Rebuilding my Solar Power System after 3 years Storage

[WooferHound]

Oh Yeah , I'm having fun buying stuff and posting pictures of it
 but
I am working so much in the last coupla weeks that there is no time for sleeping

Looks like it is slowing down enough for me to start drilling holes and making stuff
A grocery bag at my feet has all the parts to make the Lithium Battery Banks
 and
I'm anxious to convert the system over to 14-21 volts operation
 plus
the yard needs mowed, dogs need played, have new tires and tubes to go on the bike, etc . . .

[WooferHound]

I am working so much in the last coupla weeks that there is no time for sleeping

(Attachment Link)

My work schedule has not been any better. I only have a couple hours freetime daily and then stuck at work for the remainder of the day. Today I will be stuck in Dance Recital Hell for 14 hours, and this is the 4th day. There should be over 1200 people in the theater, 90% of them will be girls  of all ages, so there is some pleasantness. Barely had enough time last week to build that Lithium 18650 Charger.

I am still focused on building a couple of 18650 battery banks and converting the system over to 21 volts.

[WooferHound]

Finally got a few days off and built two 5s 8p battery banks.  Put the finishing touches on it this afternoon and went ahead and hooked them up to the system and seems to be working out very good.



These are built onto some plastic that is much like 1/4 inch thick truck bed liner. The Check & Balance wiring comes off the top, the Charge & Discharge wiring comes off the bottom. They are designed to stack  but will be used on their sides.
After I built the 18650 charger a month ago, all the batteries were topped up to 4.2 volts. Today I tested them and sorted them all by voltage High to Low. When putting them into the battery holders I the series rows with the high volt cells and worked across ending with the low volt cells, so each parallel section has a mix of low to high volt cells. After a month sitting, the cell volts varied between 4.10 - 4.18, not a big difference. when installed I will put half the batteries in each bank so they will have more even loading.



Every single cell has it's own dedicated 5 amp fuse to prevent disaster if any cell shorts or goes Reverse Polarity. The battery holders were not made for these longer 18650 cells originally deforming in order to accommodate the batteries. I just used some pliers on the Terminals and squashed them enough to work great  fully loaded.



Here is the terminal strips that will distribute the Negative and Positive power Ins and Outs. The yellow XT60 connector is part of a temporary setup to allow me to plug this into the system from here inside the house for testing.



This terminal strip is the Check and Balance connections. For testing it has two JST 5s test and balance connectors on it, one is plugged into the BatteryGO monitor/balancer, the other is for the charge controller which is also a balancer. Once installed there will also be another set of wires going into the house so I can use the BatteryGO monitor in there too. In the picture the readings are ...
20.36 volts total
19 mV difference between the lowest volt parallel bank and the highest
97% battery capacity
4.067, 4.064, 4.083, 4.066, 4.081 - individual Parallel bank voltages

It's all on my desk here, hooked up and powering the system. seems to be some good stiff power. Getting dark outside so I'll be looking at it closely over the next few hours to see what it's got ?

[WooferHound]

Yesterday I moved these two battery banks into the cooler that they were designed to fit into. At this time there are 9 parallel battery sets with a capability of 16 Parallel and 5 series making an optimistic total of 21 volts  at 40 amphours if filled with batteries. This will all be reduced to 12 volts or less using buck converters on every individual load, so there will be some amphour gain there.



I've been using this bank for 2 days now, the first day was on my desk where I could check on it frequently, and then all yesterday installed in it's permanent location under the solar panels outside.

When I was running on Wild Power using Capacitors instead of batteries, the highest voltage that the solar panels could achieve without load was 20.5 volts. Now with this Lithium Bank the highest charge seen so far is 19.1 volts, so it will likely be difficult to charge much higher using these panels. I've got 5 more of these panels that are New Old Stock and plan on adding them after getting all this stuff converted over to work on 21 volts.

This morning I woke up discovering some strange voltage readings, after some checking around I'm seeing that the Charge Controller/Balancer seems to have Cut Off. It's designed to cutoff for for the following conditions...
Over charging detection voltage: 4.250±0.040V
Over charging recovery voltage: 4.190±0.050V
Over discharge detection voltage: 2.800±0.080V
Over discharge recovery voltage: 3.000±0.100V
This is an $8 charge controller, a small PCBoard without any adjustment or indicators. I can work around it, but obviously need a better one.
Fortunately the Balance wiring is run inside too, so the BatteryGO Monitor/ Balancer can be used anytime for close inspection.

[WooferHound]

I am in the middle of another 4 days of Dance Recital Hell, worked 16 hours yesterday and should be 14 hours today and tomorrow.

