270ah DIY LiFePO4 build

[GHA]

OK, another thing to check I guess
I'm logging at 1min intervals so haven't noticed anything like that. OTOH, my panels are on the h/t nothing shading them so wouldn't see such drops anyway.

I´m logging @ 1s intervals, 1m wouldn't really flag up what's going on. Pi Zero2 doesn´t even break into a sweat

 

[geem]

This is my only data logging. A snapshot at 0700 in the morning

 

[GHA]

This is my only data logging. A snapshot at 0700 in the morning

you poor data starved thing
Though never met anyone as obsessed with measuring things as I seem to be. Can take ages to track down just which particular device charger is causing the 200mA cyclic currrent ripple overnight
The revelation didn't really happen for most cruisers though..
Will be interesting getting a load cell on the chain & compass recording, who knew poverty could be so much fun!!!!!

 

[geem]

you poor data starved thing
Though never met anyone as obsessed with measuring things as I seem to be. Can take ages to track down just which particular device charger is causing the 200mA cyclic currrent ripple overnight
The revelation didn't really happen for most cruisers though..
Will be interesting getting a load cell on the chain & compass recording, who knew poverty could be so much fun!!!!!

Have you ever been diagnosed with OCD

 

[vas]

I most likely collect much more data than GHA, but I never ever considered myself a OCD person I'll ask wife, but I think she'll start laughing...

btw, I do have a few loadcells and figured how I could use one to measure how empty the gas tank is haven't installed it yet though, too low on the to do list.
Had a discussion with Neeves re measuring loads on anchor chain, not particularly easy and not really an automated way of doing so, so atm I'll pass. If you find a decent way I'd like to know how!

 

[GHA]

Have you ever been diagnosed with OCD

Has it's benifits
Though part of it came from feeling very exposed down in south america with everything breaking & no real way to get any decent kit. Limped up to Tobago with dead batteries, bust windlass & no engine. Want to know long beforehand if anything is getting a bit tired.

 

[GHA]

If you find a decent way I'd like to know how!

Had the bare bones of what look's useful working, load cell from a broken kitchen scale with an ESP32 sending data over wifi to the Pi. Accurate to under a gram but the multiplier to load will be huge. Then a very shallow bridle to the chain load with the loadcell at the node. Same concept as a rope tension meter which I've used a lot in the past rigging live events/theatre. They are expensive & need calibrating a lot. Even if a small load cell wasn't accurate as long as it was fairly linear & repeatable it would still put out very useful data. Loads on the go atm though but one of these evenings might have a blast & put it on github.

 

[GHA]

so guys, what charge settings are you using? I'm 100% on solar & have been terminating bulk at 13.8v then 'float which isn't really float' 13.37v
Smartshunt doesn't get reset so reckon it's drifted SOC loads, too lazy to retype so details are here >

What do you think SOC is charging to 3.45v

SOOOOOOOO happy with these things, lifechanger living onboard.

 

[Poey50]

What do you think SOC is charging to 3.45v

Given time, SOC at that voltage should reach 100%

 

[geem]

so guys, what charge settings are you using? I'm 100% on solar & have been terminating bulk at 13.8v then 'float which isn't really float' 13.37v
Smartshunt doesn't get reset so reckon it's drifted SOC loads, too lazy to retype so details are here >

What do you think SOC is charging to 3.45v

SOOOOOOOO happy with these things, lifechanger living onboard.

We are using 28v, so 3.5v/cell and the same float as you. Absorbtion is 30 minutes.
We only get to 28v about once per week. We just make more water or more hot water and keep the battery condition lower. When we do get to 28v, the battery balance is doing very little. The 280Ah CATL cells are very good

 

[GHA]

Given time, SOC at that voltage should reach 100%

For a while I had 2h absorption but it seemed to just 'float' but at a higher voltage. Goes straight to 13.37v 'float' now. TBH really not keen to go to 100%, no need really.
Anyone with a smartshunt can log data like this for a 20 quid raspberry pi zero & openplotter with no programming needed.
Surprised how far out the smartshunt was, 60% when they were really up closer to 100% though that's months without a reset. Think I'll try manually resetting to 95% ater bulk getting to 14.2v(3.55v) once a week or 2 & live with that. The idea of a black and white bang on SOC measurement is a bit of a pipe dream anyway though much less so than with lead acid.

No balancing & after must be 4/5 months now cells are still within a couple mV of each other. A days balancing might go on the rather large "once a year" boat checksheet.

Was just interested how others had theirs set up, thnx.

 

[GHA]

Goes straight to 13.37v 'float' now.

Changing the victrons just now & missed it, when it hits that top knee the voltage spikes FAST..

 

[GHA]

Say hi to GHA from me hope he's doing good!

re monitoring lifepo4, after spending the last couple of months on diyBMS, which is an opensource project, I'd urge him to have a look (either for ideas or to join in!)
1+A passive balancers per cell (attiny controller on each)
esp32 main board
attiny powered current shunt
testing main board and balancers now for 2months, brilliant integration to my victron only setup onboard (working a bit on editing the rules passed on via canbus to victron atm)

cheers

V.

Only popped in for a second, still groundhog day in here then
Anyway, been monitoring my LiFePo4 cell voltage with a home made board & eager to push the button to get one made, any comments from the few on here would be welcome, not that I know what I'm doing

Easyeda schematics below. If anyone is interested enough to have a chat come over to the github page, issues I suppose -
GitHub - boatybits/esp32_lifepo4_meter at beta
or discord
Join the Openplotter Chat Discord Server!
in random chat.

In short, 4 channel opamp measure the voltage of each cell, ads1115 measures the voltage the an esp32 sends it out over wifi. Maybe BLE if it works.

This is the cobbled together version which has been running for maybe a year now.



Calibration is a bit out but really don't think it matters, voltage isn't that useful, when each cell hits the top knee is!


Huge help from perplexity comet which is their version of chromium with the perplexity Ai built in. Point the browser at the easyeda page & upload the circuit diagrams & BOM & it will hunt around all the datasheets & come up with suggestions like decoupling capacitors, which type & why it's a good choice with links to the data sheet pages. Then get it to hunt for the cheaper "basic" parts in jlcpcb of there are any. Crazy useful. Knew there were some chats relevant in this thread so told it to search the whole thing, clicking the next button didn't work so told it to edit the ip address instead to get to all the pages.

Right, feel dirty already Outa here... come talk cheap high end monitoring!