Sometimes the most good you can achieve is to serve as a warning to others…
YES!!! For the first time EVER I can set the system to Keep Charged, and the cells don’t run away!!!
And I don’t get an ESS #1 message, as I leave the system to do its thing.
Lessons learned:
- Never use cheap Lifepo4 cells other than camping systems … nope just don’t.
- Make sure you top-balance the cells … and don’t be afraid to do it again later if needed.
- And have a BMS that shows you the cell voltages … if you are DIY.
- And in a bind, you can mix and match the same quality/resistance Lifepo4 cells.
Just look at this beautiful picture … aaaa, so nice! 
Note: Frankenstein bank …
Cells 1-12 are 100ah 2nd Life cells.
Cells 13-17 are 280ah A Grade Batch Matched cells.
Bank AH is set at 100ah.
I think it is time to insert cell 18, still, 3.45v per cell - so max 62.1v, well clear of 66v max of Victron and the JBD BMS.
Unless someone has an opinion on why the not … please speak up now!?
So it was working all of 10mins, time to mess with it? I’m starting to understand 
1 Like
Been watching the 100% SOC for weeks now, no issue whatsoever.
The challenge I had when I set the SOC to 100% killed my idea for 100% SOC before evening.
From long before, the Keep Charge I was skittish of since it made dramas before. Where I got most of the DC Ripples from.
So sitting today, methinks let’s try it again the Keep Charged.
I watched the volts climb from 58.65v to +59.5v where the system discharge the batts again to +58.66v … where it has been sitting for a long time now, stable.
2 Likes
I am starting to trust these cells … when the new 280ah cells arrive later this month, I want to start 18 cell 280ah bank.
The Keep Charge function is then ONLY to be tried once they have settled a bit.
So the Keep Charged and 18 cells are two separate issues … I just have never before seen the system draw the bank down again, like it did today.
I literally sat with my finger on the "button, so to speak, to disable the option if it went past 62v. 
The Delta stayed comfortably <0.01v all the time, BMS balancing quite nicely.
Right, the Keep Charged worked perfectly the last few days. So that is sorted.
And we now have an 18 cells bank - currently testing the Keep Charged for a few moments to see where it peaks … nicely top-balanced on 3.45v from the 17 cell bank.
Inverter Settings:
- 3.375v max charge per cell
- Guessed the Float.
- Did not change much else here - open for advices.
I set BMS to these volts for SOC … using this as my guideline:
BMV and BMS SOC in line … for now:
Let’s see how this works …
Jinne, I’m nervous!
So far I can report that the Apple App has an issue showing graphically the 18 cells.
the bar graph has to be refreshed all the time to show it more realistically.
Works fine with <18 cells.
Nee wat … at 3.375v the SOC of the BMS and BMV has a huge difference, like 72% vs 100%, where it stops charging.
Yes, I know re. the volts, but it was close as damn on 3.45v.
So back to 3.45v as before, BMS SOC settings also back as before, 100% SOC at 3.45v
Ps. Obviously, when I want to test from panels to batts, Cpt has rain again. 
Result so far:
- BMV - yesterday at 3.375vvs today at 3.45v per cell, being my “benchmark”.
- BMS
It “feels like” that at 3.45v, the BMS SOC is the closest to BMV SOC.
So far the 18 cells work quite nicely.
Same stability, as one would expect, as using 16 cell bank.
1 Like
And to think it is 13 x 100ah and 5 x 280ah cells in a 100ah bank.
I think I like Lifepo4 again.
Maby derailing the tread abit, but maby not.
It still have to do with a 17s string.
So im getting my cells when the ship berths. This is somewhere is 2022 as things are currently going with shipping!!!
So I need to top balance these cells, but do not have a bench power supply.
I do however, have a buck 150w boost charger that i use for projects.
If i set that to 3.5v, and input 20w, with a fan to keep everything nice and cool. (It said max 10a, but i like to be safe. Not really want to go higer than 6a)
Will this be doable, or do i need a bench power supply being use on countless Youtube videos?
I know it will quite a while, but have to do with what i have…
I would suggest you use your BMS. Just limit the charge current <5A and use a lowe battery voltage which you increase every day. Charge this with your PV or inverter (in Victron system with the keep charged option).
So you start with 3.35V x cells and increase that to 3.45V over a few days (0.02V each day over 5 days will work well).
So if you use 16 cells you will start with 53,6V and end with 55,2V after 5 days. Make you the balancer on your BMS is on.
Thank you Louisvdw. Will do that. Im just worried that that the BMS might shut down. And cause damage to mppt.
I have all the tools to build a bench power supply, just need a variable booster board. Mantech has one. 12-80v. Will get meself one of them and charge the bms at 1a. Will take about 7 days, but then it should be right.
