Low Cost Lithium Battery Motorhome Conversion

feedhorn

Lived 11 years now full time in my motorhome using 700 watts of solar panels and 4 golf cart batteries. Over that time I learned to despise lead acid batteries. Late last year I had a single cell failure in my 2 year old set of of batteries and I was fed up! I wanted to get away from lead acid batteries for good. I had good luck converting my motorcycles to Lithium with Shorai batteries and wanted to do the same for my motorhome.

To make a long story short, I dumped 250lbs of lead acid batteries for a 23lb 100ah LiFePO4 battery and from day one life in my motorhome life has been getting better. The last three months I have been going through a steep learning curve with the lithium battery and it is just now that I think things have stabilized and I'm looking forward to maybe another 11 years without any battery problems at all. We'll see.

Choosing my first lithium motorhome battery was a risky decision for me living on a very low income. I could replace the golf cart batteries for about $400, or a bare bones 100ah lithium for $600, or a fully automated 100ah lithium for $1200. Since my lithium motorcycle batteries didn't require a Battery Management System (BMS), I thought I would go with the bare bones low cost lithium and work my way up the learning curve the hard way. In effect I would become the BMS, me, my voltmeter, and a 5 ohm resistor with clip leads on both ends (used to bleed off excess charge for high cells).

After the first week of me being the BMS, I learned that no BMS was not an option for a motorhome installation and constantly checking the battery cells with a voltmeter was too much work. I needed a better BMS solution. I had two choices, individual cell battery balance boards for $8 each or a full blown BMS with all the features for $340. I chose the cheapest route to begin with and ordered the individual cell balance boards.

My first day after installing the balance boards was a disaster, the battery went way out of balance and I didn't know why. I thought the balance boards were behaving badly due to the un-filtered DC from my solar PWM power controller. I removed the balance boards and went back to manual balancing for a while.
I searched the Internet for other cheap BMS solutions and found that the RC model car hobbyist community seemed most up to date on systems for balancing lithium batteries. I ordered a $15 BMS called a 'two meter' to see if that worked and learn a bit about integrated BMS devices. The 'two meter' was a good thing to get as I learned a lot from working with it. In the end I decided it was just too under powered to manage a 100ah size battery. I got out the balance boards again and traced out their circuity and studied how they worked.

Then came a breakthrough! While studying the operation of the balance boards I had one hooked up to my lab power supply and was watching as it turned on at 3.50 volts and turned off below that. I left one turned on for a while as I got a fresh cup of coffee and sat down to play some more. I turned down the voltage and Woah! it stayed on drawing current even below 3.50 volts! This was wrong! What happened ?

As it turned out the balance boards were defective and any of them would go into thermal runaway if left turned on for more than a minute or two. Ouch! This explains how they managed to imbalance my battery so badly. The power transistors on the balance boards needed a heat sink to prevent thermal runaway, so I glued a chunk of aluminum to the back of the boards and this completely cured the problem. Now the whole set of balance boards would perfectly balance the battery automatically during the float stage of charge control. They all turned on at 3.50 volts and off below that.

Whew, the battery balancing problem was finally solved but I still needed more information. I needed to know the State of Charge (SOC) of the lithium battery. My amp-hour meter that I used for lead acid batteries wouldn't work for lithium because it had built-in compensation for the inefficiency of lead acid batteries and was totally inaccurate. The 'two meter' wouldn't measure both positive and negative amp-hours so I went looking for an accurate SOC meter. I found it in what is called a Powerlog 6S a $50 device again developed by the RC hobbyist community.

The Powerlog 6s is a brilliant device that will monitor up to 6 cell voltages and current amp-hours too. It has a USB computer interface and comes with graphic display software that can constantly monitor and record the battery performance. It provides and extensive set of programmable alarm conditions and has an output to drive a battery disconnect relay in case something goes wrong. Now between the Powerlog and the balance boards I have a complete BMS system for very low cost.

After a month of using my automatic BMS system, I noticed that cell one of my battery was reading lower than the others by about 0.020 volt and I thought I was seeing the beginning of a defective cell. I decided to disassemble the battery and rearrange the cells to see what happened. Cell One still read low by the same 20 millivolts. As it turned out there was a ground loop in my battery wiring that was causing the voltage error. Correcting this showed all cells to be equal.

So now after three months on lithium all is well and the system is working perfectly. By noon each day my battery is fully charged. On a cloudy day I can recharge the battery in one hour flat off generator. I'm a happy camper after getting the lead out.

I'll report back later if anything changes.

Cariboocoot

Re: Low Cost Lithium Battery Motorhome Conversion

A good adventure.
For clarity's sake: you did not use the same golf cart batteries for 11 years, did you?
That would be something of a miracle.

feedhorn

Re: Low Cost Lithium Battery Motorhome Conversion

Cariboocoot wrote: »

A good adventure.
For clarity's sake: you did not use the same golf cart batteries for 11 years, did you?
That would be something of a miracle.

No, the my best set of golf carts lasted only 4 years.

Cariboocoot

Re: Low Cost Lithium Battery Motorhome Conversion

feedhorn wrote: »

No, the my best set of golf carts lasted only 4 years.