The system seems to be working fairly good. In the last post, I was complaining about the Charge Controller in Cut Off, but after a couple of days of observation, it seems to be working as advertised and cutting off properly if any Parallel bank gets below 2.8 volts. Then it waits till all 5 banks are above 3 volts before restoring to normal operation.

Seems that I may have made a Cell weak by accidentally putting it in Reversed Polarity and that bank of 9 cells shows a difference in the readings.
Otherwise  I\m real happy with the results.

[lighthunter]

Sorry for your work schedule. That can get old real fast. I was in a similar situation several years back and it isnt fun.
Proud of you for taking the pain to make these cells work for you. I am actually looking into 26650 cylindrical cells for the car project. I can cut the new cost of LFP in half by going that route over prismatic (plate pkg format). Of course there would be lots more DIY time involved, the benefits would be cheaper, shape configuration and voltage configuration is customizable. The payback even with solar charging is pretty meager like several years to return any investment on $980 of batteries = 400 cells = 4kwh LFP. the catch is it would take some time, materials and spot welder to connect all those cells. I didnt catch how you connected cells?  High power solder iron to minimize heat transfer can work i guess. Good project i love the efficiency of lithium, an electron charged is an electron available later For smaller systems or battery powered loads it really shows up!!

[WooferHound]

LightHunter >> I used battery holders and just snapped the cells in there, No Soldering, super cheap and very flexible.

I have converted totally over to 21 volt lithium power now by removing a Buck converter that was holding the output voltage down to 12.6. But it looks like my 21 volt solar panels will only push the charge up to about 18 volts max (3.6v per cell), my BatteryGO Monitor says that is only 8% full. Today I will wire the panels in series to see if I can get them to charge up any higher. This will lead to more problems eventually as it will then be over 40 volts for charging the 21 volt battery bank.

The batteries don't really have a Charge Controller right now. It's really a 5s Protection and Balancing board that will has a
discharge cut-off if a cell goes below 2.8v or 14v for the full bank
Overcharge cut-off if a cell goes over 4.2v or 21v for the full bank
Max Balancing Current is 42ma
So if the Overcharge cut-off Kicks-in when it charges up to 21v, The the voltage on the input would become Open Circuit at 42v.

I still have devices connected that don't want that high voltage. They are switched and I have tape tags on the switches that shouldn't be used until I can deal with them individually.
I've tried using 2 of those super small Mini Buck Converters that were rated 1.5 amps. One was used on a red LED string 5 in series and adjusted to 9 volts at 20ma and it is working great. Tried to use one of them to power a 12v Muffin Fan rated .6 amp 7 watts. The Coil on the converter got hot to the touch and it quit completely after about a minute. Ended up using one of my regular 3 amp converters without problem.

[lighthunter]

LightHunter >> I used battery holders and just snapped the cells in there, No Soldering, super cheap and very flexible.

Ahh, yes, that is perfect solution for your application!!

I have converted totally over to 21 volt lithium power now by removing a Buck converter that was holding the output voltage down to 12.6. But it looks like my 21 volt solar panels will only push the charge up to about 18 volts max (3.6v per cell), my BatteryGO Monitor says that is only 8% full. Today I will wire the panels in series to see if I can get them to charge up any higher. This will lead to more problems eventually as it will then be over 40 volts for charging the 21 volt battery bank.

[/quote]


Werent you using buck regulator to downconvert voltage from panels? If not, thats your solution, you can even modify them to do mppt, seems like someone here has done that if i remember right. Or if you have to buy one anyway why not get an auto buckboost model and then put your panels all in parallel again. It may come out more efficient since the volts wouldnt have to be boosted or bucked much.

[WooferHound]

Werent you using buck regulator to downconvert voltage from panels?
If not, thats your solution, you can even modify them to do mppt, seems like someone here has done that if i remember right.
Or if you have to buy one anyway why not get an auto buckboost model and then put your panels all in parallel again. It may come out more efficient since the volts wouldnt have to be boosted or bucked much.

Since I don't have much battery capacity I can drain the charge to nothing too easily, plus going a week without the Sun is common around here. So I made a voltage Maintainer which uses Mains Power to keep the voltage from going below a set point, protecting the batteries and allowing me unlimited use of the system. This uses a 12.6v transformer which is about 18vdc after rectification. That is where I'm using a Buck Converter, at this time it is set to 14v (2.8v per cell) for the new Lithium power banks.

YES ,  that is a great idea to wire the panels in parallel again and use a Buck/Boost Converter to push it all toward the 21v target. In fact I ordered one last night after I saw your suggestion. It is rated 7 amps, 10 peak. Should be here within a week. Was a little worried how well it will work with raw Solar Panels on the Input. It also has Current Control, but probably won't be needed.

