|
|
23-04-2022, 15:54
|
#736
|
Senior Cruiser
Join Date: Sep 2014
Location: puget sound washington
Boat: 1968 Islander bahama 24 hull 182, 1963 columbia 29 defender. hull # 60
Posts: 12,787
|
Re: Merged LiFeYPO4 1000Ah Winston prismatic cells and all electric galley...
Quote:
Originally Posted by CaptainRivet
You chage them to 3,65V as end off charge parameter? No room for error...
Thats no real charge and discharge 20AH from a 250AH bank and 200W charge....thats like trickle charging and discharging the LifePo4 bank like a lead...
Thats a minimal load even grade C would stay in balance witout balancer. And you do in fact an long absorption charge from 3.35 till 3.65V with tis low amps which equals to daily a top balance.
Discharge it regularly with 130A and charge with 120A and you will see them drifting apart quickly.
|
I do discharge them further at times and when on shore power I do have the ability to charge from 1 amp up to 120 amps . fully adjustable
I have not seen any drift at all in 3 years it's all the same as my initial top balancing so I see no need for doing anything different than I do now.
I am adding a nuwave2 induction unit this summer . I have an additional 200 watts of flex to put out if I feel its needed.
Also ave yet to see the need for my air x 400 marine. Wind generator . and yes I can active balance if I feel the need to my bms has that ability built into it . but after 3 years my capacity is still well above rated capacity and I have seen no loss .
Forgot to mention I am running CALB grade A aluminium case cells . https://www.cruisersforum.com/forums...ml#post3053353 is where I start describing my bank in full .
__________________
Non illigitamus carborundum
|
|
|
23-04-2022, 15:56
|
#737
|
Registered User
Join Date: Jun 2016
Location: Australia
Boat: Milkraft 60 ex trawler
Posts: 4,651
|
Re: Merged LiFeYPO4 1000Ah Winston prismatic cells and all electric galley...
Quote:
Originally Posted by s/v Jedi
.
When you order Winston cells from Julia Yu, you get sequential serial numbers and they are really well matched.
|
We got the same from Amy Wan and the EVE cells.
Paid a slight premium for matched and batched but seems well worth it.
|
|
|
23-04-2022, 15:59
|
#738
|
Senior Cruiser
Join Date: Sep 2014
Location: puget sound washington
Boat: 1968 Islander bahama 24 hull 182, 1963 columbia 29 defender. hull # 60
Posts: 12,787
|
Re: Merged LiFeYPO4 1000Ah Winston prismatic cells and all electric galley...
Quote:
Originally Posted by CaptainRivet
You chage them to 3,65V as end off charge parameter? No room for error...
Thats no real charge and discharge 20AH from a 250AH bank and 200W charge....thats like trickle charging and discharging the LifePo4 bank like a lead...
Thats a minimal load even grade C would stay in balance witout balancer. And you do in fact an long absorption charge from 3.35 till 3.65V with tis low amps which equals to daily a top balance.
Discharge it regularly with 130A and charge with 120A and you will see them drifting apart quickly.
|
I answered your query but I have a question why do you care what I do on my boat with my battery bank ?
Yes I fully expect my bank to outlive me and I have a good 40 years left in me .
__________________
Non illigitamus carborundum
|
|
|
23-04-2022, 16:17
|
#739
|
Registered User
Join Date: Sep 2019
Location: Lifeaboard
Boat: FP Lavezzi 40
Posts: 4,064
|
Re: Merged LiFeYPO4 1000Ah Winston prismatic cells and all electric galley...
My statement was for building a BMS and the fact that 8S 24V need a BMS with Balancer bevause 8S will drift apart without. The more the cells deviate the more capable the balancer must be and the more they drift.
The developer need to take into account that newbees with no glue will use the BMS.
It was not that you 2 need to do anything on your systems. You know it yourself and I know not enough of your total system to judge taht either.
Winstons are top notch when from Julia, are no Lifepo4 and eg 0.5C is no significant load, 0,66C is. In the hands of experienced user they don't need a BMS nor a balancer when not used in above 1C continous environment. I highly doubt any normal installation on a cruisers boat will come close to this spec. And if yes its someone like CatNewBee who knows what he is doing.
|
|
|
23-04-2022, 18:39
|
#740
|
always in motion is the future
Join Date: Feb 2009
Location: in paradise
Boat: Sundeer 64
Posts: 19,826
|
Re: Merged LiFeYPO4 1000Ah Winston prismatic cells and all electric galley...
