Merged LiFeYPO4 1000Ah Winston prismatic cells and all electric galley...

[s/v Jedi]

Originally Posted by newhaul

Im a retired navy engineer ( hull tech) i always think worst case and plan ahead for it so it wont happen . How many ways have you planned to stop water ingress due to hitting something at 0300? ( just an example of my thinking )

My whole boat choice was based on that. You can cut off 10’ from my bow and I will be fine. Same for the stern and the center compartment has no thruhulls plus we have a kevlar reinforced laminated at the turn of the bilge.

The first owner thought it was not enough and besides the 10,000gph engine mounted crash pump, added another electric 8,000gph pump in the keel sump

[beckett]

Originally Posted by CatNewBee

The SI is the big brother of the ABMS, it measures and reacts on cell values, it just has less outputs. It is totaly comparable to all other BMS around with either single bus switching or charge / load bus switching, and it also has CAN integration, it only misses the two extra contacts the ABMS provides.

I spoke with Tine from REC and he had some questions I could not find answers yet. Called MV and sent them emails but I think because of COVID they are still not open yet.
Do you know if can I soft control my MasteVolt 24/80/3 charger? I know that I can put my charger in Standby mode from my EasyView 5 control panel, however, I could not find any inputs or outputs on the charger itself ... The EasyView panel most likely talks with the charger using NMEA 2000 or some proprietary protocol ... do you know? I am just trying to see if I can still use my MV charger and MV Intvertet or I should just convert to all Victron systems and dump the charger and possibly keep MV Inverter as you do with Power Assist feature?

[beckett]

A question about Winston cells dimensions and their expansion forces. I have a battery box that is molded from GRP and 540mm wide. I am planning for 8 Winston 700Ah cells which theoretically should be 536mm wide, which leaves me with just 2 mm on each side. I wonder:
1. if the Winston cell dimensions are pretty accurate or they ran bigger or smaller than specified. In my case, anything +0.5mm per cell would present me with a problem and I would have to create a new battery case.
2. What force the Winston cells create when expand and what reinforcement of my battery box will be sufficient? I see people engineer different solutions and in my case, I would like not to overengineer as my battery box with 3mm GRP walls might be enough or just need to be reinforced to deal with the cell's expanding forces.

Thank you.

[CatNewBee]

MV is an own proprietary ecosystem. There are digital input interfaces on the market that speak Madterbus, and you can use those signals to switch MV gear. Not sure about your particular model.

Regarding the Winston cells, you better approach julia from Winston for the specs.

[beckett]

Originally Posted by CatNewBee

MV is an own proprietary ecosystem. There are digital input interfaces on the market that speak Madterbus, and you can use those signals to switch MV gear. Not sure about your particular model.

Regarding the Winston cells, you better approach julia from Winston for the specs.

Thanks, CatNewBee. What have you used to separate your electrical into charge and consumer buses? I wonder if Victron's Lynx DC distribution systems, Victron's PowerIn or busbars worth considering for my conversion?

[CatNewBee]

Originally Posted by beckett

Thanks, CatNewBee. What have you used to separate your electrical into charge and consumer buses? I wonder if Victron's Lynx DC distribution systems, Victron's PowerIn or busbars worth considering for my conversion?

See on the first few posts my circuit diagrams, I have used 3 ML-RBS 7700 solenoids plus one more for the legacy chargers.

You can of course use whatever you feel comfortable with, but I cannot comment on other solutions, I don't have the gear to test them, nor the time to read all the manuals and calculate the expected loads and evaluate the necessary interfaces.

But take your time and do the calculations, every use case is different, get the manuals and read through what inputs are required and how they need to be set up...

If you want to go all Victron, check their websites, they have example diagrams of the integration with smart batteries, BMS and all the other gear.

[CatNewBee]

Hi guys,

I have made a very strange observation on my cells mis-behaving, what lead to an ever growing imbalance.

I've got a cell imbalance error and started investigating of the cause.

While discharging, cell differences remained cell 1 was higher than 4, then with some distance cell 3 and cell 2 was lowest. It was plausible because it reflected the state of the cells at the error condition.

I discharged about 300Ah and then the next day the sun kicked in. Battery was charging and after a short time, surprisingly cell 2 became the highest, followed by 3 and then 1 and 2. this condition remained during the whole charge process up to 3.4V where the cells were almost even again. My balancing settings caused all the time a discharge of cell 2 and probably 3 up to this point around 98% SOC.

The remaining 20Ah become wild, cell 2 (that was balanced down all the time) remained at 3.4V almost all the time, cell 3 goes up to 3.5 and cell 1 and 2 run off high (see pictures)
Causing an imbalance over 250mV (RAZL) and firing the error. The remaining 20Ah are charged quite quickly, so the balancers could not bring cell 2 back up during this time and get the cell 1 and 2 down.

Now I have changed the BMIN to 3.45V and BVOL to 3.50V to prevent wrong balancing during the plateau phase, and allow balancing only up the shoulders. I hope this will get my cells back to normal soon. Will post how it develops.