That would make more sense.

Now, how much did you spend on your lithium system and how much would it cost to do avoiding the mistakes?

(We all know mistakes are expensive. There's no such thing as a cheap mistake, although being cheap often is a mistake.)

feedhorn

Re: Low Cost Lithium Battery Motorhome Conversion

Cariboocoot wrote: »

That would make more sense.

Now, how much did you spend on your lithium system and how much would it cost to do avoiding the mistakes?

(We all know mistakes are expensive. There's no such thing as a cheap mistake, although being cheap often is a mistake.)

No expensive mistakes yet. If the battery fails in less than 4 years, that will be an expensive mistake.

My $15 for the 'two meter' might be considered a mistake but I've rebuilt it into a single cell rechargeable battery capacity tester.

I could duplicate my whole lithium system now for about $700.

feedhorn

Re: Low Cost Lithium Battery Motorhome Conversion

Pictures:

100ah LiFePO4 battery with 300amp fuse and balancer circuit boards.

Attachment not found.

Powerlog 6S Monitor:

Attachment not found.

Computer battery data:

Attachment not found.

mike95490

Re: Low Cost Lithium Battery Motorhome Conversion

Thanks for sharing your lessons and labor with us.

PNjunction

Re: Low Cost Lithium Battery Motorhome Conversion

GREAT write up!

I too have been smitten by the lithium-iron (LiFePO4 chemistry) and NOT the temperamental lithium-cobalt (LiCo02). Despite the higher initial costs, the advantages, at least for me means lead is out of the question for my future projects.

feedhorn wrote: »

Since my lithium motorcycle batteries didn't require a Battery Management System (BMS), I thought I would go with the bare bones low cost lithium and work my way up the learning curve the hard way. In effect I would become the BMS, me, my voltmeter, and a 5 ohm resistor with clip leads on both ends (used to bleed off excess charge for high cells).

Part of the reason for Shorai not mandating the use of a bms-charging system on every charge, and the reason we pay even a bit more for them, is that the cells are closely matched in both capacity and internal resistance. Shorai sells a bms-charger, but it is optional. I have one but I do not use it all the time on my LFX36, and two LFX14's. I'm not even using it in a starting capacity either, but in low-current applications where watt-hours / real ah are in use and not the motorsports "Pbeq" or other manufacturer-invented standards.

Hobbiest RC chargers and even my simple Fluke 87V multimeter rarely ever show any cell-cell difference larger than .015v - even when taken well down into the discharge knee beyond their recommendation, nor at the top when I charge with a voltage of about 14.6 and the absorb current is essentially zero. Either of these is not good on a regular basis, and probably due to factory matching, they are still in balance.

After the first week of me being the BMS, I learned that no BMS was not an option for a motorhome installation and constantly checking the battery cells with a voltmeter was too much work. I needed a better BMS solution. I had two choices, individual cell battery balance boards for $8 each or a full blown BMS with all the features for $340. I chose the cheapest route to begin with and ordered the individual cell balance boards.

In regards to those GBS cells with their great purple tops and 4-screw intercell connections - great. And I'm glad you found a way to modify those balance boards to actually work.

However, once you have actually balanced them once, I'm wondering if in our fractional-C application, that balancing may be overboard unless you are regularly driving them into the upper and lower charge/discharge knees all the time like EV'ers / RC'ers do?

Have you tried running them only from 80% to 10% DOD, where balance should not be a major issue once an initial balance has taken place?

Unlike the Shorai's, even though they may come from the factory charged to the very same voltage, I don't know if GBS does precise capacity and internal resistance matching - unless perhaps you ask first. Ie, you may have a battery made up of a 95ah, 98ah, 100ah, and 105ah, where top-balancing them all the time may be an endless game, especially when IR starts to change a bit too.

I guess what I'm saying is that I'd hate to see your bank destroyed by the failure of a single inexpensive bms balance board, when you might be able to live happily staying out of the extreme knees of operation from 80% to 10% DOD, ie anything from about 12.80v (3.2v cell) at the low end, to about 14.2v (3.55v cell) max at the charge side. If one cell reaches that limit first, time to charge or stop charging. Obviously buying just a little bit more capacity than you need (or conserve what you got) helps a lot too.

As you can tell, I'm on the no-bms side for fractional-C applications, but know that this often turns into religious thread, which I want to avoid.

Either way, great post and thank you VERY much for the detailed info about your setup.

feedhorn

Re: Low Cost Lithium Battery Motorhome Conversion

Thanks but...


What I didn't mention in my write-up is the mind boggling complexity that LifePO4 chemistry presents to the user of a given application. I could have said that LifePO4 is not ready for prime time yet. It seems that way to me. But for now the benefits outweigh the lead-acid nightmare that I have put up with for the past 11 years.

Engine starting is a great application for lead-acid or LifePO4. Put a big 1 or 2 second load on the battery then immediately recharge the battery and float. This works for Shorai without a BMS but the reasons it works are are quite complex.