Quote:
Originally Posted by CaptainRivet
My statement was for building a BMS and the fact that 8S 24V need a BMS with Balancer bevause 8S will drift apart without. The more the cells deviate the more capable the balancer must be and the more they drift.
The developer need to take into account that newbees with no glue will use the BMS.
It was not that you 2 need to do anything on your systems. You know it yourself and I know not enough of your total system to judge taht either.
Winstons are top notch when from Julia, are no Lifepo4 and eg 0.5C is no significant load, 0,66C is. In the hands of experienced user they don't need a BMS nor a balancer when not used in above 1C continous environment. I highly doubt any normal installation on a cruisers boat will come close to this spec. And if yes its someone like CatNewBee who knows what he is doing.
|
I can’t say you’re wrong because I simply don’t have the experience with 2nd rate cells, but your statements got me tempted to build an 8S with these so called Eve cells…. would I need to buy particularly bad ones or will good ones do to get them unbalanced during use? I would need to make sure you wouldn’t claim I got the wrong cells again when I find them to stay balanced !
__________________
“It’s a trap!” - Admiral Ackbar.
|
|
|
23-04-2022, 18:45
|
#741
|
always in motion is the future
Join Date: Feb 2009
Location: in paradise
Boat: Sundeer 64
Posts: 19,826
|
Re: Merged LiFeYPO4 1000Ah Winston prismatic cells and all electric galley...
Now that I think of it, it appears I already have another 8S bank of 650mAh… cylindrical. Would these do? Pretty easy to do 1C loads
https://www.amazon.com/gp/product/B0...?ie=UTF8&psc=1
__________________
“It’s a trap!” - Admiral Ackbar.
|
|
|
23-04-2022, 19:33
|
#742
|
Registered User
Join Date: Aug 2017
Boat: Lagoon 400S2
Posts: 3,755
|
Re: Merged LiFeYPO4 1000Ah Winston prismatic cells and all electric galley...
Quote:
Originally Posted by jordanbigel
Could I impose on you with an off-topic question? I just installed a CerboGX 2 weeks ago, and I guess I still have much to learn. I also have the 4 temp sensors wired to useful things (alternators, genset, Multiplus case) - and I see the temps on the menu page, as you show in your screenshot on the right side.
But how did you get your graphic display, I don't know what you call it, the main graphics display which shows the colored boxes - you have a Temp box in the top left I see, but I don't get how you got it to appear there. You wrote then it appears when you are off shore power, I can see your Multi says Inverting, as does mine, but I just have an empty red box in the top left corner.
Also, I see additional info on your display like amps in addition to watts, etc.
How? What is the way?
Thanks in advance!
|
I run it on a RaspberryPi and I looked at the scripts of Victron Venus OS (they are open source on GitHub), then I changed and extended them to show more useful information and use the screen area better, removed some unnecessary icons, made the switch toucheable to turn on and off the inverter, wrote some device drivers for 1-wire sensors and the REC ABMS RS485 serial interface to get all the settings in the UI, cell data, balancing information, cell temperatures, all that stuff. I reused the temp sensor menus but feed the data from the 1-wire RaspberryPi GPIO, same for I2C sensors, so I can use all the advantages of the Pi on board.
1-Wire is a 2 or 3 wires bus system ( Data I/O, +5V, GND, you can even omit +5V and can use Earth as ground then they run truly on a single wire using the small energy from data transmission from the host to charge a small capacitor and use this to respond, of course a 3 wire set up is far more reliable), the sensors use very low power and have unique addresses, you can hook up up to 10 sensors on a single 1-wire bus, really easy to use and versatile. As they are digital, they send the temperature information directly in a high precision, no need to do analog / digital conversions at the host - so you can use long wires for them. They come prewired in a waterproof housing and are quite cheap in bulk orders of 10pc.
Attachment 256530
|
|
|
24-04-2022, 16:36
|
#743
|
Registered User
Join Date: Sep 2019
Location: Lifeaboard
Boat: FP Lavezzi 40
Posts: 4,064
|
Re: Merged LiFeYPO4 1000Ah Winston prismatic cells and all electric galley...
Quote:
Originally Posted by Simi 60
Why do they need to communicate with each other?
They are looking after those 8 cells, not the other 16
They do if needed, if we get the bank to 100% (only saw it while commissioning on shore power) in the final few minutes of charging the cells get slightly out of whack, but as soon as the charging is stopped they are back to equal soon after.
|
Well you have "thrown away" one off the biggest advantages of building your own bank from raw cells compared to drop in ones.
