The Battery Debate (AGM, LA, LIFEPO4) Post 2


Disclaimer – Please understand that I’m trying to simplify this as much as possible for a wide audience to understand the basics. I’m using oversimplified examples and voltages. I have purposely left out complex charging routines and such. This post is not intended to be a detailed technical reference. If I do build the system, I will post a detailed technical post with all those values.


CAUTION…. Serious technical talk ahead. This post will be much more technical and LONGER than the last one. Again I know this won’t be for everyone so may I suggest a trip over to for anyone not interested. Just be warned, not all the jokes are tasteful and careful of the ones marked NSFW!


Still with me? Good, first I would like to touch on a battery that I left out of the last post since multiple people contacted me about it. Firefly carbon foam batteries are a relatively new offering (about 5 years) that have some promise. Although I left these out of the original post, I did consider these batteries and even spoke with multiple experts about them (I can’t remember %100, but I think I actually spoke with the inventor). Although they are great product and probably have a place in the market, I rejected them for a few issues.

1. Weight. They essentially weigh the same as any quality AGM battery (heavier compared to some).

2. Supply. The company has not been able to meet demand EVER. Delivery times are usually months out. Not only is this a problem for the original purchase, but warranty.

3. Still AGM. These batteries are still essentially an AGM technology. As such they still suffer from some of the AGM downfalls, albeit to a much lesser degree (voltage sag, heat sensitive)

4. Price. Price is higher than a good DYI LifePo4 system. Personally I think they are way overpriced for what they are.


To sum up. LifePO4 batteries beat the Firefly batteries in every category. If I am going to spend the money, why not spend it on the technology that ticks the most boxes.


Alright now back to the Lithium battery debate. Remember when I said I would need to design and build a Battery Management System (BMS). Yes… Well surely you have been thinking about this and wondering “Travis, there must be companies that have already done this and offer a package”.  You of course, would be absolutely right. There are at least 3 companies (Victron, Relion and Lithionics) that offer great packages. However, those packages just aren’t in our price range. To put it in perspective a 300AH battery from Relion costs $3500USD and a DIY 400AH system is “around” $2600USD. Not only is it much cheaper to DIY, but we get a bigger bank.  Frankly, the only way I can swing lithium batteries is DIY.


Ok, so now we know DIY is the only option for lithium on Party of Five, back to the BMS design.  Before we get into the meat of the design we need to do some background on batteries. First, there really is no such thing as a 12 volt battery (at least in the technologies we are talking about). That 12V battery under the hood of your car is really 6 individual 2.1V batteries hooked in series (called a 6S pack). The voltage of those batteries is combined to achieve a nominal battery of 12.6V. The 2.1V number is a limitation of the actual chemistry of the battery. 2.1V happens to be the best voltage that the lead acid design can achieve.  Now LifePo4 being a completely different chemistry (lithium iron phosphate), has a different optimal voltage which happens to be 3.2V. So to get a “12V” battery that would be compatible you need to hook 4 cells in series (adding their voltages). This “4S” pack would be 12.8V and close enough to make anything in a automobile, RV or boat happy.  So with both of these batteries we really have multiple little batteries hooked together and we treat them as a single battery. However, just like a chain, the battery will only be as good as its weakest link. If any one of those individual cells gets out of sync bad things can happen (technically called unbalanced). Let say you draw the battery to %50 charged, but the cells are out of sync and one cell is at a higher voltage than the others. In a lead acid battery that would mean the “pack” voltage would be 12.2v (a %50 discharged lead acid battery voltage is 12.2V). However at the individual cell level it might look like this:

Cell 1 – 2.02V                                Cell 4 – 2.02V

Cell 2 – 2.02V                                Cell 5 – 2.03V

Cell 3 – 2.035V                              Cell 6 – 2.02V

The problem with this scenario is that we have now taken 4 of the 6 cell below the %50 number (potentially damaging them). The other issue, is when we recharge the pack back to %100, 2 of the cells will be pushed over the charging limit. Remember our chargers treat it as one 12V battery, they don’t see the individual cells. Make sense? This is the one place where lead acid technology has a HUGE advantage. The nature of the lead acid design means they can bleed large amounts of energy (heavy lead plates, water to absorb heat, ect). So when the pack is recharged to %100 those 2 out of balance cells will just bleed that extra energy until the other 4 cells “catch up”. Of course we still have the issue that the 4 cells were drawn too low, but that’s not really a problem if it only happens occasionally.  So lead acid batteries come back to balance by themselves if you charge to %100 often enough. Now if we tried this same scenario with a LifePo4 pack here is what it would look like:

Cell 1 – 3.13V

Cell 2 – 3.13V

Cell 3 – 3.14V

Cell 4 – 3.17

The first thing to notice is how little voltage change there actually is. LifePo4 has a very linear voltage curve. This means they don’t loose much voltage as they discharge. In fact they loose such little voltage that using voltage to determine the “Depth of Discharge” (DOD) is not reliable. However with the voltages being so close you can see how easily a single cell can get out balance. Unlike lead acid batteries, this cell will not simply bleed the extra energy. With every %100 recharge it will continue to go more and more out of balance, eventually getting to a voltage that damages that cell, ruining the entire pack. ENTER the BMS!