Yesterday morning I did wire the panels in Series. There are 6 panels, 21 volts open, about 1 amp each. It greatly speeded up charge time but seemed to top out at 18.4 volts, battery monitor reading 18% full. It was a clear day under Hazy Sun and I was keeping all the loads off.

[WooferHound]

Remembered that I already have a 6 amp Boost Converter and decided to go ahead and use that since Boost is all that will really be needed.
Rewired all six panels back to Parallel connection, Set the boost converter to 21.1 volts output and wired it between the panels and the batteries.
Today is totally Cloudy with 4-7 inches of rain forecast over the next week, so I probably won't see any results for awhile.

But , I wonder . . .
Until the batteries reach 21v Full Charge, won't the Boost Converter be heavy loading the panels down to some low unusable voltage that can't be Boosted ...

[WooferHound]

Cloudy all day yesterday, but got a few hours of Full Sun Today around Noon.
No ,  a Boost converter, directly connected to raw Solar Panels, will not charge batteries.
It may possibly work using Constant Current mode.
Going back to pairs of my solar panels wired in series. 3 pairs in all totaling 40 volts at 3 amps.

[Pete]

Hi Woof, just wondering why you chose such a high tech approach to your system.
I would have thought that going down the 12 volt battery bank road would have saved lots of complicated electronics.
It would be interesting to know why you chose a 21 volt battery bank, and why you wanted to use boost and buck converters.
I come from the Keep it as simple as possible camp, so run 12 volt batteries, 12 volt lighting, 12 volt power wherever possible and an inverter for 240 volt stuff. Has worked fine for me for the last 40 years of being on solar power.
Anyway it looks like you are having fun so keep on grinning
Pete

[WooferHound]

I had explained it a little bit earlier when I said . . .

Code: [Select]

I have decided to run the whole system at the high Battery Voltage 14-21 volts, then use Buck converters at all the loads to power them.
Most of the loads are LED lights which already use buck converters with current control to drive them, so they will work on anything between 11-32 volts. A small bank of 5v USB sockets are powered with a current limited buck converter too.
The only other items of concern are a 12v computer fan, a Bank of Cigarette lighter Sockets,  and an Audio Amplifier. They can all get buck converters so they work properly. The amp may possibly work OK at 21 volts.
I am up early and actually have some time so I'll ramble on a bit more.

My Lead Acid batteries had all died and I was looking for some replacements. Was pricing more lead acid plus looking at the future for batteries so was pricing Lithium batteries too. After learning more about Lithium batteries, I decided that even with the high prices there were too many good advantages with Lithium. Lead Acid batteries didn't make any since anymore so they are out and now moving on to the future.

Lithium Cells are 3-4.2 volts so it is hard to make a decent 12 volt battery with them. Five of them in series is 21v full charge and exactly fits in the Picnic Cooler that my Lead Acid batteries were in. Not to mention that the solar panels are about 21 volts too.

In the beginning I was going to run a high battery voltage and use One Big Buck Converter reducing everything down to 12.6 volts. Looking at my system I was seeing that most of the loads are LED lights which are Driven based on Current, so they already had Buck Converters with current limiters and could handle 11-32 volts. So if I used the One Big buck converter method, then the power to the LED lights would be going through 2 buck converters losing over 5% each time. That is the main reason for running a High Voltage and dealing with it at the load later.

A friend of mine gave me some Laptop battery banks, I broke them open and recovered over 40 batteries. They seem to be in fairly good condition. I enjoy making stuff, so making the battery banks and getting all this stuff up and running, then using it in real life is really fun to me.

Yes I'm grinning everyday . . .

[WooferHound]

When I changed my panels to Parallel doing the Boost Converter experiment, I blew the fuse that connects to the batteries.
So I do Not Know if that Boost converter would have worked.
It was discovered minutes ago when I was restoring the panels back into Series connection.
I saw the spark but didn't think it would have blown the fuse.

May try it again later but it's 20 minutes of tedious work in a cramped space.

[lighthunter]

oops, i forgot to mention buck and boost converters (at least the ones you are working with) are not meant to connect to a battery since they often use synchronous rectification for efficiency reasons, therefore an external output diode is needed between the output and the battery. Some mfgrs ship a slip of paper with them warning of this. The diode will prevent the short circuit backflow current from battery through the fet when it is powered off.

I do agree with Pete but i also enjoy doing things differently for fun like you Woof.

 The boost converter current control has to be modified to regulate panel voltage, thus you end up with your very own homebuilt mppt... well close enough to it anyway. Actually temp compensated voltage control of panels gets extremely close to true mppt. To test it out you could add the diode and adjust the current limiter for full sun mpv of panel and it would work well as long as full sun on the panels.