In my solar motorhome application things get even more complex. I have a 24/7 load of at least 1.5 amps all the time. I have a Xantrex C60 solar charge controller that is not at all optimized for LifePO4 but just barely gets me by. Then I have the darn sun that goes up and down on its own, gets clouds, and changes its length of daylight all the time. Oh, then there is my computer that I run a lot whose current draw is pure noise, randomly bouncing between 3 and 10 amps. The only good thing about this complex application is that I am here to watch it most of the time. I have a Powerlog with an alarm signal and disconnect relay operating continuously to watch the individual cell voltages for problems.

Right now my C60 solar controller is my biggest concern. There are no LiFePO4 spec sheets that tell how they will respond to PWM modulated DC charging. Seems most battery manufacturers expect you to charge your cells individually with a lab grade power supply. The C60 does a couple of things really well. For the first 98% of charge it acts like a closed switch connecting the solar panels directly to the battery. At 98% the buzzing starts and it goes into PWM mode chopping up the DC into little 60hz spikes that hammer the BMS boards, connected loads, and maybe the battery too. This doesn't seem good. What is good is if its got a good load you can adjust the bulk charge and float voltages very precisely. I have the bulk set to 14.10 volts and the float set to 14.05 volts. These precise voltages are very important for the BMS boards to work well. But, float charging a LiFePO4 battery is not recommended. Good thing the sun goes down when it does.

For a BMS boards to work a balance job on a LiFePO4 it takes time and the window of time I have is between PWM start at 98% charge and when the sun goes down. My BMS boards do a 0.5 amp shunt load across cells that are above 3.50 volts. When this shunt balancing starts it takes about 2 hours for a complete balance to be achieved. Of course this whole cycle varies depending on when I decide to vacuum my carpet or watch my big screen TV. To say the least my motorhome conversion to LiFePO4 in an on-going experiment with an un-proven outcome.

And, for whatever its worth, its all been working magnificently for that past few days.

feedhorn

Re: Low Cost Lithium Battery Motorhome Conversion

PNjunction wrote: »

GREAT write up!


However, once you have actually balanced them once, I'm wondering if in our fractional-C application, that balancing may be overboard unless you are regularly driving them into the upper and lower charge/discharge knees all the time like EV'ers / RC'ers do?

Have you tried running them only from 80% to 10% DOD, where balance should not be a major issue once an initial balance has taken place?

I'll try and answer this:

Imagine a LiFePO4 cell is like a conical martini glass with a scale inside that measures 0 to 3.5 volts. The top of the glass is sealed over where there is a small clear tube entering the glass that you will provide the filling current. As you fill the glass with charge the first 2 volts of depth holds very little charge but as you approach the full state each millivolt contains more and more of the charge capacity due to the conical shape. To accurately measure 10% charge is easy, but to measure precisely 80% is much more difficult. When the glass actually gets 100% full the liquid quickly backs up in the small filling tube and the voltage jumps up really fast. Its easy to detect this state of 100% charge when the voltage jumps up fast. So for convenience of detection its easier to top-balance the cell rather any other way.

In RC and EV applications its important to have an accurate fuel guage so bottom-balancing is the preferred yet tedious method. Google bottom-balancing to see how its done. Here all the cells are carefully discharged to their lowest point and then a very precise charge amount is added to all cells equally. This way all the cells will go empty together.

The problem in the fractional-C situation is that small errors creep in on each charge/discharge cycle and leave room for a significant cell imbalance to build up over time. A large imbalance can be dangerous to the battery even if you think you are going from 20 to 80%. Charging to 80% can easily cause the weakest cell to go into an overvoltage condition.

Better safe than sorry. Top-balancing isn't perfect but it is reliable.

PNjunction

Re: Low Cost Lithium Battery Motorhome Conversion

feedhorn wrote: »

What I didn't mention in my write-up is the mind boggling complexity that LifePO4 chemistry presents to the user of a given application. I could have said that LifePO4 is not ready for prime time yet. It seems that way to me. But for now the benefits outweigh the lead-acid nightmare that I have put up with for the past 11 years.

I'll have to disagree about the complexity for our application, which are typically 12v or perhaps 24v nominal setups. We are not the EV crowd, which uses 10's of cells, and hundreds of volts in various series/parallel combinations. We are only dealing with 4 cells in series (4S). Jack Rickard does a good job of explaining bottom balancing, among others so I'm hip to that if I ever run an EV of my own. There is also a lot of junk out there which complicates the issue from dealing with counterfeits, factory rejects, etc. Like you, I'll stick to quality sources.

My contention is that demonstrated by the high-quality Shorai's, (I also have a high quality Braille / 26650s A123 internally) is that if you have a 4S pack going wildly out of balance, then something else is wrong. That means that the cells you received were way out of balance to begin with, have differing real Ah capacities, or very different internal resistance, or have been damaged or received major shocks in shipping and handling. To me, a full-time bms on a simple 4S pack that started out balanced may be indicative of a band-aid approach to loose manufacturing, QC, or possible end-user / edge-case scenarios.

