DIY 200ah LFP underperforming

[Luk3jay]

Hi, I built a 200ah LFP battery at the start of the year (see other post on here) and it's been performing pretty well.

It lives in my camper and it's recharged by 350w solar and a 30a dc-dc charger. It has a constant load of around 2a as well as the fridge which pulls an additional 5a for a few mins when it turns on.

I also have an induction hob which pulls around 100a when used, but is very rarely used for longer than 20 mins.

The battery was nearly always above 50% SOC, usually closer to 80% and with a good voltage.

After accidently running it to 0% a few weeks ago, I tweaked some of the BMS parameters and now the battery seems to be a much smaller capacity, even though the BMS says it has a 200ah capacity. Last night for example, with only the fridge and general electrics on, it drained from 85% to 45%. There was definitely not 80a pulled during the 7 hours we slept.

The general voltage now seems to be around 12.5-13v and drops to between 11.8 and 12.2 under heavy load (100a)

Does this sound right? Or perhaps the symptoms of failing cells? Or perhaps incorrectly setup BMS?

The BMS is set according to the cell parameters. I'll prepare screenshots of the settings when I can.

N.b there isn't any extra load that I wouldn't know about. And I have a display connected to the BMS so can always check the current draw. It's never over 7a unless the cooker is on, but is at 2a around 80% of the time.

Thanks

 

[Steve_S]

Depends on what changed in settings I'd say.

 

[mikefitz]

The battery should show a higher resting voltage than 13volts, it could be the battery is not being fully charged.
More details, type of BMS and cell protection voltages, charger and charge settings.
Running the battery down may have caused damage if the cell volts were drained very low. ( drain of BMS display?)
Having an independent battery monitoring other than the BMS is useful.

Mike

 

[Sojourner1]

What are your charging parameters? Also how is SOC capacity reset... is your SOC actually the percent it claims?

 

[HRTKD]

I agree with mikefitz, it doesn't sound like your battery is actually fully charged. You'll have to come up with answers to the questions posted in the thread before can help much more.

 

[740GLE]

Is it a 12v or 24v system?

If it’s a DIY battery are you seeing any cell hit LVD or HVD before any others?

 

[Luk3jay]

Hey all, sorry for my silence! I've been without internet. I'll reply below. Thanks for your help!

The battery should show a higher resting voltage than 13volts, it could be the battery is not being fully charged.
More details, type of BMS and cell protection voltages, charger and charge settings.
Running the battery down may have caused damage if the cell volts were drained very low. ( drain of BMS display?)
Having an independent battery monitoring other than the BMS is useful.

Mike

What should the resting voltage be? it tends to be around 13v. Currently the BMS monitor screen shows 13.05V with a load of 6.2A and a 71% SOC
The BMS screen stayed on and working the whole time. What battery monitor would you suggest? I decided to use the BMS display (mounted in the van wall) as I thought that would be the best and most accurate.

What are your charging parameters? Also how is SOC capacity reset... is your SOC actually the percent it claims?

I think these are my parameters. I took this screenshot when I programmed the controller but don't have my lead on me to double check (I'm abroad)
N.B Please disregard the over/undervolt connections. They're not relevant to the charging and I'm not using the load terminals on the solar charger.
I also have a DC-DC charger and mains charger on board. both are set to 14.2v charging voltage.

The SOC does seem to be accurate but I have no historical data so it's hard to gauge (I'm looking for something similar to, but cheaper to the simarine system and will probably end up building my own monitor but that is a long time off)

Is it a 12v or 24v system?

If it’s a DIY battery are you seeing any cell hit LVD or HVD before any others?

It's 12v and there didn't seem to be and LVD or HVD. it just cut off when it reached 0%. Sadly I have no historical data.

These are my BMS settings AFTER the change. This was now about a month ago and I can't remember what or why I changed. The one that jumps out is that I changed the 'pack capacity' to 160ah from 200ah. Have I just found the answer to my own question?
The heat was 42*c (which is a lot for someone from Northern England) and my brain was incredibly foggy. I think I also changed the Pack over and under voltage settings. Any advice on what they should actually be would be great as there aren't any actual instructions so I'm trying to work it out as I go along. Thank you so much for your help!

 

[TacomaJoe]

What does your cell balance look like at different SOC?

 

[Luk3jay]

What does your cell balance look like at different SOC?

Currently 3277-3282 at 50% SOC. I'll be full soon so will update when charged. I'm trying to stay alive 50% on the road but will update when it drops. Thanks

 

[Luk3jay]

3266-3274 at 40% SOC. (taken straight after a 100a load which is I think why there's a big cell difference)

3219-3223 at 37% SOC (charging at 3a)

3027-3052 at 42% SOC with a 90a load.

