Using 4 cell 18650 li-ion with 4 100W solar panel

[Mike90250]

I understand it's a small drone boat in the middle of nowhere. if you can come up with a robust BMS to keep the cells in healthy condition, great. I don't know much about Li other than to beware.
As to the motor, have you considered AC motor with controller. Small electronics to drive it, but better efficiency, Will prop be vari-pitch or is this single speed ?

Any thought of ballast compartment like a sailboat keel ?

Anyway, our job is to give you all the choices for the battery, pros, cons and it's your job to sort it out and chose what you want. And please keep us posted.
(electrics and salt water - yum)

[drpatso]

I don't know if I was extremely clear, this is going to be for an unattended drone boat designed around 400W of panels, using compass control and gps waypoint navigation. The system will power motor, rudder servo and electronics.

I am reluctant to choose 4s LiFePO4 as I will be effectively lowering the capacity charging so far below fully charged voltage before dark as well as increasing volume and weight dramatically compared to other technolgoies. It seems charge controllers capable of running a 24V lead acid / gel bank match up quite well with 7s "Lico"? packs (lithium cells that are 4.2V fully charged,NCR18650B or similar). I think Lico or similar is only risking the boat as it will not have any occupants.

That said, with the available charge controllers it looks like these are my options. I think most likely I will be running a 12V motor, not terribly afraid of running a step down or slightly above 12V to the motor controller. A lot of motor controllers accept a wide voltage range. We already have a nice waterproof buck converter for 5v electronics.

4S LiFePO4 with MPPT set to 14.2-14.4V
5S LiFePO4 (no charge controller? I don't much care for this idea)
8S LiFePO4 with MPPT

7S LiCO (4.2V fully charged) It seems like the problems most people have with this setup is ****ty bms. If a great quality BMS is sourced I think this could be a feasible option)

[Mike90250]

The biggest problem with any of the Li chemistries, is the breakdown that happens at low cell voltages and high voltages. Ask Boeing and Samsung about Li battery fires.
Out of all the different cell types, the LFP / LiFePO4 is the most robust and forgiving. But it's not the highest power density,
If you are charging with a 10A charger, you need balance boards that can handle 10A, not 500mA or 2A.
You need to make sure the spaghetti wiring mess for the BMS does not cause a fire if a wire gets nicked.
The automotive systems have thermal controls and voltage controls to watch over the batteries, are sized just right for the battery pack and the charge and discharge patterns are well known. Taking BMS #S and attaching it to battery #G - may or may not be technically acceptable, but Marketing will sell it, they both worked last week

[inetdog]

...I have been amazed at the lack of options for batteries other than lead and gel.
...
...
Of course fire is a concern but it'll only kill the ship, I have considered lithium polymer as you can get 14.8V 16Ah for 50$ now.

1. Just to keep the language correct to avoid problems later: GEL is a particular form of Lead Acid battery, along with the open system Flooded Lead Acid (FLA) and the sealed, pressure regulated, Absorbed Glass Mat (AGM) form. AGM, being sealed, has advantages for a boat and can also deliver higher power over a short time than FLA. GEL is also sealed but has serious problems accepting charge quickly and is, with a few possible special construction exceptions, considered unsuitable for PV or engine alternator charging.
2. You are using "Lithium-ion" and "Lithium Polymer" (LiPo) as if they described chemistry and associated voltage and energy density. They do not!!!!
Lithium-ion can refer to the use of a soluble lithium compound for an electrode instead of lithium metal. Lithium polymer refers to a particular kind of construction with a soft case and soft separators. They do not tell you what the electrolyte or electrode chemistry are. You need to look at the actual chemistry, such as Lithium Cobalt (LiCo) or Lithium Iron Phosphate (LiFePO4) when making your evaluations. The physical construction as well as the chemistry affect the charge and discharge rate, resistance to thermal runaway, and intrinsic safety of the batteries.
3. OK, you say a failure would only kill the ship. Actually a failing lithium chemistry battery can also produce highly toxic fumes that can kill you before you are able to jump off the ship. Some of these batteries are 100% safe as long as you do not store them at high temperatures or charge or discharge them at any temperature.