One tip here that I rarely see mentioned is that if you DO find yourself below 80% DOD, that a very low charge current should be applied until you get back to that point before beginning the regular charge. Something like below .1C. The other tip is that if you smell sweet perfume when nobody is wearing any, then stop since the electrolyte is cooking. As usual, having a bit more capacity than you actually need is a good thing. Since I operate an 80-10% DOD ratio, I have to add about 30% to my Ah calculations. What makes that exceedingly simple as you know, is the near absence of the Peukert effect. It's there, but negligible in our fractional-C application.

This works for Shorai without a BMS but the reasons it works are are quite complex.

Nothing more complex than that they take the time and resources to match the cells for capacity and internal resistance before building them into the battery itself. Note that their own bms charger is optional, and I have never seen it recommended for daily use unless one is obsessive. 6 months, or yearly perhaps, or after an "incident".

Note that I also supplement my Shorai BMS-01 charger with a simpler Tecmate / Optimate TM-291 lifepo4 specific charger. Works great and the balance routine, although not cell specific, seems to actually work. You may want to consider that as a backup for both your Shorai's, and perhaps your GBS bank, although I'd probably bulk-charge the GBS bank first before letting the Optimate TM-291 do it's balancing.

Your 24/7 operation sounds interesting and is a nice challenge. I also charge my lifepo4's with Xantrex C12 pwm, and also various Morningstar pwm controllers without issue, although I set the absorb to 14.2 and my float to something ridiculous like 13.2 to effectively take it out of the plan for float - which like you say is unnecessary - unless you have a parasitic load, but I don't. They don't seem to buzz when they actually do go into pwm after bulk, but I have actually heard buzzing at the battery terminal itself, but that was from a very poor connection I made.

Take a look at the lifepo4 specific mppt charge controller offerings from Genasun. I'm itching to upgrade to one myself.

feedhorn

Re: Low Cost Lithium Battery Motorhome Conversion

PNjunction wrote: »

One tip here that I rarely see mentioned is that if you DO find yourself below 80% DOD, that a very low charge current should be applied until you get back to that point before beginning the regular charge.

So on a rainy day, its not such a good idea to wait for the low voltage alarm to go off before starting my 100 amp genny charger and doing a 1C recharge?

LiFePo complexity depends on the application. Shorai has it easy doing just engine starts. If you start doing deep cycling on a Shorai and 1C recharges you are going to find some serious imbalances appear. At 1C charge rate the LiFePo starts getting a bit warm with the inside cells of the pack warming the most. Temperature imbalance yields capacity imbalance.

Personally, I plan to hammer my Lifepo until it dies just to see how long that takes. In a couple of years I expect that LTO and LiS cells will become available and I'll move on to those. My dream is to get my hands on some LiS and run my entire motorhome off 4 D-cells.

feedhorn

Re: Low Cost Lithium Battery Motorhome Conversion

A day in the LiFe:

Attachment not found.


24 hours of data starting about 7am on a clear sunny day.

Fuzzy red (current line) means my computer is on. Square wave is my propane refrigerator cycling while I sleep.
Four colored lines are individual cell voltages.

Mid-day the charge rate drops off as battery gets full (green capacity line), and cell balancing takes over.
One cell balancer reads high because its mis-matched with the others.

Step at top of green line is where I reset the amp-hour counter.

PNjunction

Re: Low Cost Lithium Battery Motorhome Conversion

feedhorn wrote: »

So on a rainy day, its not such a good idea to wait for the low voltage alarm to go off before starting my 100 amp genny charger and doing a 1C recharge?

Depends on where you set the low voltage setpoint. For a typical lifepo4, an 80% DOD would be about 3.2v/cell. Since you are top balancing, your bottom may be ragged, so I'd set the alarm to whenever the first cell reaches that point.

If you do go past that point for some reason, the do NOT apply your 100A charge current to it. Use < 0.1C to get the cells back up to 3.2v, and then you can begin with your normal charge current. Unfortunately, some data sheets only show the absolute highest and lowest values that a cell can operate in without *immediate* damage, but rarely do they point out the optimum voltage range for cyclic life. Usually that just means staying in the flat part of the curve - stay out of the knees. Since your GBS cells are slightly different from standard lifepo4, (lifemnpo4 I believe) then consult their own charts to determine the voltages at the start of the discharge knee, or perhaps just slightly into it, and set your alarm there.

Much like not taking my beloved Odyssey AGM's from 10.75 to 15v all the time. Sure, that may be the limits, but doing so means killing my cycle life. Maybe SOMEDAY we'll get Enersys to come out with LiFepo4 and all the fantastic docs to go with it like they do with their Pb batts. I can only dream for now ...

The drop-dead voltages seen like 2.5 - 3v are another story. I had a friend burn up his Headway pack by thinking THAT was the lowest cell voltage that was ok to go to. Initially his pack survived a few cycles, but along with going too far, he also applied > 0.1C while inside the sharp discharge knee. Not good because he had no idea what 80% DOD was or how to measure it.

This might be a good place to mention that both Shorai and Optimate lithium chargers do the same thing - they drop to very low current if the detected terminal voltages are below 3.2v per cell, and once the battery is restored to that point, then the full output current charge begins. When I charge with other charge sources like solar, I do the same thing If I pull a bonehead maneuver and let the battery drop too far.