Each BMS cares about 8 cells independently like in drop ins. But who manages and cares about the 3 packs together means your total bank...nothing... Yes they are in parallel BUT your BMS on each off the pack interfere with that. Additionally one pack will be the weakest and this will see more wear then the other over time, again nothing managing that as they are not communicating with each other.
You can say yes but if i have 1 BMS in a 3P8S config I don't monitor/can see the individual cells. True but therfor the whole bank as 1 is managed.
As with everything on a boat no advantage without a disadvantage...
|
|
|
24-04-2022, 17:24
|
#744
|
Registered User
Join Date: Jun 2016
Location: Australia
Boat: Milkraft 60 ex trawler
Posts: 4,651
|
Re: Merged LiFeYPO4 1000Ah Winston prismatic cells and all electric galley...
Quote:
Originally Posted by CaptainRivet
Well you have "thrown away" one off the biggest advantages of building your own bank from raw cells compared to drop in ones.
Each BMS cares about 8 cells independently like in drop ins. But who manages and cares about the 3 packs together means your total bank...nothing... Yes they are in parallel BUT your BMS on each off the pack interfere with that. Additionally one pack will be the weakest and this will see more wear then the other over time, again nothing managing that as they are not communicating with each other.
You can say yes but if i have 1 BMS in a 3P8S config I don't monitor/can see the individual cells. True but therfor the whole bank as 1 is managed.
|
I believe the BMS I have is capable of monitoring all cells, up to 20s but I didn't want a single battery.
As we are not land based and never in a marina I needed redundancy.
We have already stress tested this system by disconnecting 2 batteries and running everything on board on one battery
And as each bank has its own BMS the difference between individual batteries is just noise.
Quote:
As with everything on a boat no advantage without a disadvantage...
|
Relying on one single battery is a far greater disadvantage in my eyes
|
|
|
25-04-2022, 05:08
|
#745
|
always in motion is the future
Join Date: Feb 2009
Location: in paradise
Boat: Sundeer 64
Posts: 19,826
|
Re: Merged LiFeYPO4 1000Ah Winston prismatic cells and all electric galley...
I agree with Simi.
Also, many do not realize how multiple batteries, switched in parallel work. Let me try to explain:
You do not get the effect that one battery that is weaker wears down quicker. They will all cycle in sync.
Lets say 3 batteries are fully charged. This means that when you would disconnect them, their voltage is about the same. As you discharge them, the one that is weakest, gets to a lower SOC quicker. When you would disconnect, that battery starts showing a lower voltage. Because of this, it starts supplying less current than the other two. When it is almost empty, while the others still go strong, it enters the lower knee of the V-SOC graph and it’s voltage becomes so much lower that it will all but stop supplying current until the other two come down to a similar SOC.
The same happens when charging. The weaker one will see it’s internal resistance rise quicker so that it takes less and less of the charge current until the other batteries catch up.
This is what I meant with “parallel batteries balance with each other automatically”. Look at it as a power bus with several power packs that you connect to it or disconnect from it as you wish. You can keep some in reserve or you can move some to other locations to minimize wiring losses to high power consumers etc. Laws of nature keep everything working pretty good, while you have BMS’s that monitor each individual cell and prevent all issues around parallel cells where a weaker one can pull down it’s paralleled cells, negatively impacting overall functionality.
The negative is higher cost because you need to buy extra BMS’s, extra battery monitors and extra battery switches.
__________________
“It’s a trap!” - Admiral Ackbar.
|
|
|
26-04-2022, 01:08
|
#746
|
Registered User
Join Date: Aug 2008
Location: Sydney, Australia
Boat: Roberts 45
Posts: 1,040
|
Re: Merged LiFeYPO4 1000Ah Winston prismatic cells and all electric galley...
Quote:
Originally Posted by s/v Jedi
I[...]
Lets say 3 batteries are fully charged. This means that when you would disconnect them, their voltage is about the same. As you discharge them, the one that is weakest, gets to a lower SOC quicker. When you would disconnect, that battery starts showing a lower voltage. Because of this, it starts supplying less current than the other two. When it is almost empty, while the others still go strong, it enters the lower knee of the V-SOC graph and it’s voltage becomes so much lower that it will all but stop supplying current until the other two come down to a similar SOC.[...]
|
Nice summary! Let's hope many people read it carefully and understand it.
|
|
|
26-04-2022, 10:55
|
#747
|
Registered User
Join Date: Aug 2017
Boat: Lagoon 400S2
Posts: 3,755
|
Re: Merged LiFeYPO4 1000Ah Winston prismatic cells and all electric galley...