So please check your settings and adjust them if necessary and do some cell voltage monitoring during charge and discharge.

So lessons learned, never trust voltage differences below the shoulders, they are not good as balancing set points.

All Cells at 3.4V does not mean, they are in balance at all (13.6V), they can differ of more than 2%, and balancing below this can have the opposite effect by increasing the inbalance.

It was my fault to assume I could extend the balancing time by starting earlier with balancing, actually this screwed up the balance of the cells by discharging the wrong cell for a long time.

All Cells at 3.4V does not mean, they are in balance at all

But on the other hand, great to have a BMS, that can be configured to anything necessary.

[beckett]

Originally Posted by CatNewBee

Hi guys,

I have made a very strange observation on my cells mis-behaving, what lead to an ever growing imbalance.

I've got a cell imbalance error and started investigating of the cause.

While discharging, cell differences remained cell 1 was higher than 4, then with some distance cell 3 and cell 2 was lowest. It was plausible because it reflected the state of the cells at the error condition.

I discharged about 300Ah and then the next day the sun kicked in. Battery was charging and after a short time, surprisingly cell 2 became the highest, followed by 3 and then 1 and 2. this condition remained during the whole charge process up to 3.4V where the cells were almost even again. My balancing settings caused all the time a discharge of cell 2 and probably 3 up to this point around 98% SOC.

The remaining 20Ah become wild, cell 2 (that was balanced down all the time) remained at 3.4V almost all the time, cell 3 goes up to 3.5 and cell 1 and 2 run off high (see pictures)
Causing an imbalance over 250mV (RAZL) and firing the error. The remaining 20Ah are charged quite quickly, so the balancers could not bring cell 2 back up during this time and get the cell 1 and 2 down.

Now I have changed the BMIN to 3.45V and BVOL to 3.50V to prevent wrong balancing during the plateau phase, and allow balancing only up the shoulders. I hope this will get my cells back to normal soon. Will post how it develops.

So please check your settings and adjust them if necessary and do some cell voltage monitoring during charge and discharge.

So lessons learned, never trust voltage differences below the shoulders, they are not good as balancing set points.

All Cells at 3.4V does not mean, they are in balance at all (13.6V), they can differ of more than 2%, and balancing below this can have the opposite effect by increasing the inbalance.

It was my fault to assume I could extend the balancing time by starting earlier with balancing, actually this screwed up the balance of the cells by discharging the wrong cell for a long time.

All Cells at 3.4V does not mean, they are in balance at all

But on the other hand, great to have a BMS, that can be configured to anything necessary.

Attachment 218006Attachment 218007Attachment 218008Attachment 218009

I think I understand what was the issues, however, could you please clarify why balancing earlier caused "screwed up the balance of the cells by discharging the wrong cell for a long time"? Is this in any way connected to the balancing currents and your capacity of your cells and the times needed to balance?

Just wondering what is BMIN and BVOL was set to before you made this change? What is you stop charging voltage set to? Is it the same as BVOL? Thank you, these are very interesting observations!

[CatNewBee]

the old settings...

[CatNewBee]

Todays settings tweaked to balance asap

[SweareDeep]

Originally Posted by CatNewBee

But back to the REC interface, I wrote a driver service to gather the data and write it to the DBUS, modified the systemcalc script to process the data for the GUI and added / changed the QML pages to display the data, so basically patched and extended the whole thing. All done in scripting, no compiling necessary.

Would you mind posting a link to the specs of the packets sent by the REC (0x351, 0x355, 0x356, 0x35A, 0x35B, 0x35E, 0x370)? I believe these follow so-called Sunny Island format, but I've not found a full description -- obviously you did.

Many thanks.

[CatNewBee]

Originally Posted by SweareDeep

Would you mind posting a link to the specs of the packets sent by the REC (0x351, 0x355, 0x356, 0x35A, 0x35B, 0x35E, 0x370)? I believe these follow so-called Sunny Island format, but I've not found a full description -- obviously you did.

Many thanks.

The complete protocol for the serial interface is described in the pdf user manual of the BMS.
Page 6 to 8...

[CatNewBee]

here....

[SweareDeep]

Originally Posted by CatNewBee

The complete protocol for the serial interface is described in the pdf user manual of the BMS.
Page 6 to 8...

I had assumed (incorrectly) that you were using the CAN interface rather than the RS-485 interface and that you would therefore know the answer to my question about the particular packet ids. I should add there is some documentation on a few of these packets out there but not on all of them, including whatever packet contains individual cell voltages.

So, there's two ways to skin this cat and the one that's documented seems like the better choice!

Thanks.

[Care2go]

As we are preparing to move aboard our boat, this has been excellent to follow and read. Thank you CNB for a well documented system and breakdowns. Much appreciated.

I plan to convert our FP 44 similar to yours except we don’t have a generator on board. I intend to offset that with enough solar and battery bank to not be concerned.

Thanks again