If you start doing deep cycling on a Shorai and 1C recharges you are going to find some serious imbalances appear. At 1C charge rate the LiFePo starts getting a bit warm with the inside cells of the pack warming the most. Temperature imbalance yields capacity imbalance.

Not so. I don't even use the Shorai's for starter batteries, but for low-current deep discharge from 80% to 10% DOD multiple times a day. My charge rate is also 2C (Where C is the "real" Ah rating, not the starter PbEq stuff). I hit up an LXF14 with 10A of current with no problem, because the real Ah rating is only 4.5Ah. Most of the newer ones even have stickers on the bottom to show this rating. AND, according to the specs and front-panel sticker, I can apply 14ah of current to it. What that means is that these can take the typical 3C current on charge, although I prefer to go no higher than 2C. If you use the PBeQ rating, then all bets are off!

In this case, 4.5ah (real ah) * 3 = 13.5ah. Using 10ah is no problem - the cells stay cool AND balanced. Of course when I measure top-balance, I do so by fully saturating the battery at 14.6v and let the absorb go to zero-Amps current (not recommended all the time, but I did it to prove balance). No more than 0.015 cell-cell difference on a Fluke 87V (or Hyperion rc stuff) after 1 hour rest for stabilization. I can certainly live with that.

Note that I'm brand-agnostic and not really into motorcycle batteries. They were just the best available cells I could find at my local battery dealer, and in fact the tip about the Optimate lithium charger came from the Antigravity people who also do not use a bms, or provide a port. I'd could have chosen them had they been available at the counter.

These have been invaluable lifepo4 learner batteries before forking out for the big setup. I was worried at first, but smiled when I saw how uneccessary the complexity is when you just maintain them like you are supposed to.

feedhorn

Re: Low Cost Lithium Battery Motorhome Conversion

PNjunction wrote: »

Depends on where you set the low voltage setpoint.

Setting my low voltage alarm at 3.2 volts is not possible. I regularly put large 150 amp loads on the battery when I run the microwave. That load always draws the cell voltages down below 3 volts and that would set off the alarm. I also use the alarm for the battery disconnect relay to protect from any problem that might happen when I am away from the rig.

If you have ever tested individual lifepo cells you know that when they reach 100% charge they open like a switch and no more current can pass through. This makes charging a string of lifepos like a race track for the cells with the lowest capacity cell winning the race to 100% charge. When that cell opens like a switch all the rest of the string get no charge current and their voltages start falling back toward that of resting state. If say a 14.2 volt charger stays on after the first cell goes 100% then the voltage across that cell steadily increases as the other cells fall back. If the cell reaches 4 volts or more bad things start to happen to the electrolyte and the cell gets damaged. This can happen quite fast.

That is where the balance boards come in handy bypassing 1/2 amp of current around the open switch (100% charged) cell and preventing the other cells from falling back before they catch up. If there is any significant capacity mismatch in the cells the 1/2 amp can take quite a while to bring all the other cells up to 100%. Once all the cells get fully charged and the charger is set to the right voltage, the balance boards will trun off and the cell volts will fall back. When they fall back enough the balance boards will turn on again. The cells voltages will dance around up and down until the charger is turned off or the sun goes down.

PNjunction

Re: Low Cost Lithium Battery Motorhome Conversion

feedhorn wrote: »

If you have ever tested individual lifepo cells you know that when they reach 100% charge they open like a switch and no more current can pass through.

Sure enough. We're on the same page believe it or not, but unfortunately it always turns into a bms / no bms issue. It happens, but I'm glad you brought that up.

Because my cells are high quality and matched straight from the factory, and simulating loads like yours, just on a smaller scale, I've never had bad balance when measured well up into the steep charge knee.

While individual cell charging / shunting / balancing is good, there IS another way as demonstrated by the Optimate. It is so simple I don't know why others haven't tried it. Essentially that TM-291 charger will oscillate between 14 and 14.4 volts towards the end of charge. Knowing that in an unbalanced pack one cell will reach full capacity first and cut off current to all the rest, the oscillation allows for the weaker cells to get *some* amount of charge during that very short window of time when the full cell, which has been allowed to settle off the very top, is once again climbing. During that climb, the weaker cells get some charge. It does this pretty quickly and for 5 minutes at a time (more if it detects that the battery needs it) and tests and repeats if necessary.

Case in point - I purposely paralleled two LFX-14's from different manufacturing runs together and ran them that way for 2 weeks. I pulled them apart and lo and behold, one of the pair had a 0.270 mv mismatch between the highest and lowest cells. I put it on the Shorai balance charger and it eventually faulted and stopped - maybe being over 0.2mv was too much. I put it on the Optimate, and balance was restored - without using a cell-level charger. I could have used my Hyperion EOS pack-sentry shunt, but wanted to prove the optimate.

Obviously there is only so much you can do with balance on a charger designed for 4S batteries, especially if they have something gigantic like 24 cells inside, but for simple batteries like the Shorai - hey it worked. It worked so well that if I was still running my AGM's, I'd be interested in the lead-based Optimate 6 which will do much the same thing (different voltages and whatnot however).

Thing is, I've got ways to do a balance with my Hyperion chargers, shunt loads, etc, but find that with quality lifepo4 like Shorai or Antigravity, there is no need unless you go nuts.

