270ah DIY LiFePO4 build

[Poey50]

So when do we expect a 24V circa200Ah LiPO4 system to cost circa 600euro then?

8 x 3.2v 200ah cells are probably around that price point already. But total systems cost depends on many factors. My total cost including all new charging systems (including Balmar alternator, external regulator and serpentine belt upgrade) was £3K. If I had more solar available I wouldn't have gone for an alternator upgrade which would have saved around £1400.

But you need to compare like with like - 200ah LFP is roughly equivalent to 460ah of lead acid. Multiply that by a factor for additional longevity and the financial argument is strong.

 

[JohnGC]

Not off-topic at all. Further back I mentioned some hybrid systems - detailed by Nordkyn. But an intriguingly simple method is carefully explained by ZwerkfCat. I've not heard of anyone using this method even though, as they explain it, it seems safe and obvious.

Lithium-Hybrid

A minor addition to what he suggests is desirable; namely the ability to manually disconnect the LiFePO4 isolation relay.

Without that ability, the lead acid battery can't be returned to full charge unless the LiFePO4 battery is first returned to full charge.

In a situation where the hybrid battery has been run low and has significantly discharged the lead acid.
Followed by the inability (or unwillingness) to return the LiFePO4 to 100% SOC.
The lead acid will remain at a poor SOC.
If left in this state for some time, the lead acid life will be reduced.

If the "deep discharge" event occurs shortly before the boat is to be left for some time, then you are faced with a choice;
1. Fully recharge and leave the LiFePO4 battery at or close to 100% SOC so reducing its life.
2. Partially recharge to leave the LiFEPO4 battery at less than 80% SOC but leave the lead acid at a low SOC which may reduce its life.
3. As 1, but delay departure from the boat until the LiFePO4 SOC can be reduced.

Also, constantly cycling the LiFePO4 between 80% and the low SOC disconnect point will gradually run down the lead acid.

To mitigate this, a small modification to the hybrid system which;
- allows the LiFePO4 to be manually disconnected.
- and the lead acid to be charged with the correct profile.
Would allow the owner to partially recharge the LiFePO4, then disconnect it to fully charge the lead acid.

(Unless there is a BMS that will automat this.)

 

[JohnGC]

Courier just delivered my 8 270AH Cells from China along with a metal case they sourced for me. Took just over 1 month to arrive in the UK.

View attachment 102063

Unboxing coming soon.... ?

Currently Xuba and RJ are quoting shipping by train. I wonder if yours were shipped by train or by sea?
Who delivered them to your door? Was it the Royal Mail or branded courier?

I'm curious to know how they would handle redelivery if I wasn't at home?

 

[Poey50]

A minor addition to what he suggests is desirable; namely the ability to manually disconnect the LiFePO4 isolation relay.
etc.

It's a good point.

 

[nfluester]

Currently Xuba and RJ are quoting shipping by train. I wonder if yours were shipped by train or by sea?
Who delivered them to your door? Was it the Royal Mail or branded courier?

I'm curious to know how they would handle redelivery if I wasn't at home?

final delivery was made by DPD. when i placed the original order with Xuba they gave me a DPD tracking number which was status "shipment arranged" for about 3 weeks till last friday when it updated to arived in Germany then they hit london on Monday and the DPD driver

 

[JohnGC]

final delivery was made by DPD. when i placed the original order with Xuba they gave me a DPD tracking number which was status "shipment arranged" for about 3 weeks till last friday when it updated to arived in Germany then they hit london on Monday and the DPD driver

Thanks that's good to know. If it's a mainstream courier like DPD then a redelivery isn't likely to be an issue.

 

[sailaboutvic]

Here a question and remember you couldn't even call me a novice when it comes to lifop4
Why is it that they can't be charge with a normal say 50a alternator and solar panel ?
And why to you need a b2b charge ?