Alright… I going to take this paragraph to address the argument I KNOW will come up after this post! Do you really need a BMS? There is NO argument, YES, you always need a BMS. Whether that BMS is an automated electronic contraption or a person that monitors the battery and turns shit off manually, balances the cells manually and monitors temperature its required. The real question is “Do I need an automated BMS solution”? My answer will still be “Yes”. Unfortunately humans are notoriously unreliable creatures and relying on one (even myself) to protect my investment seems foolish. Hell, most cruisers can’t remember when to add water to their batteries. Trying to monitor, maintain, manage and protect a LifePo4 installation without automation seems silly. Hell a single event where the BMS kicks in and saves potential damage to the battery pack pays for the BMS. However, I personally know cruisers that have LifePo4 and no BMS. It works for them, but its not worth the risk to me. BMS is cheap insurance.


A good BMS will not only mitigate the above out of balance scenario, but offer other features and protections. However its most important job is to protect each cell in the pack. If anything happens that puts a cell or the whole pack in danger the BMS will disconnect EVERYTHING from the batteries before they are damaged. It will then leave  the pack isolated until a human can figure out what went wrong. The most common failure mode of LifePo4 cells is under-voltage or over-voltage. Unlike other battery technologies, they can not handle being discharged below a certain voltage or charged above a certain voltage. Usually a single event that goes over the upper limit or under the lower limit permanently damages the cells. With that in mind here is what a basic BMS should look like.

– A cutoff device (relay, solid state relay, contactor, ect). This device is installed between the pack and ALL loads and charging sources

– A device capable of measuring the entire pack voltage and sending a signal to the cutoff device in the event of over or under voltage (at the pack level)

– A device capable of measure the voltage and temperature of individual cells. The device should be capable of bleeding extra energy if a cell becomes out of balance and will be over charged. It also must be capable of sending a signal to the cutoff device in the event of an “out of bounds” situation (i.e. it can’t bleed enough energy and a cell will go over critical voltage).


2 years ago when I originally looked at these batteries this is where I stopped. I literally began drawing logic trees, pulling out breadboards and wiring things up. That’s when I realized that although I could build such a system, there was no way I could build, test and debug it in the timeframe I had. Although I now have the time, I think the slow cruising lifestyle has made me lazy. I looked back over my old notes and thought “This seems stupid, there must be an easier way”. That’s when I had Daphanie fetch me a beer and wielded my GoogleFoo to see if someone had done the work for me. Of course things had changed massively in 2 years.  Companies had come and gone, some had stopped selling to DIYers and others had built some solid refined products.


My first target was the disconnect device. My old notes had a bunch of points about this device. People were mostly using large automotive type relays that were energized while the pack in a “good” state. The BMS would cut power to the relay in the event of a “bad” situation. Of course these relays were using quite a bit of power which was bad. Even worse there had been some cases where the relays stuck closed and didn’t disconnect the pack. Of course the big players had also seen these issues in the last 2 years and came up with solutions. It wasn’t long before I found the Victron Battery Protect device.


The battery protect looked to solve at least 2 of the BMS issues. One, its designed to serve as a high amperage disconnect device (100A continuous and 600A surge). Being a solid state device it draws very little power in any state and It can be controlled just like a standard relay. Second, its intelligent and constantly monitors the pack voltage. If the pack voltage goes above a certain voltage, or below a certain voltage it will perform a disconnect all by itself. The best part, a local shop in Le Marin actually had one on the shelf ($95 euro though, my butt hurt a little bit)


With those 2 functions solved I went digging for the more complicated part of the BMS. This is where I was initially stonewalled as it appeared many of the companies had either closed up shop or been absorbed. It took a quite a bit of GoogleFoo before I discovered a small Australian company building and selling the exact product I was looking for. They had designed a simple single board solution that bolted right to the top of each cell. Not only will they dissipate the extra energy for balancing (up to 2A), but they are network together and can send a signal to the Victron Battery Protect in the event of an error on any cell.  Unfortunately finding stock here in the islands would be impossible, but I could buy them online for only $22cdn/cell (yup, under $100 plus shipping for a 4S pack). However since shipping things down here can be an issue, I decided the best bet would be to ship them to Canada and get them here “somehow”.


WAIT… I can hear you saying “Aha, you bought the items so you must have decided to order the batteries”. However that would be a “No”. I decided to buy the items so I can put them on my test bench and ensure they function like I expected.  Since the BMS is key to the entire operation I felt it was prudent to ensure I have not overlooked anything. As this post is now long enough, I’m going to stop here. In my next post I will discuss more specific details of the entire system. Charging sources, voltages and modifications to the boat. Stay tuned. For now here are a couple of images of what the pack should look like with the BMS.


The pack in the picture is a different brand of cells that happen to be blue. The cells I sourced are yellow.

Pack  BMS

  4 comments for “The Battery Debate (AGM, LA, LIFEPO4) Post 2

Leave a Reply to travis Cancel reply

Your email address will not be published. Required fields are marked *