DIY Battery Is Alive !!

[johnb2713]

Looks great !

Any figures on how much energy the heater uses for keeping the batteries warm ?

The heater hasnt run at all apart from at the very start. We've had -4 degC and 40w heater managed to keep the battery compartment (empty) to 10 degC with just simple 6mm insulation board in a wooden enclosure. This showed a 31% duty ratio so the heatloss on my wooden box is around 13w.

After installing the batteries, they remained between 12 deg C and 19 deg C during operation, the heater never coming on. Theyve sat at about 9 deg C above ambient temperature.

Also how are you getting on with the 5000kva victron. Do you find you are always pulling annoying extra power from the grid ? With my multiplus II 8000 when it’s on grid and the house pulls more than 6.5kw the additional comes from the grid. When it’s off grid it can make 9.5kw sustained (although complains about overload)

I havent noticed this but it's early days.

I do have rather odd charging behaviour though. My Growatt Hybrid charges in straight line from 0 to 90% and then the charge rate reduced as you would expect. The Victron starts in a similar way but every half hour or so the charge rate reduces considerably for 5 minutes or so and the resumes back to the 4.4kW this continues until the 80% mark where the max charge rate reduces but still the same every 15 minutes or so a reduction for 5 minutes. Theres no alarms, it's reading the % correctly from the BMS, it's really odd.

T

 

[shaunuk]

Great, the heater sounds like just the ticket, I need to heat a server rack surrounded by 50mm PIR, its in a garage so mostly at outside temperatures.
Could I ask what thermostats your using, are they 240v din rail type ?

Re charging, that does sound strange: Three possible reasons that immediately come to mind:

- Do you have some loads pulling from the inverter during the night ? perhaps the Victron is conflicting with the growatt, maybe try turning off the growatt one night as a test.
-Another possibility is battery BMS, i'm using pylon tech alongside the Victron and the charging is constant until the last 10%, ive experimented with different charing currents and lower ones timed to finish as late a possible are most consistent. The pylontech batteries and Victron use DVCC to monitor current and voltage during the charge and control the inverter output. You could be getting into a BMS limiting the charge for heat reasons and then backing off until it cools a little. You could try disconnecting the BMS comms cable and charging to a particular voltage so see if its linear.
-Also, look at Victron temps during this charing, if your running the multiplus II 5000 ? its rated at 48v at 70a, well, thats 3360w, your trying to charge at 4400w continuously, it may be running in its overload mode and trying its best and then backing off because its getting to hot.
Try lowering charing current to 50a and then increasing every night until it starts backing off again.

I have a Chinese 150a bms to try out soon, just need to pick up 200kg of battery cells from my mates garage but the family won't let me leave the house for half a day over Christmas to collect.

 

[johnb2713]

Great, the heater sounds like just the ticket, I need to heat a server rack surrounded by 50mm PIR, its in a garage so mostly at outside temperatures.

Could I ask what thermostats your using, are they 240v din rail type ?

I use STC1000 temperature controllers, on ebay circa £7. Easy to setup and can have really tight control for heating of cooling

https://www.ebay.co.uk/itm/31401034...l04AyPDyNIN9unqb4U0tQKIwedog|tkp:BFBMgMK2-Klh

Re charging, that does sound strange: Three possible reasons that immediately come to mind:

Do you have some loads pulling from the inverter during the night ?

No we have loads off the grid direct, 2 x EV, Tumble drier, Washing m/c, Immersion heater, 7kWh Growatt Battery and 14 kWh DIY Battery + the standard house load, we can get to around 22 kW peak.

perhaps the Victron is conflicting with the growatt, maybe try turning off the growatt one night as a test.

I can give that a try, tonight looks good, everything charged fully today from PV. We will use a bit during the evening.

-Another possibility is battery BMS, i'm using pylon tech alongside the Victron and the charging is constant until the last 10%, ive experimented with different charing currents and lower ones timed to finish as late a possible are most consistent. The pylontech batteries and Victron use DVCC to monitor current and voltage during the charge and control the inverter output. You could be getting into a BMS limiting the charge for heat reasons and then backing off until it cools a little. You could try disconnecting the BMS comms cable and charging to a particular voltage so see if its linear.

I did think something could be getting hot but later as the charge current reduces it still continues to happen at the same rate. Certainly even at full charge currents nothing is getting even warm yet alone hot. I may try the comms cable idea and see what happens.

