Rebuilding my Solar Power System after 3 years Storage

[WooferHound]

Oh Yeah, I have a bag that is a buck converter Graveyard. Blew a couple with Reverse Polarity and blew a few discovering  that it's not a good idea to put large capacitors on a converters output because of the backfeeding voltage when turned off. I think I'm OK in this situation. The Solar panel power all goes through Diodes before being combined, then run through a buck converter and a fuse before going on to the battery box.

I would like to know more about the setup you are describing using a Boost Converter as an MPPT. Most of my Boost/Buck converters have Constant Current Control.
Tried to do a search of this site but the Search Box at the top of the page doesen't work anymore.

[eidolon]

In case you haven't seen this, this is the generalized circuit to to convert any converter to power point.  My favorite is these 30-72V 150W 15A converters for $5 that produce 12V exactly. They can honestly produce 5A all day and several can be put in parallel Easy to boost up with a couple diodes in series with the 12V zener.  One of the last converters that is easily modifiable. Interesting design that uses the switcher chip as a high side driver.  An opto provides feedback and is easy todrive from external sources without conflict.  It fakes the converter into thinking there is an over voltage condition and shuts it off.  I use one on my 60V array to maintain a battery set to 14.5V (not 20V) which powers a $4 charge controller. The rest of the panel power goes to heat water. Yup, isolate with a diode if connecting it to battery directly.

[lighthunter]

Eidolon, good to hear from you! You are a genius. I did by the way finally take your suggestion and get into arduino programming this last winter. It is a blast. I still cant believe how powerful the nano is and c++ in some ways is excellent for automation programming. In other words, i almost prefer it over plc programming now .

Thanks for posting the diagram on the converter mod!  I cant wrap my head around how this works (ive not used TL431).   To me, when panel voltage goes above preset limit, the 431 conducts, sending voltage to the vsense shutting down the buck or boost converter. This would be the opposite of what we want. When panel volts goes too high, we want the vsense to go lower and run the converter harder.  Maybe thats they way it works and im missing something.

I was thinking of the inverting 741 op amp voltage comparator.  Normally, i dont want to disturb the original voltage regulation so i leave that alone and modify the current limit input to the converter. Either way if done right, it will work perfectly as you say!

[eidolon]

The TL431 is operated in a way that is different from usually seen.  AT high pv voltages it shunts the voltage to common and the buck converter operates normally.  The diodes and LED form a zener that provides an offset to that voltage sincr the TL431 can only saturate to about 2V at best.  It is best to keep saturation to about 3V.  A transistor or FET could be used just as well in the circuit but are prone to drifting with temperature. It might be easier to visualize the circuit with one of those in it.  Certainly an op amp or comparator could be used and then temperature compensation for panels added.

I've promoted a form of linear programming for the arduino where everything id just done in sequence and each line stands on its own. Nothing but IF statements and the ever popular MAP. It is a little slower, but much easier to understand. Face it, 99.99% will never become "real programmers".  I am frustrated with all those who try to show how smart they are with little programming tricks making programming seem like an impossibility. I think this discourages a lot of people who could better manage their energy.

Have you seen my water heating circuits? One is NANO. The other is a realy easy circuit that allows the heater to be connected in parallel with the panel array and harvest any excess energy not used to charge batteries.

[lighthunter]

"AT high pv voltages it shunts the voltage to common and the buck converter operates normally." 

Ahh, yes now it makes sense, thanks for the explanation.


   "I think this discourages a lot of people who could better manage their energy."

For sure, yeah they do put people down over on the arduino forum. Its kinda like a grade school environment sometimes
I  was especially surprised the IDE is not hard to use and makes debugging the program much easier than i remembered, (i took a pascal programming class years ago), c++ actually has many similarities. Im not good at it by any stretch but i can get a task accomplished tho it be the long way. Actually got one of those cheap hobby servos mounted in a 3D printed frame with a needle valve giving me high resolution modulated propane burner control to the degree. That option costs quite a bit in a purchased appliance. Of course all the pre-made libraries simplify programming a lot.

"Have you seen my water heating circuits?"

No im guessing you posted somewhere, I will have a look. I cant say as im doing it very efficient. I just matched array to the element in full sun and when its not putting out a lot (clouds etc)  i switch the array over to the 2nd input of the grid  tie mppt. I hafta use make before brake timing of relays or the 300vdc gets real rough on the contacts

 "One is NANO, the other...."

Thats interesting, i tried, paralleling two microinverters the other day, they were identical but one dominated the other throwing off the tracking voltage fairly bad.

Woof, please keep posting your progress. Hopefully some of this will help. You could even control your boost converter with arduino, tho might b more than necessary.

[eidolon]

When time eases up I'll do a long series on the discrete water heater board which is really cheap.  I'll let this be the depository for the build.

[MadScientist267]

That's essentially the classic "pseudo mppt" design. Drift however is almost a moot point, as none of these circuits actually "track" actual power. The PV drifts more than the components in the control ever will.

Anything that "OR's" the input and output voltages and pulls the PWM of the converter back when output voltage is hit or when input voltage falls below threshold accomplishes the base functionality... Much better than direct connection or a simple converter... But not as well as a true tracking scheme that hunts for the sweet spot periodically.

