vas
Well-known member
greg, have a look at this video:
explains quite a few things re charging nicely imho!
V.
explains quite a few things re charging nicely imho!
V.
270ah DIY LiFePO4 build
- Thread starter Poey50
- Start date
Poey50
Well-known member
I think it has cropped up in several places but I can't pinpoint them at the moment. In my mind I have it marked down as a reliable fact rather than an a contested opinion. I've tested it myself over last winter and quite a lot of lockdown and it works. If I come across a source I'll post it.
Edit. Apologies - I've conflated two things in my mind. Age and tiredness are my excuses. The first thing is that the pivot point is not 13.4 pack voltage but 3.4 volts per cell. Nordkyn quotes this here Charging Marine Lithium Battery Banks | Nordkyn Design and it is also a figure that Battleborn quote. 3.4 volts per cell is of course 13.6 volts for a nominal 12 volt pack. The second thing is from experienced members of the Lithium Batteries in Boats Facebook group where members have found a 13.2 or 13.3 volt figure is a good maintenance figure - 13.6 volts being too high. It is something to experiment with but as said around 13.3 pack voltage works for me and I'm reassured that it is well under the 13.6 pack voltage pivot point.
Last edited:
sailaboutvic
Well-known member
Can I ask why not just discharge to say 60/65% and leave all solar turned off ?
We will be away for about a month to six weeks from next week and that's what I be doing , as lithium loses hardly any thing .
We will be away for about a month to six weeks from next week and that's what I be doing , as lithium loses hardly any thing .
Poey50
Well-known member
Yes, self-discharge is around 3% per month, depending on temperature, so no problem. I do have some residual loads so for a 5 month haul-out I prefer a maintenance charge on solar.
gregcope
Well-known member
If there were no residual loads just disconnecting should work for many, many months. However many systems have shunts, MPPT charge controllers and Bilge Pumps that all have a quiescent draw. Hence my line of questioning as this is driven by not being live-aboards and more occasional users of our boat that is on a mooring (every other weekend and a summer cruise or two?) and then you have the long winter haul out issue where you could disconnect and just re-charge at the end.
sailaboutvic
Well-known member
If you totally disconnect as I am planning to do there be no draw from shunt/mppt or any thing else , only possible draw would be a bilge pump and with a decent bank even at 50% that's a lot of pump working Hours and if you have that bigger in take of water the batteries be the less of your problem .
Poey says it's 3% pm I don't know what the figures are I guess unless you checked it out on your own bank you know but 3% pm on my 400 ah bank is nothing to talk about.
When I build mine going onto a year now I also build another cruiser with the same cells , he did go off for six month back to the UK when he return he found very little discharge on any of his 8 cells from when he left .
Poey says it's 3% pm I don't know what the figures are I guess unless you checked it out on your own bank you know but 3% pm on my 400 ah bank is nothing to talk about.
When I build mine going onto a year now I also build another cruiser with the same cells , he did go off for six month back to the UK when he return he found very little discharge on any of his 8 cells from when he left .
gregcope
Well-known member
I've watched that. Interesting stuff and I think this is where I am confused.
At this point;
At the end he talks about comments on videos that mention a "threshold", without going into details, where above that it will charge to 100% if you apply current. This feels like what @Poey50 has described.
gregcope
Well-known member
Agree bar the leaks from LEDS on the BMS (if any) which can add up over time. There is an RV youtuber who lost his LIfePO4 bank to LEDS after leaving his RV in a garage for months on end without charging. You could have BMS, Shunt, B2B, Bilge Pump, MPPT controllers etc... all sucking a little bit over time.
On your pack size (400Ah) this sounds less of an issue.
sailaboutvic
Well-known member
Once I trip my relay/ breaker nothing comes out of my batteries although the bilge pump is wired throught it's fuse to the batteries direct.
My 123smart BMS do have a small led light on each one (4) but I doubt if there draw anything worth worrying about , maybe Poey know the answer to what they draw.
My 123smart BMS do have a small led light on each one (4) but I doubt if there draw anything worth worrying about , maybe Poey know the answer to what they draw.
Poey50
Well-known member
Between boards (my number 2 and 3 cells) are <1mA and the begin and end boards (because of standby Bluetooth and current sensors) are <1.6mA. So max of 5.2mA = 0.125 ah per day or 1 amp hour every 8 days so 4 ah per month.
3% self-discharge on 400 ah is 12 ah per month so a maximum loss of 12 + 4 = 16 ah per month (approx.). Given that there are no worries about sulfation this is not much to be worried about with no residual loads. In winter temperatures the self-discharge would be lower.
Last edited:
vas
Well-known member
if you do have some solar, wouldn't it make sense to leave a smallish fridge at a fairly high temp running 247?