-Also, look at Victron temps during this charing, if your running the multiplus II 5000 ? its rated at 48v at 70a, well, thats 3360w, your trying to charge at 4400w continuously, it may be running in its overload mode and trying its best and then backing off because its getting to hot.
Try lowering charing current to 50a and then increasing every night until it starts backing off again.

Thats a good shout, I just assumed it would run to it's maximum and no more, I'll turn the max down now and see how it goes.

I have a Chinese 150a bms to try out soon, just need to pick up 200kg of battery cells from my mates garage but the family won't let me leave the house for half a day over Christmas to collect.

View attachment 14665View attachment 14666

Glad were not at 40 deg C, I'll keep an eye on it all.
Thanks for the info and suggestions, much appreciated.

 

[Sean4S]

Regarding the strange charging behaviour, this tech note from Victron might explain what you are observing:

https://www.victronenergy.com/uploa...ting-operating-temperature-and-efficiency.pdf
Having worked in a company that built domestic battery storage products, we found that most inverters protect themselves by derating output power when they get too hot. By doing so, it allows the temperature of the inverter to drop. As the inverter cools, power is increased. This behaviour can be cyclical as the temperature slowly increases again after power is increased.

Sounds like a great project. Thanks for sharing.

 

[johnb2713]

Regarding the strange charging behaviour, this tech note from Victron might explain what you are observing:

https://www.victronenergy.com/uploa...ting-operating-temperature-and-efficiency.pdf

I've downloaded this had a look and submitted it to my archives, thank you.

I dont think this is heat related for a couple reasons, the temperature in the little shed at the side of the inverter has less than 10 degC and I would have expected if it was heat related it would reduce it's output at the higher currents which it sustains for 2 hours or more before it reduces the output and then continuous to increase and decrease the output on a fairly regular basis even at a charge current of 10 amps. I would have expected it to be more steady state current at lower rates if heating was the issue.

Having worked in a company that built domestic battery storage products, we found that most inverters protect themselves by derating output power when they get too hot. By doing so, it allows the temperature of the inverter to drop. As the inverter cools, power is increased. This behaviour can be cyclical as the temperature slowly increases again after power is increased.

Yes, I can understand that and we will maybe experience this in the summer if we get the 42 deg C again.

Sounds like a great project. Thanks for sharing.

My pleasure, maybe it may help someone else to pluck up the courage, it took me a while

 

[Sean4S]

I've downloaded this had a look and submitted it to my archives, thank you.

I dont think this is heat related for a couple reasons, the temperature in the little shed at the side of the inverter has less than 10 degC and I would have expected if it was heat related it would reduce it's output at the higher currents which it sustains for 2 hours or more before it reduces the output and then continuous to increase and decrease the output on a fairly regular basis even at a charge current of 10 amps. I would have expected it to be more steady state current at lower rates if heating was the issue.


Yes, I can understand that and we will maybe experience this in the summer if we get the 42 deg C again.


My pleasure, maybe it may help someone else to pluck up the courage, it took me a while

John, I hope you figure it out.

Just came across this post:

https://community.victronenergy.com...ent-current-cause-seplos-bms-to-shutdown.html
Do you know if it has been fixed?

 

[johnb2713]

Do you know if it has been fixed?

I can only assume it has, I've started mine up many times and it has never failed. Seplos tech support is excellent, one their engineers created a Whatsapp group so he can help directly / remotely online.

J

 

[Ucatchmydrift]

Thank you, I am pleased with the end result, my next one will have a bespoke stainless steel enclosure.

Thank God it worked or i'd be responsible for the failure LOL


Yup, mine is currently charging from the sun at 1.8 kW, absolute result.


I have heard that Octopus arent taking people on for EV tariffs but I'm also told if you ring them up it's no problem. You have to go onto a standard tariff first and then transfer to GO etc.


Excellent, that looks good, where did you go for the batteries, was it Fogstar?

What BMS / Inverter combination are you using? any comms issues?

J

I spoke to octopus a few months back and they were taking on new people, when i phoned again a few weeks later, they said no, not any more
BUT, utilita are now doing a plan which is just 5p kw/h economy 7. With no standing charge!! They get the money back by charging 90p kw/h for the first peak hour used, then about 60p kw/h the rest of the peak hours.
But, for me/us, the fact that i never have to use any peak elec anymore means max bills of £12 a month based on 8kw a day usage (im on my own in a 2 bed bungalow with a decent wood burner), and thats in the winter NOT taking into consideration any solar production! Absolutely amazing!

