Be cautious ordering generic blue wrap LFP cells

[hour]

Oh yea, some additional testing last night... I was charging through the MPPT @ 15 amps (210w) with a 24v power supply - how I've been charging this thing during testing lately. I was down almost 60ah when I started charging so the process took about 4 hours.

Started at 70*F BMS (atop the cells) and 68*F from the BMS probe, dangling on the edge of cells and towards the bottom of the can. Put the lid on.

When charging ended the BMS was reading 82*F and the probe was 77*F. The MPPT was way hotter than 82*F but didn't reduce output at all, so I'm guessing beneath 104*F (think Victron says that's when output decreases)

This will not work at all in the summer (even though I'll never see 15a in from 200w solar), the bed of my truck in July was frequently 110*F. I have one vent directly behind the MPPT but may need to add another next summer, or an intake fan. I just don't want to cut any more holes in the box right now since the objective for the next 6 months will be keeping it warm. If the other incubator heater works out then I could leverage the separate wiring for fan vs heater and use it as a bit of an intake, gapped about 1/4" off the wall of the ammo can with a plastic louver vent through can. For summer mode I could adjust params in software and set the thermostat to an impossibly high temperature (or disconnect it entirely) and pull in a little air from outside the can. In winter just throw a square of reflectix over the vent and change settings back + re-enable thermostat.

What a pain in the ass all of this is... and for pretty much no reason. I don't venture far from camp in the winter and could have easily heated the truck bed with my portable diesel heater on morning 3 of a trip to kick start it in to accepting charge. Otherwise it'd take me being in the hospital for a month to drain the battery, assuming it was below charge temp that entire time. That's assuming the fridge wouldn't cycle much/ever in those temps, and given a < 4 watt continuous draw from the router and particle photon + arduino nano (peak of 4 when downloading large files on two wireless clients)... ~7ah/day.

FWIW I'm aware that I'm TL;DR boss, I just type myself through tasks since dogs don't offer much in terms of conversation. Not trying to continuously bounce ideas off people, but it helps me and in the future maybe this thread will help someone else planning a build. Hell of a lot better than OP ghosting with no resolution.

 

[john61ct]

A 12x12" 50 watt pad is about 170$. Not cheap, but not crazy expensive. The big thing is you can pick your wattage.

To me, crazy expensive. should be well under $50 each, ordering under 20.

 

[Rando]

I think you may be over thinking the the thermal management issues. Both the BMS and the Victron MPPT have thermal cut offs, which can prevent charging below 0C. With these running you won't hurt the battery, but may loose out on some low temperature charging opportunities. I don't know where you are planning on using this, but unless you are planning a lot of COLD camping I would just run it without a heater and see how it goes.

My 150Ah battery is installed in my FWC which sits outside with solar year round. I am sure there are some cold days when the thermal cutoffs prevent solar charging, but overall the system does just fine and sits at ~13.2-13.3V all year. When we are actually using the system, we heat the camper, so charging is no issue. After 2+ years the battery is still delivering over 150Ah (measured) so I don't think I have done any damage.

 

[Rando]

PS my bluetooth BMS arrived. It is a nice $40 upgrade from the $20 BMS I have been using for the past couple of years. It will be interesting to see how the coulomb counting on the BMS compares to the Victron BMV.

 

[hour]

Stuff

re: post below - the arduino+nano+huawei b310s-518 router consume 3-4w on the BMS. Never more than 4, never less than 3. The BMV will fluctuate between 3-4w and 0w and has been zero current calibrated. It's possible that neither the BMV nor the BMS are sensitive enough and that the BMS is only sensitive enough to report > 0 because of the hair of a load placed on it by the BMV. I hope you find yours to be accurate. I've now perfected the BMS reading code on particle photons, arduino platform, and ESP devices.. if you end up wanting to nerd out let me know.

Yeah, what you said is exactly the use case of the ammo can. The battery is in the bed of the truck and isolated from the vehicle charging system (solar or generator->power supply only), and not inside an area that is climate controlled (aka popup camper trailer that I'll be in this winter).

So the truck sits outside in whatever weather we have. The goal is that the fridge and microcontrollers will continue to run indefinitely, though the fridge will be using little to nothing in the coming weeks.

Or I could stop being a lazy SOB and take stuff out of the fridge at the end of winter trips. Disable discharging (which unfortunately will cut off solar), only re-enable a day or two before a trip if it's a sunny. Then night before departure top it up if it didn't get warm enough in those few days to allow a top off. Colorado is weird and will give a day above 45 at least once a week during winter. The occasional 60* and -10* week here and there.