 

[nfluester]

i'll have a go, it is possible to charge LifePo4 batteries from the alternator however part of the great thing about Lifepo4 chemistry is that it has extremely low internal resistance so will suck up every amp of power the alternator can deliver making them run very hot.

there are a coupe of great videos showing this one from Victron:

and the other from Sterling

which explain it better than i can

also LifePo4 unlike lead acid does not like to be at 100% charge for long periods of time so doesn't like a long / continuous "float" charge so once full you need to be able to disconnect the alternator charge which isn't very straightforward, and if u just disconnect the output from the alternator you are likely to damage it

you can use an external regulator device on you alternator such as the Balmar, wakespeed or other that will vary the alternator output based on a temperature sensor to protect your alternator but these external regulators tend to require modification of a standard alternator soldering a "field wire" on to the internal bushes

the other and easier is a B2B or sometimes called DC 2 DC charger this simply connects to the positive and negative of an existing battery and charges the lithium based on a specific charge program and act as a buffer to protect your alternator from overheating or from disconnects very easy to wire in and give a charging program designed for LifePo4

hope this helps

 

[vas]

Not off-topic at all. Further back I mentioned some hybrid systems - detailed by Nordkyn. But an intriguingly simple method is carefully explained by ZwerkfCat. I've not heard of anyone using this method even though, as they explain it, it seems safe and obvious.

Lithium-Hybrid

evening,

just now had the time to go through ZwerkfCat's work, sound brilliant, would work nicely with my setup with the Victron Multiplus and Victron MPPT for solar and the slightly tired Trojans.
Only catch is that he's posted a header file and general descriptions in github almost a year ago of an arduino based system (that I'm really happy to work and play with) but hasn't actually posted any code or diagrams of a working prototype.
Plan to contact him and ask and offer to help in integrating with N2K (although not much you can display there).
Not in a great hurry so see how it pans out.
Honestly a BMS doesn't look like a v.advanced piece of kit in need of expensive el. components (unless I'm missing something as I haven't examined one closely...)
Also had a look around Nordkyn site, same thing, lots of talking nothing to actually see and evaluate (hardware or code)

Wonder if there is another opensource project that actually has code for that.
Would be interested to port and run it on teensy 3.5 3.6 boards as I have custom boards built with powersupply, i/o, N2K, serial, I2C, ic relay control, etc.

any pointers welcomed, getting in lockdown mode on Sat down here by the looks of it, so need some new project

cheers

V.

 

[sailaboutvic]

My question .
Why didn't you use one a complete build BMS rather then the 123smart , was it just the cost ?
Wouldn't it be neat and less board projecting that can be damage , plus there no messing around with solider

 

[Poey50]

My question .
Why didn't you use one a complete build BMS rather then the 123smart , was it just the cost ?
Wouldn't it be neat and less board projecting that can be damage , plus there no messing around with solider

Five main reasons.

1. Clarity of information. Most suppliers of electronic / electrical equipment couldn't write a manual if their lives depended on it. I could understand the 123Smart manual which is also backed up by instructional videos.

2. Support. All my queries have been answered fully within 24 hours.

3. Remote monitoring and configuration via Bluetooth. The app and displays are excellent. I had only one dip switch to move for the setup, everything else is configured by the app.

4. Cost. This is a moderately priced BMS. REC Active and Orion Junior (which I also looked at) are, like-for-like, much more costly.

5. It is easily matched to their bistable relays. These have very low power demands. They are rated up to 125 amps - enough for my use.

You do of course have to physically protect the boards but that isn't difficult to do. You do need some kind of case for the cells in any event. People sometimes complain about having to solder on connecting wires - seems like a basic skill to me. More concerning have been some bugs in the two first versions but this 3rd Generation version has been trouble-free so far.

However if you need a BMS to directly control chargers, provide warning alarms in addition to bog standard catastrophic level disconnect (which mine is set up for) then you do need a more sophisticated unit. See #96 for a discussion of the issues.

 

[sailaboutvic]

And another question .
How do you top balances the batteries , say they arrive with different voltage, do the 123bms do that for you ? Or do you need to charge the battery separately?