But I'm really not trying to convince you to change a working setup! We're both on the same page really when it comes to enjoying the advantages of lifepo4.

feedhorn

Re: Low Cost Lithium Battery Motorhome Conversion

PNjunction wrote: »

Sure enough. We're on the same page

When I said the reason Shorai doesn't need a BMS is 'complicated' you are getting into this complexity in your last post.

When you describe how the TM-291 balances without a BMS you are describing what I call "Lifepo's ability to self-balance" I have witnessed this in action and hoped I could take advantage of this effect to get by without a BMS. After much testing with the big GBS battery, I concluded that it would be unreliable to go without a BMS. There are subtle differences between small and large cells that make it that way. This probably explains why the GBS cell warranty is void if you don't use an approved BMS.

I have a pair of lifepo cells in one of my power tools and one of the series cells has just a little more self discharge than the other. They didn't use a BMS and I found the only solution was to disassemble the pack and charge the cells individually using my lab power supply. If I leave that pack in the charger 24/7 the cells will eventually self-balance but as soon as you take it out of the charger the imbalance starts up again. After a couple of months on the shelf the pack is unusable.

electrodacus

Re: Low Cost Lithium Battery Motorhome Conversion

You needed exacly what I'm building right now see this video
http://forum.solar-electric.com/showthread.php?22911-My-Programable-Solar-BMS-for-any-Lithium-cells-and-Supercapacitors
There is one more detailed video on my youtube chanell and I just made a post here.
Let me know what you think since you already got the experience.

PNjunction

Re: Low Cost Lithium Battery Motorhome Conversion

Welcome!

Looks like a fun homebrew unit. Unfortunately, you've posted the same thing basically in multiple threads and now it will be hard to track. Perhaps start your own thread to keep it concentrated, especially since you are seeking free R&D. Even with free R&D, you also have marketing issues and funding, so maybe we're getting into a grey area here.

Ah, looks like this is the best place for it:
http://forum.solar-electric.com/showthread.php?22911-My-Programable-Solar-BMS-for-any-Lithium-cells-and-Supercapacitors

PNjunction

Re: Low Cost Lithium Battery Motorhome Conversion

feedhorn wrote: »

When you describe how the TM-291 balances without a BMS you are describing what I call "Lifepo's ability to self-balance" I have witnessed this in action and hoped I could take advantage of this effect to get by without a BMS. After much testing with the big GBS battery, I concluded that it would be unreliable to go without a BMS.

If this were a normal CV charger, then there would be no self-balance of significance. The Optimate rapidly swings between 14.0 and 14.4 which helps bring up the low cells during those oscillating windows. Same effect with their latest Pb charger, although I've observed that it takes a few cycles for them to come really close.

With the GBS bank, how far out were your worst cell to cell differences? Are we talking more than 50 or 100mv, or something much larger?

Were you doing top-balancing, thus leaving the bottoms ragged (or stopping when the first cell hits lvc), or bottom balancing, leaving the charger to quit after one cell reaches the upper target voltage? You can't have both, so you pick what's appropriate, ie EV users do bottom balancing for the most part.

feedhorn

Re: Low Cost Lithium Battery Motorhome Conversion

PNjunction wrote: »

With the GBS bank, how far out were your worst cell to cell differences? Are we talking more than 50 or 100mv, or something much larger?

When I first got my GBS, I charged each cell individually with my lab power supply. Then floated all cells at about 14.4 over night. The battery balanced evenly. I then did some deep discharge tests, 1C recharges, and solar charging tests. My big 100 amp charger puts out half wave pulses at 60 Hz. The PWM charge controller puts our varying pulse widths. The battery performed quite well with no significant imbalance detected. After that I started playing with BMS balancing just to understand what that whole business was about.

I bought the set of balance boards and installed them and practiced charging the battery this way to see what difference it made. After sitting overnight the next morning I found the battery badly imbalanced. 100's of millivolts. It seemed something had gone very wrong. At first I thought the pulsating chargers had damaged the balance boards and so I removed them. In recharging the battery with my lab power supply at 14.4 volts I had one cell run madly up to about 4.0 volts and that really scared me. I then recharged each cell individually and went back to square one scratching my head as to what had actually happened.

After much more careful studying of Internet literature and testing I was lucky to accidentally find that thermal runaway in the balance boards was the root cause of the balance board failure. In the meantime I realized how dangerous it is to charge a lifepo string if some of the cells are way out of balance. You need to set your charger cutoff voltage as low as possible to prevent accidental overcharging of a single cell.

I have since concluded that pulsating and PWM chargers may actually be good for lifepo batteries as they seem to promote self balancing. I'm looking forward to some new solar charge controllers to become available that take on lifepo balancing very seriously.

feedhorn

Re: Low Cost Lithium Battery Motorhome Conversion

electrodacus wrote: »

You needed exacly what I'm building right now

Actually Not, your design won't work for me.

I now realize what the proper design for a lifepo solar charger should be.

But, at this moment I can say no more.