I probably won't be out for a week in -10*F but 40*F yes. And with tame consumption compared to summer and more capacity then I could ever use, it is overthinking. I have ways to heat the truck bed in winter doomsday to get the battery warm enough to enable charging, otherwise I doubt my draw + my time out would get below 50%. Get home, plug in and charge to storage voltage, unload fridge, disable discharge. The BatteryProtect I omitted from the project would have come in handy given a shared charge/discharge port on this BMS.

------

100w heater arrived and was drawing 65-68w including power supply losses through 120v->12v brick. A nice warm breeze, enough CFM to not be scorching hot an inch away from it.

Installed.

Can was 77 inside (been charging-discharging with lid on) so i set the heat-to temp to 77 + 4 = 81*F. Got there in a few minutes while still charge-discharging, 15a in 15a out, so the MPPT was producing a bit of heat.

Opened lid of can and all the stuff it was blowing heat in to first was obviously warmer than the far side of the can. The blue sea fuse block, mppt, and router are all above the heater which is currently in the bottom of the can in the empty space.

Some of the wires going to the fuse block were pretty damn hot, which makes this not ideal - cept I was heating from 77 to 81. If those cables were 100*F then what will they be when I'm using it in real world conditions, just trying to float in the 45*F range? Certainly not as hot as they were.

This could work. I am a little alarmed though that after reaching temp I disconnected load. I'm drawing 98w from heater + microcontrollers + router, or 94w from the heater+fan. I don't really understand that. Yeah it's advertised as 100w but in 20m of free flow operation for my bench test I never saw more than 68w including 120v->12v brick losses. Would work well but damn, can I trust it going from 68w to 98w draw?

 

[john61ct]

If I had those conditions, anywhere near freezing, I'd be bringing the bank inside my living space, certainly before charging.

Be very careful about charge rates vs cold, even cool temperatures.

Also if the cells start out cold, could easily take a couple of hours before can/air heating up penetrates to where it counts.

 

[Rando]

Maybe I am wrong, but doesn't the BMS allow you to set separate charge/discharge temperature limits? I set the charge on mine to 2C and the discharge to -15C. You can leave the solar connected, when it is warm enough it will charge, when it is too cold it won't. My experience is that in Colorado, you get plenty of sunny and warm days year around. But if the challenge is part of the fun, then go for it.

I am guessing your power issue is due to the voltage drop. The power brick maybe something like 11.5 under load, whereas your battery is 13.3 under load, which would lead to a 25% difference in power to a resistive load.

 

[hour]

If I had those conditions, anywhere near freezing, I'd be bringing the bank inside my living space, certainly before charging.

Be very careful about charge rates vs cold, even cool temperatures.

Also if the cells start out cold, could easily take a couple of hours before can/air heating up penetrates to where it counts.

Wish you had said that earlier, had been wondering for the last few pages. If not constantly maintaining the temperature, how do you prevent superficial readings on the BMV that allow charging when heat has been introduced and risen to the top? Sensor will be at the top + terminal unless you flip your pack sideways.

I hear you on cells staring out cold, which they would with my plan to only warm the ammo can when I'm sensing that the day is starting and it's below charging temps. That's all to avoid needlessly heating the box... if it's 6pm and dark outside what's the point? Blowing hot air at ice cubes will result in a warm surface temperature and a frozen core...

I can enable pretty accurate time tracking with this setup and wrap the whole temperature control logic in a if(it's-between-530pm-and-5am) { don'tHeat = true} type deal, setting up a routine for preheating the box over a longer duration starting at say... 5am. But I'm still skunked if the day ends up being crap and I can't even replenish what it took to heat it in the first place. That was the big objective in being able to read panel voltage.

Won't be carrying pack inside, will trust BMS low temp charge disconnect + BMV communication low temp charge disconnect + my own memory extra disabling MPPT before I'd go through that hassle.

 

[hour]

Maybe I am wrong, but doesn't the BMS allow you to set separate charge/discharge temperature limits? I set the charge on mine to 2C and the discharge to -15C. You can leave the solar connected, when it is warm enough it will charge, when it is too cold it won't. My experience is that in Colorado, you get plenty of sunny and warm days year around. But if the challenge is part of the fun, then go for it.

I am guessing your power issue is due to the voltage drop. The power brick maybe something like 11.5 under load, whereas your battery is 13.3 under load, which would lead to a 25% difference in power to a resistive load.

Think you're right, worked that in to the amazon review I just left for the heater. If the brick starts at 12.22v and sags, and the battery is at 13.6v and doesn't sag at all (was charging during this) then I kinda get it. Makes the 150w heater that hit 190whatever watts seconds in to operation certainly suspect.

Yeah so the separate charge/discharge restrictions on the BMS... the two I purchased were common in/out. Unless the BMS itself is capable of doing this internally, I don't see how that functionality can be achieved in a common in/out... I chalked that up to the mobile app not knowing the capabilities of each individual BMS you talk to..