I'm doing that but again I'm down south so sun is visiting us often having said that three days of minimal sunshine and trojans are down to 75%
I'm doing that but again I'm down south so sun is visiting us often
having said that three days of minimal sunshine and trojans are down to 75%
sailaboutvic
Well-known member
For me to do that I first would have to rewire the fridge to the batteries as once my relay off it turn stop any power leaving the batteries IE totally isolate the batteries.if you do have some solar, wouldn't it make sense to leave a smallish fridge at a fairly high temp running 247?
I'm doing that but again I'm down south so sun is visiting us often
Plus it also mean I would have to leave the the keel cooler sea cock open to top that there a risk of some kind of short , panels/ mppt/ shunt
Personally I see no advance doing this ,
I rarely switch everything off .
Another advantage of lithium .
gregcope
Well-known member
So...
I ordered some 32700 Cells to learn/practice on. I want to make a facsimile of a much larger pack and also have something off the boat to prototype on. I intend to plugin a JDB BMS and hopefully plugin a Raspberry PI to the RS485 port. This should allow me to learn how to use the DC Charger, BMS and battery capacity device without risking a much larger cell.
I ordered these; PALO 7200mAh 3.2V 32700 LiFePO4 Battery 32700 35A Continuous Discharge Maximum 55A High Power Batteries +DIY Nickel sheets. They took a few weeks to arrive. Came with little battery frame things.
I found these off-grid-Garage. I wanted some with tabs to make it easier to interconnect them as I do not have a cell welder. Capacity wise various youtubers shows between 5000-6000 mAh not the advertised 7000 mAh. Capacity is not a huge concern, just want to get something working.
Basic plan is to;
1. Assemble pack in final design topology
2. Top balance with some custom cables @3.2V nominal
3. Re-cable @12V nominal and get the BMS working
4. Capacity test
5. Test JDB BMS bluetooth / settings
6. Config a Raspberry PI with the RS485 port and Victron OS
1. Assemble pack in final design topology
First off I ripped two tabs off one cell when trying to drill it with a 6mm drill. On the next ones I sandwiched the tab between bits of wood which worked. I then soldered the now mangled tab back on. Looks a mess. Should work... I appreciate this might damage the cell.
The logic here is to practice for a larger pack where they will be assembled in final configuration. ie Wrapped with tape to offer some "compression" when partially charged to avoid charging them to 100% without compression. The the busbars used in the 12V config (ie serial). This means using cables to top balance them when in parallel.
2. Top balance with some custom cables
I then made some balance cables - 10mm2 cable with M6 ring terminals connected with brass bolts. I had previously ordered the Desktop charger than @Poey50 recommended and learnt how to configure that. Presently charging about 2A to 3.4v as I am stepping up to 3.65v to understand how they behave.
I did short some of the cells when assembling. One was not my fault! The others were.
Lessons learnt so far
1. Assembling a small pack is quite tricky due to space being limited. Larger 50Ah cells would be easier
2. Check the polarity with a multimeter. The blue cover on one cell had the positive mark on the negative! Hence why I initially shorted that cell which was a surprise!
3. Be very, very careful on preventing short circuits.
4. Wear glasses/safety glasses!
5. The Nickel tabs are very fragile - Cells with bolts would have been better
6. I am learning lots, so the plan of cheap cells is working!
Photo's of the frankenstein thing to follow.
Thanks for the inspiration @Poey50
I ordered some 32700 Cells to learn/practice on. I want to make a facsimile of a much larger pack and also have something off the boat to prototype on. I intend to plugin a JDB BMS and hopefully plugin a Raspberry PI to the RS485 port. This should allow me to learn how to use the DC Charger, BMS and battery capacity device without risking a much larger cell.
I ordered these; PALO 7200mAh 3.2V 32700 LiFePO4 Battery 32700 35A Continuous Discharge Maximum 55A High Power Batteries +DIY Nickel sheets. They took a few weeks to arrive. Came with little battery frame things.
I found these off-grid-Garage. I wanted some with tabs to make it easier to interconnect them as I do not have a cell welder. Capacity wise various youtubers shows between 5000-6000 mAh not the advertised 7000 mAh. Capacity is not a huge concern, just want to get something working.
Basic plan is to;
1. Assemble pack in final design topology
2. Top balance with some custom cables @3.2V nominal
3. Re-cable @12V nominal and get the BMS working
4. Capacity test
5. Test JDB BMS bluetooth / settings
6. Config a Raspberry PI with the RS485 port and Victron OS
1. Assemble pack in final design topology
First off I ripped two tabs off one cell when trying to drill it with a 6mm drill. On the next ones I sandwiched the tab between bits of wood which worked. I then soldered the now mangled tab back on. Looks a mess. Should work... I appreciate this might damage the cell.