Yes i got the batterys from fogstar, the 280ah grade a. (now just £1150 for 8, = ~7.2kw)
I have 16 now, built as two seperate banks linked in parallel via some copper busbars, both using JK bms's at 200amps each. No problems so far, and have 2amps of balancing capability if needed and a good bluetooth app. Have a epever 3kw all in one inverter, not grid tied, but with pass through ability. Not used any bms comms as dont really need it, not sure the inverter will accept the protocol from jk anyway..

Using a shunt with relay control to switch grid supply on and off to the inverter via a contactor and wifi smart plug so it only charges from grid when i want it to (economy 7 hours for example) - this works really well..

 

[johnb2713]

John, I hope you figure it out.

Sad but true, I'm sat here at 02:00 this morning watching the battery charge LOL. I know what is causing the strange charging behaviour.
The first issue is the calibration of the BMS, it thought I had 150a/h whereas I actually have 280a/h. I have changed one of places this appears in the SEPLOS data file and calculating the power input it confirms it's correct with the % SOC which the Victron inverter uses.
The major cause of the regular reduction in charge rate is one cell tripping the high voltage alarm at 3.5v. This seems a little low, the EVE spec sheets and data suggests 3.65v. Unfortunately the SEPLOS software wont let me change the value. Looking at the one cell thats slightly high compared to the others I have 90mV difference between highest and lowest, the difference is reducing over time which suggests the balancer is doing it's job.
I'll watch it over the next few days and see how it goes.

 

[Sean4S]

Sad but true, I'm sat here at 02:00 this morning watching the battery charge LOL. I know what is causing the strange charging behaviour.
The first issue is the calibration of the BMS, it thought I had 150a/h whereas I actually have 280a/h. I have changed one of places this appears in the SEPLOS data file and calculating the power input it confirms it's correct with the % SOC which the Victron inverter uses.
The major cause of the regular reduction in charge rate is one cell tripping the high voltage alarm at 3.5v. This seems a little low, the EVE spec sheets and data suggests 3.65v. Unfortunately the SEPLOS software wont let me change the value. Looking at the one cell thats slightly high compared to the others I have 90mV difference between highest and lowest, the difference is reducing over time which suggests the balancer is doing it's job.
I'll watch it over the next few days and see how it goes.

John, I really would prefer it if instead of calling it sad, you called it passionate. I don’t want to fall into the category of sad old fart - passionate I can handle

Well done on the diagnosis.

I’m going to be lazy and quote
“there is no power above 3.45....
you should never go above 3.6 ( 3.55 preffered ) in day to day charging, or below 2.7 in discharging, as this will prologue cell life at no cost to KWH available”

 

[Sean4S]

John, I really would prefer it if instead of calling it sad, you called it passionate. I don’t want to fall into the category of sad old fart - passionate I can handle

Well done on the diagnosis.

I’m going to be lazy and quote
“there is no power above 3.45....
you should never go above 3.6 ( 3.55 preffered ) in day to day charging, or below 2.7 in discharging, as this will prologue cell life at no cost to KWH available”

also

“your cells last longer if used between 20% and 80%.”

source: https://diysolarforum.com/threads/seplos-can-bus-rs485-48v-200a-8s-16s-bms.20051/page-14

Page 15 of the same also details the calibration process if you have not already figured it out.

Thanks again for sharing

 

[Ucatchmydrift]

also

“your cells last longer if used between 20% and 80%.”

source: https://diysolarforum.com/threads/seplos-can-bus-rs485-48v-200a-8s-16s-bms.20051/page-14

Page 15 of the same also details the calibration process if you have not already figured it out.

Thanks again for sharing

Eve cells are fine from about 10% to 99% state of charge. Anything above about 3.5v causes degredation and from about 3.45-3.65v gives you a total of about 1% soc. Just not worth charging above about 3.45v-3.5v. And are quite happy down to 10% or a bit less according to the experts. Have a watch of will prowse on youtube, seems to know his stuff better than most.

 

[shaunuk]

Lots of super info here. Sounds like a manual top balance may be in order John. Put in parallel and then charge to 3.45 perhaps ?

 

[Ucatchmydrift]

Lots of super info here. Sounds like a manual top balance may be in order John. Put in parallel and then charge to 3.45 perhaps ?