For a good discussion of LiPo and Li-ion, check out this thread on CandlepowerForums: http://www.candlepowerforums.com/vb/...=1#post4471076



While we are on the subject, I should also mention some other manufacturer recommended terminology:

Lithium batteries do not catch fire, they vent with flame.
Lithium batteries do not explode, they undergo spontaneous rapid disassembly.

[drpatso]

Appreciate all the input guys. Manual balance is not in the question for me, and I'd like to get better density than LiFePO4. I have been amazed at the lack of options for batteries other than lead and gel.

I have a budget of about 600$ for the battery and charger.

I'll let the cat out of the bag here. We have a hull design in progress and a pid controller on the rudder servo with input from tilt compensated compass, iridium satellite link, GPS, and voltage monitoring. This power system will hopefully be floating around the ocean next summer..

Now I'm at a complete loss.. I just want a light, cheap efficient method of storing the energy from 4 x 100W panels. The guy who made the sea charger has a 5 cell lifepo4 connected to solar panels directly to the 5 cell pack with a balancing circuit, maybe that's not a great method but the thing has been out at sea and running for months and months. Ofcourse fire is a concern but it'll only kill the ship, I have considered lithium polymer as you can get 14.8V 16Ah for 50$ now.

[PNjunction]

I agree - but everyone needs to keep their eye on the ball on the application at hand - and this is where it gets so contentious by forgetting that application matters.

Here, the application is a solar housebank, which differs in usage/design than those for an EV, or an RC model environment - yet many from the latter categories have NO CLUE as to what it takes to successfully run the former!

Additionally, many come to just get a shopping list of online junk to slap onto their cells, and get generic advice, such as a 14.6v charge for LFP, which when given without taking the ops skill level or application into account leads to financial bluners, safety, or performance problems.

If you want info on solar housebank use of LFP, talk to actual owners - even if they disagree amongst themselves - their ownership and experience are immensely better than those of non-owner speculators.

[jflorey2]

Putting together raw cells and component balancers without understanding how they should interact is a disaster waiting to happen.
JMO

Indeed it is - as is true with any battery (or solar) technology. Hopefully forums like this one help people make better decisions on such issues.

[inetdog]

Indeed it does. And yet EV manufacturers, renewable energy manufacturers and battery manufacturers use balancing circuits - and they have much more to lose than your typical hobbyist. (And have the years of experience to make those good decisions.) Personally I will go with them, rather than hobbyists.

You are observing high quality gear sold into a specific market, with generally knowledgable DIY hobbyists or specifying engineers. Or gear designed for the average consumer for a very specific use pattern.

What the PV energy storage experimenter will be exposed to will be a wide range from component parts which must be put together intelligently to cheap "systems" that in fact are not well engineered for his particular use and may be in a very price competitive market.

If the Tesla Powerwall includes balancing and not just monitoring circuitry, as it probably does, I would expect it to be engineered for the discharge regime(s) the unit is expected to be used in, including essentially unlimited access charging power for grid time shifting and limited access time and power charging from a PV source.
Believe me, that is not what you get with a typical "hobbyist" Li pack with built-in BMS. The BMS is there to allow it to be advertised that way, not to do a particularly useful job.
Putting together raw cells and component balancers without understanding how they should interact is a disaster waiting to happen.
JMO

[jflorey2]

Sorry man, can't resist this one about seatbelts - if they are made of straw, then it is a feel-good strategy.

Yep. But they're not.

Here's the deal - put some money on the line, and get an actual LFP prismatic bank. A 4S GBS, CALB, etc bank will do. Becoming an owner kind of changes things. You'll like it, and the eyeballs will glaze over as you investigate top-balance with no boards, top-balance with junk boards, bottom balance and the all fun stuff most other non-owners speculate about. Kinda' changes things when your wallet is on the line.

Indeed it does. And yet EV manufacturers, renewable energy manufacturers and battery manufacturers use balancing circuits - and they have much more to lose than your typical hobbyist. (And have the years of experience to make those good decisions.) Personally I will go with them, rather than hobbyists.

[PNjunction]

It isn't an argument, it's real-world ownership experience. The consumer grade junk you speak about is what the average DIY'er is going to use in the first place, like a Pb based mppt controller on his new LFP, without knowing that you need to disable temp-comp if you are going to shoehorn the engineering.