So news from our system, first time had an cell out of balance alarm and charger disconnect event in 4 years.
Last time I balanced it 1 year ago.
We had not charged full for over 40 days, PSOC use between 40 and 80..90% SOC, all looked well - until we charged full. The balancers were not able to get the bank in a safe balanced state almost 0.3V voltage difference to the lowest cell.
Will have to charge more frequently to 3.650V full in the future. May add additional balancers.
Now I balance the cells manually with ceramic resistors (5A), takes a couple of hours to get the cells to the same level.
So think twice if stopping charging earlier is a good thing or a bad thing.
|
|
|
26-04-2022, 11:36
|
#748
|
Registered User
Join Date: Nov 2018
Location: San Diego
Boat: Shannon 50 Ketch
Posts: 732
|
Re: Merged LiFeYPO4 1000Ah Winston prismatic cells and all electric galley...
Quote:
Originally Posted by CatNewBee
I run it on a RaspberryPi and I looked at the scripts of Victron Venus OS (they are open source on GitHub), then I changed and extended them to show more useful information and use the screen area better, removed some unnecessary icons, made the switch toucheable to turn on and off the inverter, wrote some device drivers for 1-wire sensors and the REC ABMS RS485 serial interface to get all the settings in the UI, cell data, balancing information, cell temperatures, all that stuff. I reused the temp sensor menus but feed the data from the 1-wire RaspberryPi GPIO, same for I2C sensors, so I can use all the advantages of the Pi on board.
1-Wire is a 2 or 3 wires bus system ( Data I/O, +5V, GND, you can even omit +5V and can use Earth as ground then they run truly on a single wire using the small energy from data transmission from the host to charge a small capacitor and use this to respond, of course a 3 wire set up is far more reliable), the sensors use very low power and have unique addresses, you can hook up up to 10 sensors on a single 1-wire bus, really easy to use and versatile. As they are digital, they send the temperature information directly in a high precision, no need to do analog / digital conversions at the host - so you can use long wires for them. They come prewired in a waterproof housing and are quite cheap in bulk orders of 10pc.
Attachment 256530
|
To restate, you are running a modified version of Venus OS on a RPI, not a CerboGX device - so that screenshot did not come from a CerboGX device, in which case, there is no way that I (using a CerboGX) could replicate that screen.
Since I asked this question, I've found out there is a way to customize the CerboGX to display custom screens (or custom components on an existing screen maybe?) but I have not delved into the details yet. I thought your screenshot was from a CerboGX.
The stuff about the sensors was also fascinating.
|
|
|
26-04-2022, 16:47
|
#749
|
Registered User
Join Date: Aug 2017
Boat: Lagoon 400S2
Posts: 3,755
|
Re: Merged LiFeYPO4 1000Ah Winston prismatic cells and all electric galley...
Quote:
Originally Posted by jordanbigel
To restate, you are running a modified version of Venus OS on a RPI, not a CerboGX device - so that screenshot did not come from a CerboGX device, in which case, there is no way that I (using a CerboGX) could replicate that screen.
Since I asked this question, I've found out there is a way to customize the CerboGX to display custom screens (or custom components on an existing screen maybe?) but I have not delved into the details yet. I thought your screenshot was from a CerboGX.
The stuff about the sensors was also fascinating.
|
There is a Cerbo mod with an installer, that just tweaks the display, but does not change / extend the functionality of the Victron hardware. Many people use it on top of cerbo / ccgx / venus / octo. It works out of the box and there are YT videos on how to install and configure it.
My version for the pi is too much adapted to non-standard sensors and interfaces and during the development I was not thinking about to make it run it on a different system - , so it has some hard coded stuff inside.
|
|
|
26-04-2022, 22:19
|
#750
|
Registered User
Join Date: Aug 2017
Boat: Lagoon 400S2
Posts: 3,755
|
Re: Merged LiFeYPO4 1000Ah Winston prismatic cells and all electric galley...
https://github.com/kwindrem/GuiMods
Here is the link to the project of the Victron user interface modification, you can search on YT for installation videos.
This is not what I have programmed, this is the Victron patch, I personally have not used it, but it is well documented...
|
|
|
|
|
Thread Tools |
Search this Thread |
|
|
Display Modes |
Rate This Thread |
Linear Mode
|
|
Posting Rules
|
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
HTML code is Off
|
|
|
|
Advertise Here
Recent Discussions |
|
|
|
|
|
|
|
|
|
|
|
|
Vendor Spotlight |
|
|
|
|
|