 

[TacomaJoe]

I agree with others, you're not getting a full charge. Your cells have good balance so I think the battery is fine. Your cell voltage vs SOC parameters match mine (factory default, I think). I'd run the FLOAT voltage to 14.2 and let it charge up, watching cell voltage (not SOC) to get a good charge on the battery.

 

[Browneye]

BMS is generally not a good indicator of SOC unless or until it gets fully charged, and then it will show the 100%. Battery voltage just isn't a good indicator of SOC, and BMS will lag depending on loads and charging. Same for a shunt meter - it needs to get to full 100% charge then reset the meter/indicator, if it doesn't do so automatically.

Cell voltages are low - 30-40% seems about right. It could well be your solar panels are not bringing in enough juice to cover your loads AND get the battery fully charged.

What is a 'hob' for 100A? If it runs 20 minutes that's 33Ah just itself. Figuring 80% production out of your 340W of PV, let's say they're bringing 20-25A, you could use nearly an hour of your charging time just to put that back. Generally a solar day is about 6 hours, but kind of depends on if you have tilting or flat-mounted panels, or any shading.

I run a 230A system with 430W of PV array, and an extra 160W I can set out as portables. Similar draws, but no 'hob', whatever that is. The battery gets fully charged by the end of each day.

If I set my charging that high I get cell over-voltage and charging disco at high knee.

Here's what I run on my Tracer Xtra 3210N for in-use, when we're using the coach. When it's parked in storge, I have another file I load with lower values to just maintain with no loads.

 

[Luk3jay]

3266-3274 at 40% SOC. (taken straight after a 100a load which is I think why there's a big cell difference)

3219-3223 at 37% SOC (charging at 3a)

3027-3052 at 42% SOC with a 90a load.

3270-3276 at 90%.

BMS is generally not a good indicator of SOC unless or until it gets fully charged, and then it will show the 100%. Battery voltage just isn't a good indicator of SOC, and BMS will lag depending on loads and charging. Same for a shunt meter - it needs to get to full 100% charge then reset the meter/indicator, if it doesn't do so automatically.

Cell voltages are low - 30-40% seems about right. It could well be your solar panels are not bringing in enough juice to cover your loads AND get the battery fully charged.

What is a 'hob' for 100A? If it runs 20 minutes that's 33Ah just itself. Figuring 80% production out of your 340W of PV, let's say they're bringing 20-25A, you could use nearly an hour of your charging time just to put that back. Generally a solar day is about 6 hours, but kind of depends on if you have tilting or flat-mounted panels, or any shading.

I run a 230A system with 430W of PV array, and an extra 160W I can set out as portables. Similar draws, but no 'hob', whatever that is. The battery gets fully charged by the end of each day.

If I set my charging that high I get cell over-voltage and charging disco at high knee.

Here's what I run on my Tracer Xtra 3210N for in-use, when we're using the coach. When it's parked in storge, I have another file I load with lower values to just maintain with no loads.

Thanks. When I'm next with my download lead I'll tweak my settings and report back.

I have never noticed the resting voltage above 13.5v, even when using the mains or the dc-dc charger which both charge at 14.2v, but I'll double check next time. I'll be doing a lot of driving in the coming days so I'll check when it's at 100%

The solar generally gets 5-25a depending on weather, but generally replenishes what I use in 24hrs plus 10-20%. On days with less solar, I use less power.

FYI, this is a hob.

 

[Browneye]

Yes, 13.5 is full resting voltage for LFP. In fact, you can get full charge with 13.8, it just takes a long time for that last 5%.

Boost duration can be adjusted based on charge voltage as well. The higher the volts, the shorter the duration should be. It gets there either way, it's just faster if you use a higher charge voltage, but if you have cell runaway then a lower voltage prevents that.

Hob? Had to google it. Evidently it's a euro thing. Those 120V burners seem to be all the rage for campers. I like my LPG suburban range thank you. ?

 

[Luk3jay]

Yes, 13.5 is full resting voltage for LFP. In fact, you can get full charge with 13.8, it just takes a long time for that last 5%.

Boost duration can be adjusted based on charge voltage as well. The higher the volts, the shorter the duration should be. It gets there either way, it's just faster if you use a higher charge voltage, but if you have cell runaway then a lower voltage prevents that.

Hob? Had to google it. Evidently it's a euro thing. Those 120V burners seem to be all the rage for campers. I like my LPG suburban range thank you. ?

Thanks. So it sounds like my mains and Dc-Dc charger are set too high at 14.2v then?
As for the solar, I'll have to re evaluate it all when i can connect the laptop as I didn't think that lfp used 'boost' charging modes.