The booster is R400, mppt R12k. I can blow many R400 before i get to R12k!! 
18 balanced cells worked really well.
And because the idea was always to have 18 cells, 3 as “spare”, let’s see how a 15 cell bank works as I had to sell 3 of the 280ah cells to another DIY’er who also wants to try 18 cell banks.
The new order must now arrive, the ship is anchored in front of Blouberg. Hopefully will get them before the end of Jan 2022. 
@TheTerribleTriplet I wonder if they ever heard of cape town pirates 
There she sits …
Supposed to enter the harbor on 11 Jan 2022 at 8 pm, offload at 10 pm, so we could expect the goods on the 15th but then I got informed SARS wants to inspect the cargo again.
This order was paid for on ± the 15th of Sept 2021 if memory serves. Only after 2 failed shipping attempts, we got lucky on the 3rd one. Was supposed to be here Nov.
Absolutely NOTHING one can do but wait.
Sold 3 of the 280ah cells, playing with 15 cells now.
Man, what a blast to be able to go 15, 16,17 or 18 cells. Victron rules!
Here is my new volts “bible” … anyone cares to comment?
(With Compliments of Wibla from DIY Solar Forum.)
On a 18 cell bank, keeping them between 3.2v bottom, and 3.375v top, that it will prolong the life of the cells tremendously.
The BMS settings on the side … interesting too!
Why? More cells rather than harder cell use, is the idea behind it all, as per @Phil.g00
So I sold 16 of the 32 cells. Keep in mind, of the 16 cells left, one was from the previous 280ah batch, 4k cell. The new batch is 6k cells. Silly me sold 4 cells. 
EVE says don’t mix the two models 280ah cells. But I had one cell short, want 16, the 100ah Frankenstein bank worked, so let’s see what happens.
From here when I started:
To here - cell 8 being the 4k previous batch cell, where it starts to shoot out:
Ps. Have not had a DC Ripple with these new BMS’es yet.
Ordered 3 cells of the same batch we bought last, will arrive in a few weeks’ time, having waited months due to Chinese Covid lockdowns recently.
18 cell 280ah cells are now becoming a reality, done messing around … I think. 
No no no … I’m done.
Titbit to share:
How did I get the bank balanced this round?
The weather was bad, so I thought, bugger this, Enough! I’m so over stressing each time with bad weather and recharging the bank, just put the system on Keep Charged and forget about it. 24-40watt on some days, hell with this.
Started on Bulk 3.412v per cell, Float 3.375v, and let it be. Victron and BMS worked together and got the cells balanced over 24 hours.
Then I upped to Bulk 3.45v, Float 3.375v still, discharged it a wee bit, and put it back to Keep Charged. It did its thing again over 24 hours.
As the bank got to the volts set, you can see Victron charge a bit, then discharge, keeps it up till all is settled and happy, cells nicely balanced, bar no 8. Could never get that on the original bank, the DC Ripple bank. Now I can. YES!
BUT, I still could not figure out WHY is it when discharging, it is balancing. I set the BMS to Charge Balance!? This is NOT good for balancing the bank overall.
Lost in translation - keep it off.
Right, next level …
Have 15 cells currently, got 4 more, I call it my “Warranty Extension” as I was getting really stressed.
All the cells are from the same batch, same resistance, being A Grade Batch Matched.
I don’t run any interface between BMS/Cerbo … on 3.45v currently, I end up with a ±0.008v Delta every day.
Victron max volts are 66v, right - sorted.
BMS can go higher than that - sorted.
The tricky part is … how to add 3 more cells into an existing bank, with LS getting the "Top Balancing “thing” right.
The ideas I’m stewing on:
Idea 1:
- The 4 cells, using an EV.Peak C3, balance the 4 new cells.
- Then, very carefully, over a week or so, take the 15 cells up to close as I can get to 3.65v.
- Take everything apart, connect all 18 cells parallel, and connect the EV.Peak … and hope like everything there is no LS if it takes too long.
Idea 2:
Leave step 3 above and just wing it, and coach the BMS to do the top balancing over weeks.
Settings:
- 18 x 3.375v = 60.75v so have 17.01kWh
- 18 x 3.45v = 62.1v and have a 17.388kWh
… Or … - 19 x 3.3v = 62.7v a 17.556kWh - and don’t have to worry to keep the one cells charged.
What are the risks!?
Does anyone have any other suggestions?
What would you do @Phil.g00 … YOU are the “cause” of this idea. 
More info …
Inverter is set to 1500w max at 6 pm so that by the time the sun takes over I’m between 20-35% SOC depending on what loads at night.
3 geysers and batts are recharged every day, I start looking for stuff to switch on on good days.
The current batt is idling.
“Warranty Extension” as I said … and IF we get really bad weather and cables are stolen, or a transformer pops, I don’t have to rush out for a gennie.