Good Luck

CDN_VT

Re: Low Cost Lithium Battery Motorhome Conversion

Great Read Gents !
Please keep all of us reading and learning .It might be nice to also include links to an article , so we can read up and learn from that also. I had to gogle & read to find the correct article on Jack Rickard ... after many try's.

VT

PNjunction

Re: Low Cost Lithium Battery Motorhome Conversion

feedhorn wrote: »

My big 100 amp charger puts out half wave pulses at 60 Hz. The PWM charge controller puts our varying pulse widths. The battery performed quite well with no significant imbalance detected. After that I started playing with BMS balancing just to understand what that whole business was about.

Caution! I wonder how much ripple that half-wave circuit in the charger has? I forgot where I read it now, but I saw a reference that lifepo4's should not be hit with anything larger than 3% ripple. I also have a large half-wave charger, and stopped using it until I can gather more facts about an acceptable ripple current. Wish I could remember but I think it was a lifepo4 distributor in Australia maybe that spec'ed this. I haven't seen it anywhere else but I should have bookmarked it. Until then, I'm sticking to the pwm cc's and higher-quality chargers.

If you find a ripple spec, I'd love to know. Wonder if your bms boards might have had a problem with the half-wave ripple?

100's of millivolts. It seemed something had gone very wrong.

Ok, that's a problem. I just wanted to make sure you weren't trying to reach zero volts diff. You know that for the first 10 cycles or so, the balance will wander around a bit. Thus if cell1 and cell3 are today's major players, tomorrow it could be cell2 and 4. Eventually a pattern will develop and stabilize - of course until aging sets in.

You need to set your charger cutoff voltage as low as possible to prevent accidental overcharging of a single cell.

I can understand that. Still for me, I'm just going to use a bit larger capacity to provide me some autonomy / unexpected loads, and loaf in the flat-discharge curve. I may never even reach 80%, but plan on running perhaps 70% - 10% DOD. I figure I'll pay a little more initially for the capacity, but I'll have removed a failure point of an external bms and get some headroom for out of balance. I guess to each his own.

feedhorn

Re: Low Cost Lithium Battery Motorhome Conversion

PNjunction wrote: »

I wonder how much ripple that half-wave circuit in the charger has?

From what I can tell the charger outputs 100% ripple, there is no filtering just like most ordinary car chargers. It would take huge filter capacitors smooth out a 100 amp 12 volt charger. I don't know if my PWM CC is any better, there is no filtering in that thing either. Those chargers expect the battery to be the filter capacitor and for lead acid that works pretty good. As far as I can find nobody has done any research on lifepo and ripple currents. Fortunately, I rarely run my big genny charger for the sun here in Arizona is quite reliable.

I chose to go with a minimum capacity lifepo so I can get a real good idea of the abuse they can take. I cycle it 60 to 80 percent each day. It looks like when capacity sizing lifepo all you need is enough capacity to get you through the night.

I read an account of how lifepo batteries are used in LEO satellites and they have been working very well. DOD is about 80% every 88 minutes and then quickly recharged for the next orbit.
This equates to over 100,000 deep cycles in the expected 20 year lifespan of the satellite. The ISS also does massive deep cycling on its Nickle Iron batteries but I hear they may be headed toward lithium in the near future.

My lifepo is balancing perfectly in about 2 hours each day. I wish I could extend the bulk charge hold time so I don't have to go into float mode to complete daily balancing.

feedhorn

Re: Low Cost Lithium Battery Motorhome Conversion

Six Month Report:

All is well with my Lithium conversion everything is working perfectly.

These long days show my DoD of only 25 Ah over the short night time.

Battery is usually fully charged by 9 AM.

With only 100 Ah capacity it more than covers what the 4 golf cart batteries did.

I would expect that I would need 1000Ah of lead acid to give the same charge/discharge performance as these lithium do.

With no maintenance on the lithium battery, its easy to forget you even have a battery just perpetual free electricity always at the proper voltage.

I would definitely do the conversion again and am now sure that I don't need any additional capacity.

FH

PNjunction

Re: Low Cost Lithium Battery Motorhome Conversion

Thanks for the report!

I put my money where my mouth is by joining you with GBS cells, although I'd have no problem with CALB or Winston. The tops that come with the GBS are convenient. I got a lot of practice with smaller A123 cells, which to me, are really more intended for SLI duties, although you can press them into low-current service at quite an expense.

One of the first things I had to get over was that after an initial balance at the top, there really is no need to bring them up to a high SOC every time - in fact it is better not to do so if you have some capacity to spare. Since it appears you do, you may want to look at lowering your overall voltage to maybe no more than 14 - 14.1v or so. You'll still get nearly a full charge, and absorb will take just a smidgen longer.

Lifepo4's have a full charge at 3.45v per cell, and it takes about .05v higher than that to get there, and when current is about C/20, absorb is finished. If you are charging at less than C/20 to begin with, just stop when the cell voltage hits 3.45 or so. Taking the cells higher in voltage than this is merely heating the electrolyte unnecessarily.

When my new cells arrive, I'm thinking about using a very light coating of Penetrox or Noalox under the cell bridges, and a light cleaning of the terminals, as we are dealing with dissimilar metals.

feedhorn

Re: Low Cost Lithium Battery Motorhome Conversion

PNjunction wrote: »

Thanks for the report!


