If there is one topic that is new to me, it’s battery banks. I’m going to use this post to dump some useful discussions for future reference.
Those are some really good points!
Could you add capacitors to your buck converter’s outputs for the Pi?
I don’t have a stock of components, otherwise I would have added a cap and an inductor. I’ve only recently been learning electronics in the real world (past just watching YouTube videos). This is useful, thank you.
On another note, you estimate of 1.5 hours to charge your battery up from 75% suggests you’re not using the full current capacity of your generator. Is that by design to reduce load on the generator or strain on the battery? Or is it a limit of your charger and wiring?
That’s a good question! I became aware of so much battery stuff over in my crosspost where I learned that charging my bank to only 95% was bad news. https://www.reddit.com/r/vandwellers/comments/awgta3/first_weekend_fully_under_solargenerator_power/
I think I’ll be brain dumping all I’ve learned onto my blog as a notepad so I don’t forget.
Honestly, I don’t know all of the specifics! I mean, my AIMS inverter claims to charge at up to 85A (and I have it cranked up all the way), I’m using 4/0 Gauge wire (as big as my thumb) and have MIG welded, used an angle grinder off the setup. Their claims of pulling up to 7,500W for up to 20 seconds are valid, and nothing seemed hot or failed (low frequency inverter) while I load tested.
At 60% I can usually push 70-75A into the bank (specs are PICOGLF25W12V120AL 85+/-5A, and a maximum charge current of 80A per battery), with in mind the temperatures are between 15F-25F this weekend (battery seems to hover around 40F).
I think in this case it’s more of a limitation of the inverter’s charger. I still need to learn what C ratings mean and more battery jazz. On the plus side, my inverter and solar charge controller aren’t compatible to be networked together (should have got a Victron inverter, but $500 vs $1,500), so I don’t have to worry about pumping more than the combined output of 135A (batteries are good for 160A MAX).
Adding a second, parallel battery (and charger) would allow you to double your charging current at no additional stress on an individual battery and half your charge/discharge cycles thereby doubling the service life of each battery. Your generator would run just as long but more efficiently and half as often.
Ditto! A LifePO4 battery has awesome current soaking compatibilities. They pretty much accept max current throughout the whole charging cycle, instead of AGMs which raise internal resistance the more they are charged. I learned this from an Amazon review, some person prefers to throw an LifePO4 battery in the mix with their lead batteries to be available on cloudy days to accept maximum current, and increase lead lifespan.
I would love to throw another 400Ah into the mix of LifePO4. Like you said (and that I didn’t think of before), I could dump in a lithium specific charger and maximize my generator output. I could also run them down 80% with good cycle numbers.
Some problems though. I’d need to add temperature dependent charging monitoring, as I read that charging lithium in freezing conditions damage them. Meaning a transfer switch and/or battery warmer would be required (latest AWD Priuses use Ni-cad for this reason). I’d need to be able to intelligently manage my charge current so I don’t exceed (and trip) my generator, as it simply stops generating when it’s overloaded (instead, just running at idle until restarted). Victron inverters have the ability to drop off charging depending on load… Yet another reason why I should have spent the $$$.
So it seems the long term plan is to eventually add some LifePO4 batteries, swap out the inverter for a Victron Inverter, switch my bank voltage to 48V or something higher than 12V (wiring gauge cost savings).
I wonder if I can find a LifePO4 who’s BMS is temperature monitored and can cut off charge if the temperature is too low, so I don’t have to account for it externally. Just dumping in 100Ah or two packs would be nice in the mid term.
I have no idea yet! Sadly, it’s cloudy and gloomy out here on my first test weekend, with snow scheduled for tomorrow.
I do know that I can usually pull 600W on a sunny day, 200W on a cloudy one. I’m sure that I could run off just solar if I switched the fridge to Propane and was mindful of my usage.
The reason I asked is because lead acid batteries don’t like partial state of charge cycling. Lithiums don’t seem to mind it but leads get a capacity loss. If you only recharge to 95% that becomes the new full.
Think of a dry sponge standing on edge. As you spritz it on one side with a squirt bottle it absorbs the water. That’s like lead plates taking up sulfate ions becoming lead sulfate as a battery discharges. At first the outer surface gets moist then the moisture gets deeper as you spritz it more. When you charge the battery it is like blowing it with a hair dryer. Quickly the surface dries but the drying process slows as the remaining moisture is deeper inside the sponge. While holding 14.4 volts the charging current tapers away. Eventually it dries. If you moisten it 60% full of water, 60% discharged, and then dry it to 95% dry, 5% still wet, there is some moisture remaining. The remaining moisture is deep. Then if you moisten and remove 10% that buried 5% is still there. Lead sulfate hardens and cannot be removed. Think of mold growing in your sponge. Once the mold grows and takes up that deeply buried 5% moisture it is stuck as the mold will hold the moisture even with the surrounding sponge dry. As you apply charging voltage, blow with the hair dryer, you can only reach the surface directly.
Your 100 amp hour battery is now a 95 amp hour battery. Repeat monthly and the once proud 100 amp battery is only 95-4.75=90.25 amp hour. Next month it is only 85.7375 amp hour.