 

[vas]

Vic, from my reading on the topic and what previous posters mention, you do charge them individually to same voltage, then get them all together and the bms takes over. Considering bms balancing deals with (what seems to be the norm of 1A loads shifting about), it would take ages to balance on the bms alone, so you more or less assist the system by balancing manually ahead.
The more I study the more I think that a hybrid system is a more failsafe option (both for the boat and the LifePO4 bank!)

to the experts and from my relatively limited studying on the topic:

hybrid seems to be reasonably well defined as a concept, but no implementation as yet.
The two links pointed in the last dozen posts are basically talk, no h/w, no data.
It may sound naive, but seems to me that tracking with a conservatively tuned bms (say 123bms) voltages (and via Victron BMV and shunt amperage) one could with two MASSIVE contactors (sp? relays!) get the LifePO4 bank online and offline and be done with. Any PLC/arduino/raspberry pi could run the logic, it's easy especially if you integrate that in Venus OS (ordered the kit to built a raspberry pi hopefully next week)
Keeping your existing FLA bank as the buffer and running your loads mainly off the LifePO4, seems to be the best of both worlds.
You don't gamble killing your alternator as bat load is never going to be offline, there's always going to be the FLA to dump to
LifePO4 takes over on consumption as and when needed.
All sounds too good to be true, and simple!

What I've not grasped fully yet (and could be a show stopper) is the linking between FLA and LifePO4 in charging mode. The docs imply that you just wire them together and "it'll be fine" which I'm not sure its enough. Should there be some special B2B charger in between?
In perspective, I'm interested in the hybrid system as I have two main sources charging my batteries (600W solar in the Med through a Victron MPPT100/30, 60A@24V alternator with a Sterling advanced regulator) When the going gets tough, I may fire up the 8kVA diesel generator which then charges via (and that's the main issue) a Victron Multiplus 3000/24/60 charger inverter.
If I am to reconfigure everything on a LifePO4 only method, Multiplus is a no go (cannot isolate charging to consumption!), and it's not going away for sure!
Trojan T105RE X4 225Ah@24V currently in okayish state, they manage the consumptions on my 43ft mobo fine.
Could add a 200Ah LifePO4 bank and be able to run my watermaker longer and safer (80A draw @24V as the Trojans suffer badly and only make 20-30min each time)
Aim to be able to be able not to fire up an engine/generator for up to a week or certain quiet anchorages. [that's the blackwater tank limit really...]

sorry slightly long winded...

cheers

V.

 

[Poey50]

And another question .
How do you top balances the batteries , say they arrive with different voltage, do the 123bms do that for you ? Or do you need to charge the battery separately?

Vic, this is covered in #9. There is a link to a page on the DIY Solar Forum which includes a Will Prowse video. In the top right-hand corner of that page is an orange download button. Click on that to download top-balancing instructions. I followed those to the letter.

From browsing on forums the top-balancing stage is the moment when people can seriously get it wrong and damage their investment before they have even used it. This has broken the hearts of several people on DIY Solar.

You will also need a bench-top power supply unit to do the final stage once they are all connected in parallel. I can certainly recommend the one i bought from Amazon.

 

[Poey50]

Vic, from my reading on the topic and what previous posters mention, you do charge them individually to same voltage, then get them all together and the bms takes over. Considering bms balancing deals with (what seems to be the norm of 1A loads shifting about), it would take ages to balance on the bms alone, so you more or less assist the system by balancing manually ahead.

See my reply to Vic above. Basically I connected the cells in series to begin with (4S) and added the BMS and relays. I programmed the BMS to disconnect when any cell reached 3.65 volts and then charged the whole pack with a 12 volt charger until the first cell that reached 3.65v disconnected the charge relay. This meant that all cells, although of course still not balanced, received the majority of their charge quickly. (The alternative would be to connect them all in parallel to start with creating a 3.2 volt battery of over 1,000ah which, with my 10amp power supply would take a very long time indeed.) After this series stage of charging I then put them in parallel and used the power supply unit to charge them to 3.65v (follow all warnings in the link on how to get that figure) and stopped when cells reached this and when current they would accept was very low. From memory I stopped when current was 0.04 amps. This last stage is very slow. What people do is lose patience, turn up the voltage on the unit to increase the amperage (having noted that the voltage measured at the cells is still under 3.65v forgetting that the charging is still in constant current mode, not constant voltage mode) and then come back to find the cells bloated and over-charged. The use of BMS alone to achieve a top balance is not generally recommended.