For a proper top balance, should always charge to 3.65v to get all the cells balanced as best as possible. Then just charge to 3.45v afterwards. Wont hurt them unless charged to 3.65v many times..

 

[Sean4S]

Eve cells are fine from about 10% to 99% state of charge. Anything above about 3.5v causes degredation and from about 3.45-3.65v gives you a total of about 1% soc. Just not worth charging above about 3.45v-3.5v. And are quite happy down to 10% or a bit less according to the experts. Have a watch of will prowse on youtube, seems to know his stuff better than most.

What do you mean when you say that the cells will be fine from 10% to 99%? I know they will work but are you saying the degradation won’t be any worse than implementing say 20% to 80%?
The datasheet says “Recommended SOC window 10%~90%”.

https://batteryfinds.com/wp-content...O4-LFP-Battery-Cell-Product-Specification.pdf

 

[binky]

What do you mean when you say that the cells will be fine from 10% to 99%? I know they will work but are you saying the degradation won’t be any worse than implementing say 20% to 80%?
The datasheet says “Recommended SOC window 10%~90%”.

https://batteryfinds.com/wp-content...O4-LFP-Battery-Cell-Product-Specification.pdf

Battery life is dependent on charge cycles, and that is dependent on depth of discharge. The thing that has given lithium batteries longer life of say 6000 cycles is not discharging them fully, it makes a big difference. By not going below 10-20% it puts a lot less stress on the cells. It's not that you can't do that occasionally, but if you do drop below those sort of figures it takes an age to get the battery trickle charged back up to the level where you can put a decent charge rate in.

 

[Ucatchmydrift]

What do you mean when you say that the cells will be fine from 10% to 99%? I know they will work but are you saying the degradation won’t be any worse than implementing say 20% to 80%?
The datasheet says “Recommended SOC window 10%~90%”.

https://batteryfinds.com/wp-content...O4-LFP-Battery-Cell-Product-Specification.pdf

Pretty much, yes, there will be some degredation, but its so minimal, its not worth worrying about. I used to think 20-80% also, based on the very data sheet you have posted, but after much reading and watching videos from people with lots of experience and testing in the real world, i now go with the generally accepted soc of 10%- 99% is good for a long battery life. (10-20 years discharging each day)

This guy has ALOT of lifepo4 experience :
https://diysolarforum.com/threads/r...e-for-diy-lifepo4-batteries-sticky-post.5101/
His youtube chanel:


Also, this guy does alot of testing also, not going through all his videos to find the info, but it is a very good resource:


Diysolarforum.com is extreemely handy also as there is alot of experience on there about this sort of information.

A typical lifepo4 voltage chart for example:



I think the temperature has a bigger affect on battery life for lifepo4 than charge profiles.

 

[johnb2713]

John, I really would prefer it if instead of calling it sad, you called it passionate. I don’t want to fall into the category of sad old fart - passionate I can handle

The latter I can just about cope with

Well done on the diagnosis.

I do enjoy tracking faults down and resolving issues.

I’m going to be lazy and quote
“there is no power above 3.45....
you should never go above 3.6 ( 3.55 preffered ) in day to day charging, or below 2.7 in discharging, as this will prologue cell life at no cost to KWH available”

Seplos has 3.5v as alarm high and 2.9v as alarm low and you cant adjust them, they're locked (well they are until Seplos tell me how to get into them).

 

[johnb2713]

also

“your cells last longer if used between 20% and 80%.”

source: https://diysolarforum.com/threads/seplos-can-bus-rs485-48v-200a-8s-16s-bms.20051/page-14

Page 15 of the same also details the calibration process if you have not already figured it out.

Thanks again for sharing

My intention once I have everything working correctly is to drop back to 15% to 85%.
The calibration instructions for mine was to charge at 0.5C until the BMS switches off and then discharge at 0.5C until the BMS switches off. I did attempt to do that with the Victron Multiplus as the charger and the load but I could'nt fool it into running until the BMS switched off, too many alarms etc. I dont have a suitable PSU or load to do it any other way.

 

[johnb2713]

Lots of super info here. Sounds like a manual top balance may be in order John. Put in parallel and then charge to 3.45 perhaps ?

Looking at the balance in situ, they dont look too bad, maybe the BMS has done it's but slowly? I'll check again tomorrow after another charge cycle.



It does still display rated capacity as 150AH but the % and actual capacity remaining is correct and ties up with the load etc.