 

[john61ct]

Wow, talk about entitled attitude. You think I'm parsing the posts closely, as if this is a paid gig?

I haven't been, to be honest since the cell commisioning, all these design & build ideas, I have no idea if workable or not,

just quickly scanning here and there, trying to help out with an occasional quick heads up .

These variable curves of cell damage (longevity hit) vs charge C-rate vs cell temperatures are not at all well understood even by top PhD researchers in the field, every chemistry formulation is different.

So, just saying, LFP is not good in the cold, automated systems are fragile, consider using personal attention and care if not instead, then in addition to sensors and heaters and microcontrollers.

 

[hour]

Wow, talk about entitled attitude. You think I'm parsing the posts closely, as if this is a paid gig?

I haven't been, to be honest since the cell commisioning, all these design & build ideas, I have no idea if workable or not,

just quickly scanning here and there, trying to help out with an occasional quick heads up .

These variable curves of cell damage (longevity hit) vs charge C-rate vs cell temperatures are not at all well understood even by top PhD researchers in the field, every chemistry formulation is different.

So, just saying, LFP is not good in the cold, automated systems are fragile, consider using personal attention and care if not instead, then in addition to sensors and heaters and microcontrollers.

Hah! No, I don't think it's a paid gig or that input is expected. I even went as far as to mention at the bottom of one of my posts a few hours ago why I blab so much - but I understand this isn't a paid gig and you probably didn't catch that. I'll bold text next time in hopes that I don't get called entitled . I don't expect input from anyone - I'm just documenting my build and talking myself through problems, whether I resolve them or not. If people would like to chime in, great. That doesn't mean I can't wish I had your input on temperature control practices before now. Not much to be found googling 'environment control battery box lithium' so your contribution would have represented approximately 33% of the feedback I've received.

In any event, yes - I agree. I have the ability to shut it down remotely and this has been quite the initiative for me, which I won't soon neglect. I can surely take care of the pack even if it remains in the truck and sits idle for days, weeks. That has always been an option, my drive to automate is to simply pass the time and improve my skills surrounding hardware interaction. This is hobbyist activity. I'm not whacking it in the middle of the ocean on a sailboat. I bought the cells because I won on a scratch ticket and wanted something to ******** around with. My outgoing DIY LFP was fine for my purposes and the setup would have likely been just fine all winter making no changes.

 

[Rando]

Think you're right, worked that in to the amazon review I just left for the heater. If the brick starts at 12.22v and sags, and the battery is at 13.6v and doesn't sag at all (was charging during this) then I kinda get it. Makes the 150w heater that hit 190whatever watts seconds in to operation certainly suspect.

Yeah so the separate charge/discharge restrictions on the BMS... the two I purchased were common in/out. Unless the BMS itself is capable of doing this internally, I don't see how that functionality can be achieved in a common in/out... I chalked that up to the mobile app not knowing the capabilities of each individual BMS you talk to..

The BMS knows which way the current is flowing, so it can operate as an 'ideal diode' - turn off the MOSFETs when the current is flowing into the battery, turn them on when the current is flowing out. This is pretty standard functionality for the drop in lithiums which don't have separate charge/discharge connections. It would be easy enough to test this......

But anyway, occasional charging at low rates (~ 0.1C) at moderately cold temperatures (0 to -15C or so) is not going to dramatically decrease the cycle life of your batteries. There is research on this, but you need to look for it.

 

[hour]

The BMS knows which way the current is flowing, so it can operate as an 'ideal diode' - turn off the MOSFETs when the current is flowing into the battery, turn them on when the current is flowing out. This is pretty standard functionality for the drop in lithiums which don't have separate charge/discharge connections. It would be easy enough to test this......

But anyway, occasional charging at low rates (~ 0.1C) at moderately cold temperatures (0 to -15C or so) is not going to dramatically decrease the cycle life of your batteries. There is research on this, but you need to look for it.

Thanks for the clarification, I wanted to ask that on endlesssphere's topic surrounding the chinese BMS units but I'm not registered there and the conversation seems to has lost steam. I was tempted to test this early on but didn't want the BMV to reset. A lot of this has been centered around that exact thing - another layer between BMS and BMV + MPPT so that they seldom-if-ever get shut off.

 

[Rando]

Here is an interesting plot of lithium battery degradation as a function of charge rate and temperature:

Source

At a charge rate of ~0.1C, there is no degradation down to -10C. If you are charging at rates >0.5C temperature is a significant issue, even above freezing.

 

[john61ct]

Even within the LFP family, between say Winston cells and CALB cells, there are huge differences wrt specific number values.

Extrapolating between LI research and what the IRL reality for LFP might be, can only be done in the most general terms.

One sweeping statement seems clear - the relationship between (potential) lost cycles increases with either factor exponentially, not linear.