT is not the same as a regulator or charge controller. And it's not a "free energy" device. Rather, you can think of it as a sort of variable compensator- much in the same manner as a multi-speed transmission or a controllable pitch prop is a variable compensator of sorts.

In an electrical system, power = voltage * current. Depending on the lighting conditions and the electrical load, a solar panel might operate at higher voltage / lower current, or lower voltage / higher current. The graph in post #34 shows the operating curves for some particular solar panel under different illumination levels.

Now, for any given illumination level, there will be some combination of voltage and current that is optimal- ie, maximizes power output. But it's very rare for the load- the battery, motor, whatever- to demand exactly this combination of voltage and current. A Maximum Power Point Tracker (MPPT) serves as a bridge between load and panel- it loads the panel to its optimum operating point, takes that power (the maximum the panel can deliver), and outputs a different voltage and current that are better matched to the load. In this way, it prevents the load from dragging the solar panel into an operating regime where it would be inefficient- hence the 30% (typical) improvement in performance.

They may be expensive, but they do pay off- all solar cars use them, all solar-powered boats use them, most big solar farms and rooftop installations use them. Only small trickle-chargers, like the kind fitted to many cruising sailboats, do not.

 

Why are Mppt expensive?

 

I guess it depends on what terms you consider expensive. Here is one that costs less than a big panel:
http://www.energymatters.com.au/apo...t-batttery-charger-management-syst-p-872.html
When you are looking at household applications with maybe 6 x 200W panels you get a free panel if you have MPPT because tis will be roughly the benefit gained by optimising recovery. A big panel is worth about AUD1800 so the MPPT is good value.

Now when you look at the electronics required you might have a good case to say they are expensive but they are not yet made by the millions so design costs have to be recovered from a small number.

Rick W

 

Well, here it is Friday morning and I've had "one of those" weeks. Which is to say I didn't get as much done as I wanted to.

But I did manage to formulate a plan of attack in between other work, so let me give you a rough outline here . . . and with a little luck maybe I can get it detailed over the weekend.

The "Front end" circuit will give me a logic signal when there is power stored in a capacitor ready to be harvested, and will automatically take care of the MPPT tracking. What remains is to transfer that power into the battery, with the voltage and current that is appropriate for the battery state, whatever that may be (discharged, charged and floating, etc.).

The LM5118 regulator has a handy "enable" input that shuts the device down to a 10 microAmp standby state, so no problem there, and it also has an input to set the output voltage value. This regulator does boost and buck seamlessly, so my voltage coming in from the solar panel via a capacitor can be thought of as independent of the output voltage and current, for as long as that input power lasts.

So, if I know that my solar panel can put out 2 amps, for example, and I set the regulator to suck off 3 amps, then the capacitor storing power from the solar panel will be sucked down to below the "enable" threshold fairly quickly and the enable will go "false." When the solar panel re-charges the capacitor sufficiently, the enable will go "true" and the process repeats. This sporadic behavior is useful for two reasons: First, bursts of current into the battery are useful for keeping the sulfide (sulfate?) off the plates; and second, the off time is when I want to sample the voltage on the battery to check it's charge condition.

Once I get the battery voltage, I hope to feed it to a LM3914 Dot/bar Display Driver, which is normally used as a "voltmeter" display. I will be using it in the "dot" mode only. Internally this IC has a comparator cascade (10 of them) that trips sequentially in equal steps as the voltage increases at their inputs. So, if I am charging a "12" volt battery, I have a range of roughly 10.5 volts (discharged) to 15.5 volts (fully charged) that I can indicate with half-volt steps . . . which I plan to implement as a battery condition indicator. I will put some circuitry in there to sense when the enable goes false, delay, then sample the voltage briefly. This way the appropriate LED will flash and indicate the battery's voltage state. But in addition, the same outputs that operate the LEDs can be picked off to set or reset a Flip Flop (FF). I think that I only need two voltage settings for my regulator, 15.5 volts and 13.2 volts (or so, I need to check this). The current will take care of itself and be whatever is available. So if the device starts up with the FF calling for 15.5 volts and the actual battery voltage is less than that the comparators will just keep setting the FF to 15.5 volts every time they read the battery voltage. If, however the battery voltage actually gets to 15.5 volts, that comparator output should re-set the FF to the lower, 13.2 volts, and the regulator will self-limit the current to maintain this voltage, i.e., it will be in "float" mode. When a load causes the battery voltage to drop below 13.2 to the next lower step the FF is again set and the process starts over. To summarize, battery voltages less than 13 volts (I need to work on the exact values for these windows) will set the regulator to try and put 15.5 volts into the battery and thus dump maximum current into it. Voltages from 13 to 15 volts will not effect the FF. Voltages above 15 volts will set the FF to deliver 13.2 volts where it will remain until the voltage drops below 13.

Any input on what these voltage values should really be will be appreciated!