This is the argument. Let's use NiFe as an analogy:

1) Op buys NiFe batteries and wants to know how to charge and maintain them.
2) Based upon specs alone, and being a well-intentioned non-owner, I just parrot back what I read on the net.
3) Consult an actual owner, like Mike, for proper guidance.

I'd choose option #3.

This is why you don't see me making any recommendations about NiFe that I can't back up. I don't own any, so you shouldn't trust me until I do!

Sorry man, can't resist this one about seatbelts - if they are made of straw, then it is a feel-good strategy.

Here's the deal - put some money on the line, and get an actual LFP prismatic bank. A 4S GBS, CALB, etc bank will do. Becoming an owner kind of changes things. You'll like it, and the eyeballs will glaze over as you investigate top-balance with no boards, top-balance with junk boards, bottom balance and the all fun stuff most other non-owners speculate about. Kinda' changes things when your wallet is on the line.

[jflorey2]

Bzzt. Most balancing circuits are the CAUSE for killing packs. The "balancers" are what kills them.

Completely wrong. You are talking about consumer grade junk, not the balancing circuits used in most lithium-ion batteries. Contrary to what some think, battery, EV and renewables designers are not fools, and would not add devices to their batteries that reduce their life (and thus cost them money.)

The balancing circuits are too small in the first place, vastly extending the time to do a so-called top balance and turn into vampire parasitic draw cell killers. They are frequently out of spec within regards to each other. And when they fail, all hell breaks loose, unbeknownst to the owner who likes watching blinky lights. Or watch their EV which operated ok for a month go into flames as their infrastructure burns first, regardless of which li-ion chemistry they choose.

Your argument is akin to claiming that most seatbelts are what cause traffic fatalities, based on stories of seatbelts trapping people in their cars, and homemade seatbelts doing more good than harm in accidents.

[PNjunction]

In general, yes. Most lithium battery BMSes provide some kind of balancing. This can be very simple (accurate zener function at 4.2 volts per cell) or complex (charge transfer.) They are often not needed for packs that start out well balanced, but will generally prolong the life of an average pack.

Bzzt. Most balancing circuits are the CAUSE for killing packs. The "balancers" are what kills them. The balancing circuits are too small in the first place, vastly extending the time to do a so-called top balance and turn into vampire parasitic draw cell killers. They are frequently out of spec within regards to each other. And when they fail, all hell breaks loose, unbeknownst to the owner who likes watching blinky lights. Or watch their EV which operated ok for a month go into flames as their infrastructure burns first, regardless of which li-ion chemistry they choose.

This lesson was learned many years ago. Most newcomers to LFP and lithium in general disregard the effects of TIME, coming from a Pb based world to their banks detriment.

In other words, even if the balancing circuits don't fail, they are usually too feeble to do the job quickly enough cycle to cycle, extending the time the OTHER cells spend at high voltage twiddling their thumbs waiting for the feeble balancer to bring the lower cell(s) up. This time factor is what the marketers of balancing boards don't tell you.

Well, we know one thing. No mention was made about DISABLING temp-comp with your typical Pb based controller you mentioned, which usually incorporates that.

You know that you can fully charge a 4S LFP cell anywhere from 13.8v to 14.6v right - the only difference being that of time to absorb to reach full charge. Again TIME is the factor here, so a fixed voltage is not always the right answer with LFP. It depends on your application. and of course assuming good balance in the first place.

Much of this has been discussed elsewhere here.

My biggest issue is that we can't just toss out a laundry list of things to buy to someone who doesn't have a clue about li-ion in the first place. It isn't that easy from a financial and safety standpoint.

The second biggest problem I personally faced when getting my LFP setup going, is that most of the material passed to me came from well intentioned NON-OWNERS, who were just repeating back tech specs that "should" work, when in fact they were wrong, or at most ineffective.

[jflorey2]

Thanks Mike, scratch the HQST MPPT charger. I guess my last 2 questions are:
When I find a good charge controller when using potentially LFP 26650s or huge single LFP cells in a 4s pack, do I need some kind of balance circuit on the pack?

In general, yes. Most lithium battery BMSes provide some kind of balancing. This can be very simple (accurate zener function at 4.2 volts per cell) or complex (charge transfer.) They are often not needed for packs that start out well balanced, but will generally prolong the life of an average pack.