With regard to the second part of your post it might be a good idea to write to Zwerfcat. I imagine they would be happy to provide advice. Alternatively, the best source of general LFP advice that I have come across is the Lithium Batteries in Boats Facebook Group. Rod Collins, the author of the renowned Marine How To article on LFP, is a member.

 

[TernVI]

Not off-topic at all. Further back I mentioned some hybrid systems - detailed by Nordkyn. But an intriguingly simple method is carefully explained by ZwerkfCat. I've not heard of anyone using this method even though, as they explain it, it seems safe and obvious.

Lithium-Hybrid

That's interesting but raises a lot of questions.
Like how does the system respond dynamically?
What happens when big loads like inverters switch on and off quickly?
Does the BMS work quickly enough?

This is really slightly scary amounts of power for amateur experiments.

 

[Poey50]

That's interesting but raises a lot of questions.
Like how does the system respond dynamically?
What happens when big loads like inverters switch on and off quickly?
Does the BMS work quickly enough?

This is really slightly scary amounts of power for amateur experiments.

What's interesting is that it is clear from all the information on the site that that this is a person (or these are people) who have a deep knowledge of LFP yet are proposing the simplest imaginable solution - one that makes drop-in batteries a real possibility. I can't comment further on the fine detail as having invested in my 'dual bus' system I'm not really motivated to explore any other in detail but it looks worth interrogating further - just not by me. You will see how recently Zwerfcat posted their documents so that may explain that, of the dozens of systems I have come across in some way, I've never heard of one using that simple hybrid solution.

I totally agree with you that LFP isn't something to suck-it-and see given the amount of power in those little cells. Everyone needs to do their research first.

 

[vas]

good questions TernVI, I guess if one could "route" all charging to the FLA, and all loads on the LifePO4 (except for the mains charging through the inverter charger - but you can deal with that separately) So whatever loads are dealt by the LiPO4, and whatever charge from engine and solar goes first to the FLA.

Now, when LiPO4 is under whatever level you've set:
A. if there's juice coming in from solar/engine contactor (or B2B) switches on and it charges the LiPO4
B. if there's no solar/engine activity, it just drops offline and loads are dealt with from FLA
C. debatable if when there's both load and solar/engine activity if you should be filling the FLA first and the LiPO4 afterwards, or you should be pumping as much as you get first to cover your needs, but I guess it's a suck it and see (with your bms setup correctly and conservatively)

When there's new current intake from solar/engine you evaluate as per C above and charge goes to where you deem appropriate, according to DOD levels
something along these lines could be dead simple to accomplish using plcs and/or micros (imho). Not sure there are small/cheap suitable 200A contactors though

Poey50, regarding charging one cell by one, I'd confused what you explained with a video of a guy charging them with a massive bench power supply doing 20+A on the 3.something V cell... (mine only does 5A so will take a few weeks to do them all )

You mention that recently Zwerfcat posted docs of his work, got any links? The only things I've seen are from Dec 2019 where he's explaining the whole idea, he does mention his github which is basically empty, so not too sure what he's willing to contribute (or can for that matter).
Will mail him tonight, but not holding my breath.

V.

 

[Poey50]

Poey50, regarding charging one cell by one, I'd confused what you explained with a video of a guy charging them with a massive bench power supply doing 20+A on the 3.something V cell... (mine only does 5A so will take a few weeks to do them all )

Will Prowse is using a 10amp PSU - i have the same - not too pricey from Amazon. You don't use it to charge the cells in parallel until they are mostly charged. As said above go to this page

... Top Balancing LiFePo4 Cells using a low cost benchtop power supply.

... and download the document from the orange download tab in the top righthand corner. Those are the instructions to follow. This is a safe and fast method.

 

[vas]

Will Prowse is using a 10amp PSU - i have the same - not too pricey from Amazon. You don't use it to charge the cells in parallel until they are mostly charged. As said above go to this page

... Top Balancing LiFePo4 Cells using a low cost benchtop power supply.

... and download the document from the orange download tab in the top righthand corner. Those are the instructions to follow. This is a safe and fast method.

thanks, bookmarked it!
since I haven't yet bought cells, I'm not in a hurry, more keen to understand how a hybrid system would work (and how safe that would be)
Emailed Zwerfcat and will let you know when I have news from him.

cheers

V.