The logic here is to practice for a larger pack where they will be assembled in final configuration. ie Wrapped with tape to offer some "compression" when partially charged to avoid charging them to 100% without compression. The the busbars used in the 12V config (ie serial). This means using cables to top balance them when in parallel.
2. Top balance with some custom cables
I then made some balance cables - 10mm2 cable with M6 ring terminals connected with brass bolts. I had previously ordered the Desktop charger than @Poey50 recommended and learnt how to configure that. Presently charging about 2A to 3.4v as I am stepping up to 3.65v to understand how they behave.
I did short some of the cells when assembling. One was not my fault! The others were.
Lessons learnt so far
1. Assembling a small pack is quite tricky due to space being limited. Larger 50Ah cells would be easier
2. Check the polarity with a multimeter. The blue cover on one cell had the positive mark on the negative! Hence why I initially shorted that cell which was a surprise!
3. Be very, very careful on preventing short circuits.
4. Wear glasses/safety glasses!
5. The Nickel tabs are very fragile - Cells with bolts would have been better
6. I am learning lots, so the plan of cheap cells is working!
Photo's of the frankenstein thing to follow.
Thanks for the inspiration @Poey50
Last edited:
gregcope
Well-known member
Photos
vas
Well-known member
good thinking, that's what I'm about to do as well, just ordered 4 18650 cells with a case for them, so that I can try the bms and controller boards when they arrive from China (FedEx got them today, so anything from 10days to 2m )
Also planning to pick from a friend a couple of slightly dead 18650 cells so that I can play with a small 12V inverter and check that all the ballancing features and reporting/shutting down work before I hook them up to the real cells
)Also planning to pick from a friend a couple of slightly dead 18650 cells so that I can play with a small 12V inverter and check that all the ballancing features and reporting/shutting down work before I hook them up to the real cells
gregcope
Well-known member
@vas. Have a good think about how you are going to interconnect cells. I hope the case you are getting has tab/bolts/something to make that easy.
My get tabbed ones and drill has not worked as well as I hoped! It appears to work, but is not pretty and certainly not fit for use on a boat. .. too many exposed / movable connections that could short.
My get tabbed ones and drill has not worked as well as I hoped! It appears to work, but is not pretty and certainly not fit for use on a boat. .. too many exposed / movable connections that could short.
vas
Well-known member
well, I'm getting these:
so pretty easy to get a battery in each one
I'll have a massive 3.2Ah 12V bank to play with, easy to empty and recharge
so pretty easy to get a battery in each one
I'll have a massive 3.2Ah 12V bank to play with, easy to empty and recharge
gregcope
Well-known member
That is a good approach. I would do similar if doing it again.
vas
Well-known member
lesson of the day: check the drawings of the batteries before moding the BMS cell balancer to fit!
diyBMS custom cell balancers ordered are like this with an 89mm spacing between the terminals - fit and forget:
ends up that the EVE280Ah terminals are 123mm appart though :-(
For the record, I modded again the design and that's how it should look like:
(left the old terminal for reference...)
oh well, will use short copper bars to join the one side to the terminal.
Currently thinking about the spacers I should place between the cells before compressing them or not (if you follow solar garage guy's opinion)
Cheap and easy to use a 4mm ply, worried about expansion/soaking dampness, other option being thin grp sheet (not too keen building them tbh!) or else some plastic thin sheeting, perspex, polycarbonate, etc.
what have you actually used guys?
Thinking of getting these inbetween liners/spacers higher so that when fitted with the cells, there's going to be a 10-20mm space under the cells to the floor and another 50-60mm above so that it wont be easy to have things dropping on the live cell terminals and these bms balancer boards. There's going to be a clear perspex cover anyway so that I can see that everything is fine without having to lift the lid. Two leds on each board give an immediate view of comms or battery issues.
V.
diyBMS custom cell balancers ordered are like this with an 89mm spacing between the terminals - fit and forget:
ends up that the EVE280Ah terminals are 123mm appart though :-(
For the record, I modded again the design and that's how it should look like:
(left the old terminal for reference...)
oh well, will use short copper bars to join the one side to the terminal.
Currently thinking about the spacers I should place between the cells before compressing them or not (if you follow solar garage guy's opinion)
Cheap and easy to use a 4mm ply, worried about expansion/soaking dampness, other option being thin grp sheet (not too keen building them tbh!) or else some plastic thin sheeting, perspex, polycarbonate, etc.
what have you actually used guys?
Thinking of getting these inbetween liners/spacers higher so that when fitted with the cells, there's going to be a 10-20mm space under the cells to the floor and another 50-60mm above so that it wont be easy to have things dropping on the live cell terminals and these bms balancer boards. There's going to be a clear perspex cover anyway so that I can see that everything is fine without having to lift the lid. Two leds on each board give an immediate view of comms or battery issues.
V.