One of the first things I had to get over was that after an initial balance at the top, there really is no need to bring them up to a high SOC every time - in fact it is better not to do so if you have some capacity to spare. Since it appears you do, you may want to look at lowering your overall voltage to maybe no more than 14 - 14.1v or so. You'll still get nearly a full charge, and absorb will take just a smidgen longer.

Lifepo4's have a full charge at 3.45v per cell, and it takes about .05v higher than that to get there, and when current is about C/20, absorb is finished. If you are charging at less than C/20 to begin with, just stop when the cell voltage hits 3.45 or so. Taking the cells higher in voltage than this is merely heating the electrolyte unnecessarily.

I have my C60 solar controller Bulk set at 14.10v and Float set at 14.05v. I have individual balancer boards on each cell that regulate their voltage to 3.50 all day long. Indivdual cell voltages wander around a bit during the day as I turn loads on and off but as far as I can tell everything is working well.

DoubleDD

Re: Low Cost Lithium Battery Motorhome Conversion

feedhorn wrote: »

I have my C60 solar controller Bulk set at 14.10v and Float set at 14.05v. I have individual balancer boards on each cell that regulate their voltage to 3.50 all day long. Indivdual cell voltages wander around a bit during the day as I turn loads on and off but as far as I can tell everything is working well.

Great Thread

If you had to do it all over again, would you have gone with a 12 volt Lithium Battery with built-in BMS. I've been looking at http://www.lithiumion-batteries.com/products/12v-50ah-lithium-ion-battery/ or http://store.starkpower.com/12V-50AhStarkPower-UltraEnergy-Lithium-Ion-Battery-LiFePO-Energy-Storage-Battery_p_42.html. I have a single 250 watt solar panel which I will be adding another panel in the spring with a 40A MPPT solar charger. I don't do much boon-docking and when I do its for 3 days at max. I keep my RV at a storage lot with no power when not in use. The slow discharge rate of Lithium should be offset by my solar panels.

In the past 5 years I replaced the original 12v batteries with 2 T-105 batteries that don't seem to last long anymore. I'm sure it has to do with the bad converter/charger (WFCO 55AMP) that came with the RV. I've read they don't fully charge batteries and I have overly discharged them a couple of times. One of the batteries is also bulging, so I'm sure it is damaged. Not wanting to see myself at the same point in 4 years, I'm considering Lithium Ion Batteries.

Going with a Lithium battery with BMS will ensure I don't overcharge/discharge. The solar panels will charge Lithium batteries much faster as they take a charge more efficiently.

What do you think about starting out with a 50AH battery and add more 50AH batteries as I'm able to afford them. People have been recommending AGM (cost) as a good replacement for the T-105.

DoubleDD

Cariboocoot

Re: Low Cost Lithium Battery Motorhome Conversion

Welcome to the forum DoubleDD.

You are correct that your converter/charger does not charge the T105's properly. Unfortunately it won't charge any replacements properly either. These things are just big 12 VDC power supplies; they don't do the Bulk/Absorb/Float stages that deep cycles need to stay healthy.

Therefore replacing the T105's with AGM's will just be throwing good money after bad. As a rule AGM's are even more expensive per Watt hour capacity than flooded cells. I don't know why people would think they are a good replacement. Unless you actually need the AGM advantages (higher current handling, sealed container, lower internal resistance) there's no point in buying them.

As for the LiFePo chemistry, check out the numerous threads about using them in the Advanced section before you take that plunge. There are different types of "lithium" batteries and they aren't all equal. A good choice for RV applications where the weight savings can be a big plus, but if you managed with lead-acid before perhaps the less expensive solution is solving the improper charging - and buying batteries that don't cost as much as nor demand the high Absorb Voltage of Trojans.

DoubleDD

Re: Low Cost Lithium Battery Motorhome Conversion

Cariboocoot wrote: »

Welcome to the forum DoubleDD.

You are correct that your converter/charger does not charge the T105's properly. Unfortunately it won't charge any replacements properly either. These things are just big 12 VDC power supplies; they don't do the Bulk/Absorb/Float stages that deep cycles need to stay healthy.

Therefore replacing the T105's with AGM's will just be throwing good money after bad. As a rule AGM's are even more expensive per Watt hour capacity than flooded cells. I don't know why people would think they are a good replacement. Unless you actually need the AGM advantages (higher current handling, sealed container, lower internal resistance) there's no point in buying them.

As for the LiFePo chemistry, check out the numerous threads about using them in the Advanced section before you take that plunge. There are different types of "lithium" batteries and they aren't all equal. A good choice for RV applications where the weight savings can be a big plus, but if you managed with lead-acid before perhaps the less expensive solution is solving the improper charging - and buying batteries that don't cost as much as nor demand the high Absorb Voltage of Trojans.

Yes I agree that the charger is useless and am going to be changing that out too. Just wondering if anyone sees a problem with using a 50AH battery to start and adding another next year. I see many benefits going with Lithium. The biggest drawback is cost but I try and set aside $1500 per year for maintenance and upgrades.

DoubleDD