As for the TL431 implementation, yep sometimes you gotta get crafty hehe... That said, yes, a TL431 and an opto work well in tandem on this, and the isolation allows for super simplicity on literally any converter topology (no jumping thru hoops, just have the transistor side tug on the Vfb pin all but directly). The most stable version of it I've messed with involved a TL431 on either side into a relatively low impedance divider to set up the Vfb.

Either way, always interesting to see another version of this circuit... They're very useful where the cost (or parasitic nature) of a true MPPT can't be justified.

[eidolon]

It is not quite like that. A lot od MPPT is just crap and doesn't really operate at the true power point. As there is a bypass diode about every 10V on a panel and I've seen controllers sync to lower voltages and stay there in a shading situation even after it passes.  One of my MPPT controllers has a micro monitoring the panel voltage and when it drops low for a period of time it disconnects the panel to reset it.  Immediately higher wattage and voltage. MPPT is solar for muggles so they can connect any panel to any battery and not know anything. When heating water, voltage is everything. Constant voltage ends up just as good as power point at a lower voltage.  Constant voltage can track.

[WooferHound]

Building and installing my two Lithium Power Banks.

[WooferHound]

I'm still having trouble getting my new Lithium Battery Bank to take a full charge. When Full it should read close to 21 volts, the highest I have seen it is 19.5 volts (3.9v per cell) which is only 25% according to the BatteryGO monitor.
For the last 2 weeks I have left the 6 Solar Panels in Series Pairs. On a Sunny day with heavy haze the total Panel output was measured open circuit 42 volts at 2.5 amps. The battery bank should be around 20 amphours. Right now the panels are connected directly to the battery bank, bypassing the charge protections.  Haven't messed around with the Buck / Boost converters again for charging. Have been looking at some small MPPT controllers.

I have fully converted the system to work at any voltage between 12 & 22 using Buck converters to control the voltage or current for every individual load.
My Audio Amplifier is the only item that is running straight off the battery, it will work up to 25vdc. I made it from an Amp Module and is a really beefy little thing at 15 watts per channel. Added a green indicator light to it while I was inside it to check it's voltage capability. It sounds louder and cleaner than ever running at the higher voltage.

[eidolon]

I have a couple of flush mount ceiling LED lights that work off those same boost converters.  The commercial lamps are so cheap now. I just rip out their inverter and put one of these in.  They normally work on 41V, but are more than bright enough at what these put out.

I just got a small stereo that has a 120V wall wart like converter built in.  At 50V panel power it will run. It does take about 15 seconds before it will start up.  Need to change that 2 meg startup resistor to 100K.

[dochubert]

Hey Wooferhound!
How goes the battery charging?
You probably have it humming along by now.

Anyway, best wishes on your system's success!

[WooferHound]

Hey Wooferhound!
How goes the battery charging?
You probably have it humming along by now.

I have been feeling a bit slow lately and lost a lot of energy and motivation, to the point that I didn't want to do anything. Not sure about what my problem is, maybe Thyroid or High Blood Pressure. Well that is over now and I'm feeling more pushed to get things accomplished.

But old age is catching up to me. I'm 62 years old and and my ankles, knees and hips are starting to go. I was riding my bicycle almost exclusively for the last couple of years, but last month I bought a small cargo van that was built to be a Postal Delivery vehicle. The bike doesn't get much use now and I'm starting to gain a little weight. One thing that I noticed was that I never got sick during the time that I was riding.

So what I'm trying to say is that I haven't done anything to my Solar Power system since I finished converting everything to run on the odd 21 volt battery bank. But the outdoor temperatures are sometimes going below 90F - 32C degrees and I'm feeling better so it's time to Get Back To It !

The the 5S 9P Lithium battery bank has been only charging to a max of about 19 volts over the high Sun Summer months. The Solar panels have been set to the Spring/Fall Solstice angle which  is coming up soon on September 23rd 2019, so the Sun angle will be great for a coupla months. It has been working better too. In the last week I have seen it charge up to 19.5 frequently and over 20 volts a few times. This is plenty of power for my small hobby system since I rarely use more than 1 amp load for over 10 hours.

Now I am a little bit concerned cause it is working better. It was working so poorly that I never bothered to add a charge controller, but there is battery Overcharge Protection which will disconnect the battery if any cell goes above 4.25 volts. If that happens then the system will jump up to the 42 volt Solar Panel voltage which is more than any of the Buck Converters are rated for.

[WooferHound]

Well Well . . .
It has been a clear cloudless sunny day and I am getting a 20.6 volt (4.1v per cell) reading on the battery bank. The battery monitor shows 100% charge with a 70mv difference in cell charge.
Will be drinking beer tonight and checking out how much power that I can store now.

[WooferHound]

I figured out that I just need a second Charge/Discharge Protection board, also does balance.
Will charge the Battery bank through one protection board which will use the Overcharge protection and cutoff the panels when charged.
And discharge through the second protection board which will use the discharge protection and cutoff the batteries when discharged.
Ordered a second identical protection board as the on I'm currently using which should control everything properly.
Plus they both do Balancing, so the single board 60ma balancing will double to a more appropriate 120ma balancing current with both protection boards.

https://www.ebay.com/itm/5S-20A-w-Balance-Li-ion-Lithium-18650-Battery-Charger-BMS-Protection-PCB-Board/264077586312?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2060353.m2749.l2649