With AGM batteries you cannot use a hydrometer to measure the electrolyte to determine how much sulfur is in the sulfuric acid and how much is still buried in the plates. The only thing you can do is fully charge the battery, as described in my prior post, and reset your coulomb counter. When you set up your battery monitor one of the things you told it was your battery capacity. If you were to take 60 amp hours from a 100 amp hour battery that’s 60%. If you take 60 amp hours from a 4 month old 81.45 amp hour battery that’s almost 74% not 60%. The situation spirals. What you see is the charging puts in amp hours and the battery voltage goes up and everything is good. As your capacity vanishes, in your case, the generator will start more often. Without the autostart generator everything is fine until the fridge beeps and gives a low battery voltage error code. Then you are puzzled. The solar is charging and the battery is full by 10AM. It used to take until noon but now it works better than ever, full by 10. The green light on the charge controller says full and the voltmeter says 14.4 then float mode 13.8 so nothing is wrong. But the fridge won’t run past 3 AM. In your case the generator used to take 1.5 hours but now it fills the battery in only 30 minutes, better than ever.
This is, in my opinion, the main benefit of LiFePo4 batteries. They can go 10 years cycling in the 20% to 80% range with no need to ever get to 100%.
With flooded lead acid batteries you can see the problem developing. Specific gravity drops and water consumption rises. you can do something about it. You can discharge to 10.5 volts, leave stuff on till it’s dead. Get that sponge wet all the way through then recharge using an elevated absorption voltage, up to 16 volts, and get back some of the lost capacity. While doing that don’t overheat the battery and keep adding distilled water. It will take a lot.ReplyGive AwardShareReportSavelevel 3zorrobyte1 point· 14 hours ago
You may have just saved my bacon https://www.practical-sailor.com/blog/Dont-Kill-That-New-AGM-Battery-11662-1.html
Damn batteries. I swear. I’ll program my controller for 100% charge.
Love bacon, must save bacon.
That’s a nice web page that practical sailor.
Practical sailor tips 1 and 2 are right on.
Tip 3, only 50% discharge, I’m not so sure. If you have a 100 ah battery and cycle to 70% according to the vendor’s chart you should expect N cycles. With 35% discharge you should get somewhere around 2N cycles. Not exactly 2N, probably better than 2N. The price per cycle would be better with 2 batteries of 100 ah discharging each 35 ah rather than one going 70 ah. However, you have twice the space and twice the weight. You have twice the money tied up from day one. The beyond 50% discharge doesn’t hit a switch and kill the battery.
In your case where you can set the generator to start at 30% or 60% there’s not really much difference. Since the charging takes longer at lower current with a deeper cycle it would matter if you had to run the generator longer. If you use solar to finish the charge but get to 80% quickly with the generator 30% or 60% won’t matter much. Sailing is different.
The tip about knowing your state of charge is pretty far down the list. I don’t have a Victron but it is my understanding that they need to be told when the battery gets to full or they get farther off with each cycle. Given that the Peukert effect makes input and output amp hours not equivalent it seems reasonable that maintenance there would be necessary. Given your dependence on the Victron it may deserve a position higher on your list.
> Tip 3, only 50% discharge, I’m not so sure. If you have a 100 ah battery and cycle to 70% according to the vendor’s chart you should expect N cycles. With 35% discharge you should get somewhere around 2N cycles. Not exactly 2N, probably better than 2N. The price per cycle would be better with 2 batteries of 100 ah discharging each 35 ah rather than one going 70 ah. However, you have twice the space and twice the weight. You have twice the money tied up from day one. The beyond 50% discharge doesn’t hit a switch and kill the battery.
I went with that advise and put in 400Ah of Capacity, as I wanted 200Ah usable.
>In your case where you can set the generator to start at 30% or 60% there’s not really much difference. Since the charging takes longer at lower current with a deeper cycle it would matter if you had to run the generator longer. If you use solar to finish the charge but get to 80% quickly with the generator 30% or 60% won’t matter much. Sailing is different.
So, you are saying that since I can immediately recharge the bank that I can get away with deeper discharges? Is the difference letting the battery stay at a deep discharge for a period of time, like in sailing for other typical uses? Maybe I’m misunderstanding here, are you saying that it shouldn’t matter over a 24 hour period if I run the generator to charge up to only 95%, as the solar should take over in the morning and ensure 100%?
I may be misunderstanding here. I draw 50% DoD to expect N cycles before dropping to 80% life (or whatever that number is, 70%). I also keep worst case in mind, I only allow my inverter to draw up to 50% to have reserve capacity in case of an emergency, and I don’t “trust” my solar, generator to be immediately available to recharge the bank.
>The tip about knowing your state of charge is pretty far down the list. I don’t have a Victron but it is my understanding that they need to be told when the battery gets to full or they get farther off with each cycle. Given that the Peukert effect makes input and output amp hours not equivalent it seems reasonable that maintenance there would be necessary. Given your dependence on the Victron it may deserve a position higher on your list.
Peukert effect. This rings a bell. The Victron 712 goes on about how they account and track it. I also added the thermal probe for the Victron, which is supposed to help calculations of some sort. While you can manually sync the 712, it can sync itself by keeping track of charging voltages. Since I have a Venus GX (and also Bluetooth in the solar controller/BMS, in case I didn’t want to use the Venus), Victron advertises that the data is supposed to be shared between their equipment to make everything more efficient.
It’s obvious that I have more learning when it comes to battery bank health and maintenance. The fact that they aren’t these “reliable, solid state components” kind of irk me, but it’s not like I can afford 400Ah of Graphene Supercapicators.