 

[Luk3jay]

SOC update:
We used loads of power last night and took the battery from 60% down to 22% when we went to bed at around midnight. I didn't get a cell range reading at 22% unfortunately.

When we woke up around 7am, the bms was showing 0% and had disconnected discharging.

My tracker gives me intermittent voltage readings and shows that we were at around 12.87v at 10pm.

And at 1am we were at 12.45v


With the BMS cutting out shortly before 2am.


There was only a 2a load and the SOC reading was 22%at 10pm so something isn't right. The bms screen is annoyingly basic so doesn't show me why it disconnected, but I would guess it is due to it believing that the SOC is 0% which is perhaps based on pack voltage?

Thank you very much for all your help so far! I'm actually really enjoying learning about this.

 

[Browneye]

What is the 'auxiliary battery'? Is that just what your lfp battery is called?

13.8 is fully charged - 12.45 is way discharged. You either have some kind of parasitic draw, or a failing cell. I'm not understanding a cause for your use situation.

What is kicking in and charging in the middle of the night? Or is that daytime, with solar gain?

My thought would be to do a load test with a known load for a duration that tells you usable amp hours. If it's way short you can go looking for a failed cell, although they appear to show well balance at low knee. [shrug]

 

[Luk3jay]

What is the 'auxiliary battery'? Is that just what your lfp battery is called?

That's my lfp battery. The tracker also shows 'main battery' which is the engine battery.

13.8 is fully charged - 12.45 is way discharged. You either have some kind of parasitic draw, or a failing cell. I'm not understanding a cause for your use situation.

What should the discharged voltage be?

What is kicking in and charging in the middle of the night? Or is that daytime, with solar gain?

If you are referring to the right hand side, ignore that. That is the solar coming in in the early hours, but not being strong enough to charge the battery (but it does power the tracker)

My thought would be to do a load test with a known load for a duration that tells you usable amp hours. If it's way short you can go looking for a failed cell, although they appear to show well balance at low knee. [shrug]

That is a good idea. It turns out that the Bluetooth module for the battery has a memory and event log! I'm going to leave it connected for the next few days and try to get a full power cycle. Hopefully it'll help out!

 

[Browneye]

I just looked up the default for low voltage cutoff on JBD BMS from my Overkill manual:

Cell over voltage 3650 mV recover: 3500 mV time: 2 sec
Cell under voltage 2500 mV recover: 3000 mV time: 2 sec

Frankly, I think 2.5Vpc is too low - that's 10V for a 12V battery. But that's what mine was left at. I've never used the whole battery. [shrug]
Anything under about 3Vpc is a danger zone - you're at or under 10% capacity. If you go below regularly the cells will severely degrade.

I have bounced it off the top a few times, but have adjusted boost charge voltage and charge termination so they don't.

Now in review I think I'll adjust my low voltage cutoff to 2.9Vpc with a recovery at 3Vpc.
Discharge cutoff creates all kinds of other problems, like turning off your charge controller, sensors, everything. Just a really bad idea for my installation.

 

[Luk3jay]

So I may have found it, There's this parameter which sets the battery to 0% at a predefined cell voltage. (2.9v)
There is also voltage targets(for want of a better word) for 100%, 80%, 60%, 40% and 20% So I assume that when the cell voltages hit one of those 'target' voltages, the SOC reading resets to that percentage regardless of the calculated load in/out (This means the BMS only has to keep track of 20% at a time instead of the full 100% of the battery, and it gets 'reset' at every 20% - not a bad idea.



20% is set to 3.266
0% is set to 2.900

according to your chart above, that means the battery would jump to 0% at around 10% SOC, so a 10% (20ah) use at 20% SOC would cause the pack to show fully discharged. Hrmm. I think I'm getting closer!

I'm going to tweak a few things and make a spreadsheet to track progress. Now I've got a history graph, things should be a lot easier. Thank you, the chart above is especially useful!

I just looked up the default for low voltage cutoff on JBD BMS from my Overkill manual:

Cell over voltage 3650 mV recover: 3500 mV time: 2 sec
Cell under voltage 2500 mV recover: 3000 mV time: 2 sec

Frankly, I think 2.5Vpc is too low - that's 10V for a 12V battery. But that's what mine was left at. I've never used the whole battery. [shrug]
Anything under about 3Vpc is a danger zone - you're at or under 10% capacity. If you go below regularly the cells will severely degrade.

I have bounced it off the top a few times, but have adjusted boost charge voltage and charge termination so they don't.

Now in review I think I'll adjust my low voltage cutoff to 2.9Vpc with a recovery at 3Vpc.
Discharge cutoff creates all kinds of other problems, like turning off your charge controller, sensors, everything. Just a really bad idea for my installation.

They're close to the settings that mine has.
I think I'll slightly lower them to match you.