The issue here is that while lead-acid batteries can deal with overcharge by boiling off electrolyte (which can then be replaced) lithium ion batteries have no such mechanism. So if one cell has a slightly higher leakage than the others, it will drop in voltage, and will continue to drop until it is almost dead even when all the other cells are fully charged. At that point the BMS will look at cell voltages and shut the pack down to prevent overcharge and overdischarge. That will greatly restrict the usable range of the pack. (And while some BMSes do not offer balancing, ALL lithium chemistry BMSes must offer over/under voltage monitoring.)

People who use large lithium cells (generally LiFePO4) for home storage sometimes balance them manually, then forego any balancing at all in favor of monitoring the cells. If the cells get out of balance they rebalance them manually. Several people have had good luck with this approach.

So for 4S you are generally going to want to use balancing. There are plenty of good BMSes out there that will do this for you.

Is there a charge controller you guys might recommend for 400W ( 3 x 12v 100w panels)? I was thinking 30 amps, since the panels would go over that maybe for an hour mid summer.

You are going to want an MPPT charge controller just because of its better control of output voltage and current. Make sure the input can deal with your voltage from the 3 panels in series. Then you are going to want to set the voltage to approx. 16.8V for a 4S lithium cobalt pack (or 14.6V for a 4S LiFePO4) and you are going to want to set the max output current to less than the maximum C rate of the pack. If the pack is 10ah and has a maximum charge rate of 1.5C then limit the maximum current to 15 amps.

[karrak]

As far as I can see the HQST MPPT controller is actually a Tracer MPPT controller made by EPSolar who are based in China. I have installed a more upmarket version of their MPPT controller in a friend's LFP offgrid system that has been working well for over three years.

Ideally for a 12 volt LFP battery for the longest lifespan you want to set the charge/bulk/boost voltage to 13.8 volts and the float voltage to 13.4 volts. At a pinch you can use the "Gel" settings. I think you can set user programmable values via an external programmer. Here is a link to the manual https://www.epsolarpv.com/en/index.p...wnloadF/id/678

As with your RC batteries which are most likely Lithium Cobalt (LCO) batteries you have to balance an LFP battery before you use it and have some way of checking that it remains in balance and that the individual cell voltages do not go outside the voltage range 3.65-2.50 volts.

One big advantage of LFP batteries over LCO batteries is that you will get maybe ten times the lifespan if you use them conservatively. The batteries in the Powerwall are another lithium-Ion variant, Lithium Manganese (LMO)

One question springs to mind, what exactly do you want to use this power supply for?

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger github.com/simat/BatteryMonitor
Latronics 4kW Inverter, homemade MPPT controller

[PNjunction]

Been digging through google for weeks and I can't find out if using a 4-6 cell 18650 battery pack is even feasible.

Easy commodity item - I run a few myself. I prefer the ANKER brand of li-ion battery packs, since their 18650 cells are usually Panasonic or LG-Chem or similar high-quality cells. I use a generation back, the Anker "Astro" series, the E5 and E7 right now.

Charge via solar with a 21w folding panel. Also by Anker. There are others. 5v usb output regulator built-in to the panel. Direct attachment to the battery packs.

Why hack together what today is a common commodity item shipped to your door overnight if you want. Easy peasy.

No large panels - no mppt controller, nothing like that. Brainless operation for the most part.

I'm thinking with 400W of solar panels peak time with 80% of rating (peak like 11am-3pm?)I'll be pulling down like 26amps at peak. A 65-75amp hour 4 cell (14.8V average) should be able to get charged from this, but how? Is my best option to switch to a 5s 16V lifepo4 hooked directly to panels?

I run a small 40ah LiFeP04 nominal 12v 4S battery bank quite successfully from a simple 100 watt panel. GBS cells. Similar cells would be CALB. Different chemistry than what is in the Anker packs, which is important because if you don't differentiate the chemistries and their voltages, you'll definitely burn stuff up.

Do some more reading found right here to get the background you need, rather than having us write up a laundry-list of componentry. This way you'll know what you are doing, and be safe not only to yourself, but also to others. Close wallet. Start reading.

Actually, open wallet and pick up an Anker Astro battery pack, and a folding panel to satisfy your li-ion appetite. Close again and start reading before thinking about going big with lithium.