And I think that should do it. The "front end" takes care of the MPPT function and also produces a handy pulsing signal used further down the line. The regulator takes care of the battery, and the display driver indicates battery condition. The remaining detail will be how to calculate values for different sized panels and/or batteries, and I think a spreadsheet might be the way to go there. I want to build a prototype before doing that, though. There is a lot of room for errors here.

BillyDoc

 

Some progress, but not done yet . . .

Attached below is a circuit diagram for the control part of the circuit (I hope). It's a first draft and I'll have to breadboard it to be sure I haven't screwed up somewhere, so please be patient! The good news is that it does seem to be possible, and at not much cost.

Let me try to walk through the circuit now.

At the left is the MPPT circuitry as above, with the only new part being a tap on pin 5 of the AD8214, the "enable" output. This feeds a 5 K resistor where 5 volts are developed (the enable output is a 1 mA current) which are fed to pin 6 of an ICM7556 "dual one shot" timer. Pin 6 is a trigger input for the first timer, and triggers on the falling edge of the input. So, when the enable is asserted high, the timer does nothing. Then when the enable is released to ground the falling edge of this signal triggers the first timer (pins 1 through 6) which has timing elements to give a roughly 11 mS pulsed output. This output pulse goes out pin 5 and is connected to pin 8, which is the trigger input for the second timer. Again, it is the falling edge that does the triggering. This second timer has components that give it a roughly 1.1 mS output pulse, which goes out pin 9.

The point of the 11 mS delay followed by a 1.1 mS pulse is to allow the battery to settle for 11 mS after being charged by a string of current pulses from the solar panel (not implemented yet), then to initiate a 1.1 mS "sampling" period where the voltage will be determined and displayed, and if the voltage is 15.5 volts a Flip Flop will be set to reset the charging voltage to the lower float voltage of 13.2 volts.

So, the output from the second timer is wired to the gate input of a MOSFET transistor with a very low on resistance (less than 0.1 ohm). The pulse from this timer turns the MOSFET on, which is to say that it's resistance between the Drain and Source drops to less than 0.1 ohm, thus completing a power path from a LT3010-5 high voltage regulator through some LEDs, and a LM3914 display driver. The battery voltage (Vbatt) has been presented to pin 5 of the LM3914 via a voltage divider, where it is compared to a reference voltage derived from an internal voltage reference by a string of 10 comparators. Only one of these comparators can be "true" at any given time, so the LED attached to that comparator has current drawn through it and lights, thus indicating the voltage level of the battery. When a LED has power applied to it, which means it has it's cathode effectively grounded, this ground may also cause current to be drawn through an associated Schottky diode. Any of the voltage indicating LEDs below 13 volts will cause a 10 K resistor to be pulled to near ground, and at the same time will put a negative going pulse on the "clear" input (pin 1) of the 74AC74 Flip Flop, thus causing the "Charge" output to be set high. Power applied to the LEDs between 13.5 and 15 volts will do nothing to the Flip Flop, but power applied to the 15.5 volt LED will pass through an associated Schottky diode to the "Set" input of the Flip Flop, thus causing the "Float" output to be set high. When the Float output is high a pulse from the IRLD024 causes an eleventh LED to pulse as well, thus indicating that the battery is on float charge. This same Float output will be used to toggle the charging voltage back to 13.2 volts when we get to the buck/boost controller, and the Charge output will be used to cause a charging voltage of 15.6 volts.

Two high voltage regulators are used, with the power draw being distributed between them such that there is never more than a few mA being drawn at any given time, and for the pulsed stuff that power is only on briefly.

When the system is in operation the flashing rate of the LEDs will be a good indicator of the rate of power being captured because the input capacitor will be re-charged after every discharge cycle more or less rapidly in direct relation to the solar panel output. The particular LED doing the flashing will indicate the battery voltage, with the Float LED also flashing when in the float phase.

The only thing left now is the buck/boost controller, and for that I will be following the design paradigm given in the data sheet pretty strictly. I think I'm going to take the time to put it into MathCad, then I'll be able to calculate the variables for the four battery voltages anticipated easily (12, 24, 36 and 48 nominal voltages) and can just reduce it to a table of values for you guys.

But not today! The sun is shining and the wife wants to go for a long walk, so I'll get back to it tomorrow. This was the hard part though, the rest is grunt work. Please review and let me know what you think so far!

BillyDoc

 

I understand your MPPT regulator but have a few points on batteries for you....what is happening at the battery is lost in the explanation of what the electronics are doing ( for me)

First a good battery charged and left for 24 hours will show a little over 12.5v..it does not matter the state of charge as long as its over 20%(no load)
If you charge a run down battery it will pass through 14.4v and start to gas until its fully charges at about 15.3v...so if its unattended for say 1 month + you dont want to go past 14.4v or you will loose water ..acid concentration increases and it buggers the plates . If you want to float it then 13.8 v is the typical figure but you would not want to leave a wet battery at that level or you will sulphate the plates ....perhaps each day when the sun comes up it is brought back to 14.4v before being allowed to drop back to 13.8v....Although the battery technology is 1850 its not simple to keep it happy ,,,,